* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Acknowledgement - ResearchSpace@Auckland
Survey
Document related concepts
Psychopharmacology wikipedia , lookup
Compounding wikipedia , lookup
Polysubstance dependence wikipedia , lookup
Neuropsychopharmacology wikipedia , lookup
Pharmacogenomics wikipedia , lookup
Neuropharmacology wikipedia , lookup
Theralizumab wikipedia , lookup
Prescription drug prices in the United States wikipedia , lookup
Drug design wikipedia , lookup
Pharmaceutical industry wikipedia , lookup
Nicholas A. Peppas wikipedia , lookup
Prescription costs wikipedia , lookup
Pharmacognosy wikipedia , lookup
Drug discovery wikipedia , lookup
Transcript
http://researchspace.auckland.ac.nz ResearchSpace@Auckland Copyright Statement The digital copy of this thesis is protected by the Copyright Act 1994 (New Zealand). This thesis may be consulted by you, provided you comply with the provisions of the Act and the following conditions of use: x x x Any use you make of these documents or images must be for research or private study purposes only, and you may not make them available to any other person. Authors control the copyright of their thesis. You will recognise the author's right to be identified as the author of this thesis, and due acknowledgement will be made to the author where appropriate. You will obtain the author's permission before publishing any material from their thesis. To request permissions please use the Feedback form on our webpage. http://researchspace.auckland.ac.nz/feedback General copyright and disclaimer In addition to the above conditions, authors give their consent for the digital copy of their work to be used subject to the conditions specified on the Library Thesis Consent Form and Deposit Licence. Note : Masters Theses The digital copy of a masters thesis is as submitted for examination and contains no corrections. The print copy, usually available in the University Library, may contain corrections made by hand, which have been requested by the supervisor. Design and Characterisation of Niosomes for Ocular Delivery of Naltrexone Hydrochloride Hamdy Abdelkader Mohamed Abdelkader A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy, School of Pharmacy, The University of Auckland 2011 Abstract Abstract Background: Recent reports have demonstrated that topical ocular administration of naltrexone (an opioid growth factor antagonist) is able to reverse the diabetic complications on the cornea (diabetic keratopathy). The topical administration of naltrexone accelerates corneal wound healing, restores corneal sensitivity and enhances corneal epithelialisation in diabetic rats, rabbits and humans. Naltrexone can be considered as a new therapeutic agent for treatment of diabetic keratopathy. To our best knowledge, an ophthalmic formulation has not yet been developed for naltrexone hydrochloride (NTX) nor has it been properly formulated in a liquid dosage form. Niosomes are non-ionic surfactant vesicles and have the same closed bilayer structures and properties of the well-known phospholipid vesicles, liposomes. Niosomes are considered more chemically stable and more economically viable alternatives to liposomes. At the same time, they could offer the same advantages as vesicular ocular delivery systems such as prolonged precorneal residence time and enhanced corneal penetration. Objective: The aim of this work is to design and characterise niosomal formulations for ocular delivery of NTX having the convenience of being delivered as eye drops. Methods: An HPLC method was developed and validated for NTX; preformulation studies were performed to study the physicochemical properties of NTX, such as aqueous solubility, lipid solubility, melting behaviour, spectrometric identification and construction of the Arrhenius plots for predicting NTX stability. Two classes of non-ionic surfactants, sorbitan fatty acid esters (Span®) and polyoxyethylene alkyl ethers (Brij®), were selected and investigated for their ability to form niosomes and encapsulate NTX using different levels of cholesterol (a bilayer membrane stabiliser). Also, different membrane additives [dicetyl phosphate (DCP), Solulan C24 (C24) and sodium cholate (CH)] were studied to modify the physical properties of niosomes, such as their percentage entrapment efficiency (EE %), size, morphology, rheology and spreading ability characteristics. Four different niosomes; F-S60, FDCP, F-C24 and F-CH were prepared and evaluated for ocular irritation, using the hen’s egg chorioallantoic membranes (HET-CAM) test, bovine eye test and histological corneal examination. Finally, they were studied for in vitro NTX release; transcorneal permeation of NTX; and physical stability. Results and discussion: The preformulation studies showed that NTX is a hydrophilic drug with a log P value of 1.61 at 35oC. The main degradation pathway of NTX in aqueous solutions was found to be autoxidation. Span 60-based niosomes, consisting of Span 60: cholesterol 7:3 mol/mol (F-S60), demonstrated superior EE % for NTX. Cryogenic scanning electron microscope (Cryo-SEM) images of niosomes showed onion-like structures, indicating multilamellarity. Incorporation of the membrane additives into bilayer membranes produced both spherical niosomes and discomes (giant disc-like niosomes), as confirmed by confocal laser scanning microscope (CLSM). The vesicular incorporation of NTX imparted a protective effect against light-induced oxidation, and the prepared niosomes demonstrated minimal to no irritation potential; significantly controlled NTX release; enhanced transcorneal permeation of NTX through excised cow corneas compared with the aqueous NTX solution; and were physically stable at least for three months at 4oC. Conclusion: The developed niosomes were able to protect NTX chemically, impart a penetration-enhancing effect and control NTX corneal permeation through bovine corneas, and demonstrate good ocular tolerability, suggesting that they are potential ocular delivery systems for NTX. i Dedication Dedicated to My parents, brother Mohamed Abdelkader, My wife Zeinab, my son Omar, my daughter Malak and the 1000 young people who gave up their lives for Egypt’s freedom in 25th January revolution ii Acknowledgment Acknowledgement It is with a great moment of pride and relief that I come to the end of a challenging and exciting project of my PhD Research. The completion of this thesis could not have been possible without the help, support, encouragement and input of many people. Foremost, I would like to express my gratitude to the main supervisor Dr Raid Alany for his thorough guidance, mentorship, editing my manuscripts and financial support of the third year. I would also like to cordially congratulate him on the post of professor at Kingston University, London, UK. I also greatly appreciate the expertise and guidance provided by my co-supervisors Dr Zimei Wu and Dr Raida Al-Kassas for their keen interest, valuable input and encouragement. I would like to thank my advisors Prof. Sayed Ismail (Department of Pharmaceutics, Faculty of Pharmacy, Assiut University, Assiut, Egypt) and Dr Amal Kamal (Department of Pharmaceutics, Faculty of Pharmacy, Minia University, Minia, Egypt) for their help with scholarship application to the Egyptian government. My sincere thanks go to Mr Ahmed Abu Moussa and Dr Moustafa Ibrahim at the Egyptian Embassy in Canberra for connecting me to the Culture and Mission Bureau, Ministry of Higher Education, Cairo, Egypt. I would like to pay my immense gratitude to Professor John Show and the School of Pharmacy for providing me the travel grant for international conference which was an important part of my PhD process and the professional career. A big thank you goes to Mrs Satya Amirapu from Histology Lab. for preparing histological sections of the bovine corneas for my research project. Another big thank you also goes to Hilary Holloway (the Biological Imaging Research Unit), who helped me to get niosome images using confocal laser scanning microscopy. A big thank you also goes to Associate Professor Brian Palmer (Auckland Cancer Society Research Centre) for help interpret H1-NMR and mass spectra of naltrexone hydrochloride. I would like to acknowledge the Association for Research in Vision and Ophthalmology as the copyright holder to grant me a permission to reprint the figure “the Architecture of human iii Acknowledgment corneal nerves” in my PhD thesis. I also would like thank the head of the English Language SelfAccess Centre Dr Penny Hacker for her great help and assistance and Mr. Ben Jackson who helped with proofreading. I have been fortunate to have some of my dearest colleagues during my studies. Thank you guys Alvin, Darren, Akhlaq, Thilini, Sara and James. Finally, I would like to take this opportunity to acknowledge my family who have always been supportive and have never given up on me. iv Publications Publications to date arising from this thesis • Abdelkader, H., Kamal, A., Ismail, S., & Alany, R. ( 2 0 1 0 ) . Preparation of niosomes as an ocular delivery s ystem for naltrexone hydrochloride: Physicochemical characterization. Die Pharmazie, 65, 811-817. • Abdelkader, H., Patel, D., McGhee, C., & Alany, R. (2011). New therapeuti c approaches in treatment of diabetic keratopathy. Clinical and Experimental Ophthalomology, 39, 259-270. • Abdelkader, H., Wu, Z., Al-Kassas, R., Brown, J. E., & Alany, R. (2011). Preformulation characteristics of the opioid growth factor antagonist- naltrexone hydrochloride: stability and lipophilicity studies. Journal of Drug Delivery Science and Technology, 21, 157-163. • Abdelkader, H., Kamal, A., Ismail, S., & Alany, R. ( 2 0 1 1 ) . Design and evaluation of niosomes and discomes for controlled release ocular delivery of naltrexone h ydrochloride. Journal of Pharmaceutical sciences, 100, 1833-1846. Conference papers to date arising from this thesis • Abdelkader, H., McGhee, C., & Alany, R. (2009). Naltrexone from alcoholism to corneal wound healing: a formulation approach. Paper presented at Australasian Pharmaceutical Science Association (APSA) conference, Hobart, Australia. (Poster). • Abdelkader, H., Kamal, A., Ismail, S., & Alany, R. (2010). Naltrexone hydrochloride a new ophthalmic pharmaceutical: physichochemical characterisation. Paper presented at the 12th Conference on Formulation and Delivery of Bioactives, the University of Otago, Dunedin, New Zealand (Podium). • Abdelkader, H., Kamal, A., Ismail, S., & Alany, R. (2010). Development of a validated HPLC-method for naltrexone hydrochloride-encapsulated in niosomal formulations. Paper presented at the 12th conference on formulation and delivery of bioactives, The University of Otago, Dunedin, New Zealand. (Poster). • Abdelkader, H., Kamal, A., Ismail, S., & Alany, R. (2010). Effect of selected ionic and nonionic membrane additives on the entrapment efficiency of naltrexone hydrochloride in niosomes. (2010). Paper presented at the 37th Annual Meeting and Exposition of the controlled release Society (CRS), Portland, USA. (Poster). v Publications • Abdelkader, H., Kamal, A., Ismail, S., & Alany, R. (2010). Characterization of ocular naltrexone niosomes using differential scanning calorimetry (DSC) and cryo-scanning electron microscopy (Cryo-SEM). Paper presented at the 37th Annual Meeting and Exposition of the CRS CRS, Portland, USA. (Poster). • Abdelkader, H., Wu, Z., Al-Kassas, R., & Alany, R. (2010). Evaluation of niosomes and discomes for ocular delivery of an opioid growth factor antagonist. Paper presented at Australasian Pharmaceutical Science Association (APSA) conference, Brisbane, Australia. (Poster). • Abdelkader, H., Wu, Z., Al-Kassas, R., & Alany, R. (2011). In vitro release, transcorneal permeation and stability studies of an opioid growth factor antagonist (naltrexoneHCl) niosomes and discomes. Paper presented at the 38th Annual Meeting & Exposition of the CRS, National Harbor, Maryland, USA. (Poster) • Abdelkader, H., Wu, Z., Al-Kassas, R., & Alany, R. (2011). Bovine corneal opacity, permeability (BCOP) and histological examination for evaluating ocular irritation and toxicity of non-ionic surfactant vesicles (niosomes). Paper presented at the 38th Annual Meeting & Exposition of the CRS, National Harbor, Maryland, USA. (Poster). • Abdelkader, H., Wu, Z., Al-Kassas, R., & Alany, R. (2011). Niosomes as a potential ocular delivery system for the opioid growth factor antagonist-naltrexone hydrochloride: influence of vesicular encapsulation on photoxidation. Paper presented at the 3rd PharmSciFair, Prague, Czech Republic. (Poster). vi Content Content Abstract ............................................................................................................................................. i Acknowledgement .......................................................................................................................... iii Publications to date arising from this thesis .................................................................................... v Conference papers to date arising from this thesis .......................................................................... v Content .......................................................................................................................................... vii List of Figures ................................................................................................................................ xii List of Tables ............................................................................................................................... xvii List of Abbreviations .................................................................................................................... xix 1. General introduction .............................................................................................................. 1 1.1. Ocular bioavailability from ophthalmic dosage forms ..................................................... 1 1.1.1. Anatomical and physiological considerations ........................................................... 2 1.1.2. Pharmacokinetic considerations ................................................................................ 8 1.1.3. Drug and formulations factors ................................................................................. 10 1.2. Drug delivery systems providing controlled and continuous ocular delivery ................ 14 1.2.1. Ophthalmic inserts ................................................................................................... 15 1.2.2. Collagen shields ....................................................................................................... 16 1.2.3. Polymeric gels ......................................................................................................... 17 1.2.4. Colloidal delivery systems ...................................................................................... 20 1.3. Diabetic keratopathy and the available treatment options .............................................. 28 1.3.1. Physiology of the cornea ......................................................................................... 28 1.3.2. Diabetes mellitus and the cornea ............................................................................. 28 1.3.3. Corneal innervations ................................................................................................ 29 1.3.4. Mechanisms of corneal maintenance and healing ................................................... 33 1.3.5. Treatment of diabetic keratopathy ........................................................................... 34 1.4. Thesis aims and structure ................................................................................................ 45 vii Content 2. Preformulation studies of naltrexone hydrochloride ........................................................ 47 2.1. Introduction ..................................................................................................................... 47 2.1.1. Physical properties of naltrexone hydrochloride ..................................................... 48 2.1.2. Pharmacological uses and dosage for systemic administration of NTX ................. 48 2.1.3. Pharmacokinetics of NTX ....................................................................................... 48 2.1.4. Adverse reactions of systemic administration of NTX ........................................... 49 2.1.5. Warnings on the systemic administration of NTX .................................................. 49 2.1.6. Naltrexone as a new ophthalmic pharmaceutical .................................................... 50 2.1.7. Chemical stability of NTX ...................................................................................... 50 2.1.8. Oxidation ................................................................................................................. 53 2.1.9. Arrhenius plot and predicting drug stability ............................................................ 56 2.2. Chapter aims ................................................................................................................... 57 2.3. Materials and methods .................................................................................................... 58 2.3.1. Materials .................................................................................................................. 58 2.3.2. Methods ................................................................................................................... 58 2.3.3. Statistical analysis ................................................................................................... 64 2.4. Results and discussion .................................................................................................... 65 2.4.1. DSC study ................................................................................................................ 65 2.4.2. Spectrometric analyses of NTX powder.................................................................. 66 2.4.3. UV-spectrophotometric assay.................................................................................. 70 2.4.4. HPLC method validation ......................................................................................... 71 2.4.5. Forced degradation studies ...................................................................................... 74 2.4.6. Degradation products identification using ESI-MS ................................................ 81 2.4.7. pH-degradation rate of NTX.................................................................................... 83 2.4.8. Accelerated stability and chemical kinetic studies .................................................. 85 2.4.9. Solubility studies of NTX ........................................................................................ 87 viii Content 2.4.10. Determination of n-octanol/PBS distribution coefficient (D) and partition coefficient (P) of NTX ........................................................................................................... 87 Conclusion ................................................................................................................................. 89 3. Preparation of niosomes using biocompatible surfactants ............................................... 90 3.1. Introduction ..................................................................................................................... 90 3.1.1. Niosomes ................................................................................................................. 90 3.1.2. Non-conventional niosomes .................................................................................... 99 3.1.3. Niosomes as drug delivery systems ....................................................................... 100 3.2. Chapter aims ................................................................................................................. 103 3.3. Materials and methods .................................................................................................. 104 3.3.1. Materials ................................................................................................................ 104 3.3.2. Methods ................................................................................................................. 104 3.3.3. Statistical analysis ................................................................................................. 108 3.4. Results and discussion .................................................................................................. 108 3.4.1. Niosome formation and imaging using plane /polarised microscopy ................... 108 3.4.2. Niosome imaging using Cryo-SEM ...................................................................... 111 3.4.3. Niosome size measurements.................................................................................. 111 3.4.4. NTX EE% .............................................................................................................. 114 3.4.5. DSC studies ........................................................................................................... 121 Conclusion ............................................................................................................................... 126 4. Effect of membrane additives on the physical properties of niosomes.......................... 127 4.1. Introduction ................................................................................................................... 127 4.1.1. Bilayer membrane additives .................................................................................. 127 4.1.2. Surfactant and ocular delivery ............................................................................... 130 4.2. Chapter aims ................................................................................................................. 131 4.3. Materials and methods .................................................................................................. 132 ix Content 4.3.1. Materials ................................................................................................................ 132 4.3.2. Methods ................................................................................................................. 132 4.3.3. Statistical analysis ................................................................................................. 138 4.4. Results and discussion .................................................................................................. 139 4.4.1. Effect of selected membrane additives on NTX EE% .......................................... 140 4.4.2. Effect of preparation methods on NTX EE% ....................................................... 142 4.4.3. Niosome size and distribution measurements ....................................................... 144 4.4.4. Morphology and lamellarity .................................................................................. 147 4.4.5. DSC studies ........................................................................................................... 155 4.4.6. Surface tension measurements............................................................................... 160 4.4.7. Contact angle and spreading coefficient measurements ........................................ 161 4.4.8. Rheological properties measurements ................................................................... 162 4.4.9. Effect of niosomal encapsulation on NTX oxidation ............................................ 167 Conclusion ............................................................................................................................... 172 5. Evaluation of niosomal formulations for ocular delivery of naltrexone hydrochloride ...............................................................................................................................................173 5.1. Introduction ................................................................................................................... 173 5.1.1. Eye irritation tests .................................................................................................. 174 5.1.2. Toxicity of niosomes ............................................................................................. 179 5.1.3. Physical stability of niosomes ............................................................................... 180 5.2. Chapter aims ................................................................................................................. 181 5.3. Materials and methods .................................................................................................. 182 5.3.1. Materials ................................................................................................................ 182 5.3.2. Methods ................................................................................................................. 182 5.3.3. Statistical analysis ................................................................................................. 196 5.4. Results and discussion .................................................................................................. 197 x Content 5.4.1. Conjunctival (HET-CAM) test .............................................................................. 197 5.4.2. Cornea (bovine eye) test ........................................................................................ 200 5.4.3. Histopathological evaluation of the bovine corneas .............................................. 203 5.4.4. In vitro release studies ........................................................................................... 216 5.4.5. In vitro release kinetic studies ............................................................................... 220 5.4.6. Ex-vivo corneal permeation studies ....................................................................... 221 5.4.7. Physical stability of the prepared niosomes .......................................................... 227 Conclusion ............................................................................................................................... 234 6. General discussion, conclusion and direction for future research................................. 235 6.1. General discussion and conclusion ............................................................................... 235 6.2. Limitations and future direction ................................................................................... 241 References ................................................................................................................................... 245 Appendix 1 .................................................................................................................................. 286 xi List of Figures List of Figures Figure 1.1 Gross anatomy of the anterior segment of the human eye ............................................. 1 Figure 1.2 Factors limiting ocular bioavailability of an ophthalmic formulation following topical administration to the eye surface ..................................................................................................... 2 Figure 1.3 Structure of the precornal tear film ................................................................................ 4 Figure 1.4 Photomicrograph of H & E stained bovine cornea ........................................................ 4 Figure 1.5 Human corneal nerves penetration and distribution..................................................... 30 Figure 2.1 Chemical structure of NTX .......................................................................................... 48 Figure 2.2 Schematic of morphine autoxidation in aqueous solutions .......................................... 55 Figure 2.3 DSC thermogram of NTX ............................................................................................ 65 Figure 2.4 APCI mass spectra of NTX .......................................................................................... 66 Figure 2.5 FT-IR spectrum of NTX .............................................................................................. 67 Figure 2.6 1H-NMR spectrum of NTX in CD3SO, 400 MHz ....................................................... 68 Figure 2.7 UV-Spectrum of NTX (100 µg/ml) in PBS pH 7.4 ..................................................... 70 Figure 2.8 Standard UV-spectrophotometric calibration curve of NTX in PBS at pH 7.4 ........... 71 Figure 2.9 Representative HPLC chromatograms of standard NTX (0.4 mg/ml) solution and NTX resolved from Span 60: cholesterol: DCP-niosomal formulation ................................................. 73 Figure 2.10 PDA- spectra of the NTX peak acquired from 4 different positions throughout the eluted peak for peak purity assessment ......................................................................................... 74 Figure 2.11 Percentage (%) remaining of NTX after 24 h exposure to various stress conditions 75 Figure 2.12 HPLC chromatograms of the blank solution and aquous NTX solution (0.4 mg/ml) pH 10 after 24 h exposure to heat at 60oC ..................................................................................... 77 Figure 2.13 PDA spectra of NTX eluted at 4.16 min (A), the degradation product resolved at 3.7 min (B) and the second degradation product resolved at 3.3 min (C) after exposure to pH 10 at 60 o C ................................................................................................................................................... 78 Figure 2.14 HPLC chromatograms of the blank solution and aquous NTX solution (0.4 mg/ml) of pH 1.2 after 24 h exposure to heat at 60oC................................................................................ 79 Figure 2.15 HPLC chromatograms of the blank solution and aquous NTX solution (0.4 mg/ml) of PBS pH 7.4 after 24 h exposure to heat at 60oC ....................................................................... 79 Figure 2.16 HPLC chromatograms of the blank solution and aquous NTX solution (0.4 mg/ml) of PBS pH 7.4 after 24 h exposure to H2O2 (2% v/v) ........................................................................ 80 xii List of Figures Figure 2.17 HPLC chromatograms of the blank and aquous NTX solution (0.4 mg/ml) of PBS pH 7.4 after 24 h exposure to artificial daylight illumination (10,000 lux) at 40oC...................... 80 Figure 2.18 ESI mass spectra of product I and Product II of NTX oxidation ............................... 81 Figure 2.19 Schematic of the proposed oxidative degradation pathway for NTX ....................... 82 Figure 2.20 First-order degradation plots for NTX at 60oC in phosphate buffers of various pH values ............................................................................................................................................. 83 Figure 2.21 pH-rate profile for NTX at 60oC ................................................................................ 84 Figure 2.22 First-order degradation plots for NTX in PBS pH 7.4 at various temperatures......... 86 Figure 3.1 Locations of entrapped hydrophilic, lipophilic and amphipathic drugs in niosomes .. 91 Figure 3.2 Major types of niosomes, MLV (multilamellar vesicles), OLV (oligolamellar vesicles) and ULV (unilamellar vesicles)..................................................................................................... 91 Figure 3.3 Schematic of a single-chain surfactant ........................................................................ 93 Figure 3.4 Chemical structure of cholesterol................................................................................. 95 Figure 3.5 Suitable niosome sizes for particular routes of administration .................................... 98 Figure 3.6 Representative micrographs of Brij 72:cholesterol 7:3 mol/mol-based niosomes under plane light and polarised light microscope ................................................................................ 110 Figure 3.7 Niosomal dispersions containing various molar ratios of Brij 52:cholesterol and Span 60:cholesterol ............................................................................................................................. 110 Figure 3.8 Cryo-SEM micrographs of Span 60: cholesterol 7:3 mol/mol and Brij 72: cholesterol 7:3 mol/mol niosomes ................................................................................................................. 111 Figure 3.9 Effect of HLB on Span-based niosomes composed of surfactant:cholesterol at 7:3 molar ratio ................................................................................................................................... 113 Figure 3.10 Effect of gel/liquid transition temperature on EE% for niosomes composed of surfactant:cholesterol at 7:3 molar ratio ..................................................................................... 115 Figure 3.11 Effect of cholesterol concentration on EE% for Span-based niosomes .................. 117 Figure 3.12 Effect of the total surfactant/lipid content on EE % for Span 60: cholesterol ......... 118 Figure 3.13 Effect of the total surfactant/lipid content on DL % for Span 60: cholesterol ......... 119 Figure 3.14 Effect of the initial amount of NTX on EE % for Span 60: cholesterol .................. 120 Figure 3.15 Effect of the initial amount of NTX on DL % for Span 60: cholesterol .................. 121 Figure 3.16 DSC thermograms of Span 60-based niosomes containing different concentrations of cholesterol .................................................................................................................................... 124 xiii List of Figures Figure 3.17 Effect of NTX on gel/liquid transition temperature of (Span 60: cholesterol 7:3 mol/mol) niosomes ...................................................................................................................... 125 Figure 4.1 Equilibrium between forces acting on a drop of liquid on a solid surface ................. 135 Figure 4.2 KSV-CAM 101 goniometer setup for measuring contact angle ............................... 136 Figure 4.3 Effect of selected membrane additives concentration on NTX EE% for niosomes prepared using the TFH method ................................................................................................. 140 Figure 4.4 Effect of the preparation method on NTX EE% ........................................................ 143 Figure 4.5 Size-frequency distribution curves of the prepared niosomes ................................... 146 Figure 4.6 Cryo-SEM micrographs of F-S60 and F-DCP prepared using the REV method ...... 148 Figure 4.7 Cryo-SEM micrographs of F-C24 and F-CH prepared using the REVmethod ......... 149 Figure 4.8 CLS micrographs of F-S60 niosomal formulation loaded with CF produced by the REV method ................................................................................................................................ 151 Figure 4.9 CLS micrographs of F-DCP niosomal formulation loaded with CF produced by the REVmethod ................................................................................................................................. 152 Figure 4.10 CLS micrographs of F-C24 discomes loaded with CF produced by the REV method ..................................................................................................................................................... 153 Figure 4.11 CLS micrographs of F-CH niosomal formulation loaded with CF produced by the REV method ................................................................................................................................ 154 Figure 4.12 Effects of different membrane additives on gel/liquid transition temperatures of the prepared niosomes ....................................................................................................................... 155 Figure 4.13 Hypothetical position occupied by DCP in the bilayer membrane of F-DCP based on DSC ............................................................................................................................................. 157 Figure 4.14 Hypothetical position occupied by C24 in the bilayer membrane of F-C24 based on DSC ............................................................................................................................................. 158 Figure 4.15 Hypothetical position occupied by CH in the bilayer membrane of F-CH based on DSC ............................................................................................................................................. 159 Figure 4.16 Surface tension measurements for the prepared niosomes compared with NTX aqueous solution .......................................................................................................................... 160 Figure 4.17 Representative rheograms for the aqueous vehicle (PBS) and the prepared niosomes 35oC ............................................................................................................................................. 164 Figure 4.18 Viscosity values for the aqueous solution (PBS) compared with the prepared niosomes at two different temperatures 25oC and 35oC .............................................................. 165 xiv List of Figures Figure 4.19 Effect of niosomes encapsulation on the chemical stability of NTX against oxidation and daylight illumination............................................................................................. 168 Figure 4.20 First-order degradation kinetics for PBS solution of NTX and NTX encapsulated niosomes under artificial daylight illumination at 40oC in PBS pH 7.4 ...................................... 170 Figure 4.21 First-order degradation rate constants for PBS solution of NTX and NTX encapsulated niosomes under artificial daylight illumination at 40oC in PBS pH 7.4 ................ 171 Figure 5.1 Development stages of the growing embryos ............................................................ 183 Figure 5.2 Temperature, percentage relative humidity (% RH) and dew point inside the egg incubator over the 10-day incubation period ............................................................................... 184 Figure 5.3 Vascular responses used to score the test substances................................................. 187 Figure 5.4 Excised cow eyes immersed in normal saline with silicone O-rings centered on top of the cornea and incubated a water bath thermostatically equilibrated at 37oC ± 0.5oC ................ 189 Figure 5.5 Degree of corneal opacity and fluorescein permeability used to score the test substances .................................................................................................................................... 191 Figure 5.6 Degree of corneal opacity and fluorescein permeability used to score the test substances ................................................................................................................................... 192 Figure 5.7 Excised bovine cornea, modified parts and top view of the final assembly of the Franzdiffusion cell ............................................................................................................................... 195 Figure 5.8 Cumulative HET-CAM scores for the controls and representative examples of the test substances ................................................................................................................................... 198 Figure 5.9 Cumulative bovine eye scores for the controls and representative examples of the test substances .................................................................................................................................... 201 Figure 5.10 Photomicrographs of H&E stained corneal sections of negative control treated with saline for 30 s ............................................................................................................................... 204 Figure 5.11 Photomicrographs of H&E stained corneal sections of acetone-treated cornea for 30 s ..................................................................................................................................................... 206 Figure 5.12 Photomicrographs of H&E stained corneal sections of NaOH (0.5 M)-treated cornea for 30 s ......................................................................................................................................... 208 Figure 5.13 Photomicrographs of H&E stained corneal sections treated with CH Powder, C24 powder and the prepared niosomes for 30 s ................................................................................ 210 Figure 5.14 Photomicrographs of H&E stained corneal sections treated with the prepared niosomes for 1 h .......................................................................................................................... 212 xv List of Figures Figure 5.15 Photomicrographs of H&E stained corneal sections treated with the prepared niosomes for 3 h .......................................................................................................................... 213 Figure 5.16 Photomicrographs of H&E stained corneal sections treated with the prepared niosomes for 8 h .......................................................................................................................... 214 Figure 5.17 In vitro release profiles of NTX from the NTX solution and the prepared niosomes ..................................................................................................................................................... 217 Figure 5.18 Effect of sample volume on the diffusion of NTX from the aqueous solution ........ 218 Figure 5.19 Transcorneal permeation profiles of NTX from the NTX solution and the prepared niosomes using excised cow corneas ......................................................................................... 223 Figure 5.20 Relationship between NTX release rate and apparent permeability coefficient (Papp) for NTX from the NTX solution and the prepared niosomes ...................................................... 225 Figure 5.21 Relationship between contact angle and lag time from the NTX solution and four niosomal formulations ................................................................................................................. 226 Figure 5.22 Effect of storage temperature on NTX retention in the prepared niosomes over three months ......................................................................................................................................... 232 Figure 5.23 Effect of niosome composition on NTX retention over three months ..................... 233 xvi List of Tables List of Tables Table 1.1 Causes of decreased corneal sensitivity ........................................................................ 32 Table 1.2 Major attributes of insulin ............................................................................................. 37 Table 1.3 Major attributes of murine NGF .................................................................................... 39 Table 1.4 Major attributes of naltrexone ....................................................................................... 43 Table 2.1 Summary for HPLC conditions of NTX determination ................................................ 52 Table 2.2 Assignments of 1H-NMR spectra for NTX ................................................................... 69 Table 2.3 Reverse predicted concentrations, % recovery and regression coefficient (R2)............ 72 Table 2.4 Precision and accuracy data of the QC samples ............................................................ 73 Table 2.5 First–order kinetic parameters and correlation coefficient of NTX degradation in aqueous solution of PBS pH 7.4 at different temperatures............................................................ 86 Table 2.6 n-octanol-buffer distribution coefficient (D) and partition coefficient (P) and calculated pKa at different temperatures ....................................................................................................... 88 Table 3.1 Composition of the prepared niosomes using various surfactant: cholesterol molar ratios ............................................................................................................................................ 105 Table 3.2 Chemical structure, phase transition temperature, HLB and vesicle-forming ability of the used surfactants ...................................................................................................................... 109 Table 3.3 Effect of surfactant type and cholesterol level on the D [4,3] and EE % of the prepared niosomes ..................................................................................................................................... 112 Table 3.4 DSC parameters of Span 60-based niosome dispersions containing different concentrations of cholesterol ....................................................................................................... 123 Table 4.1 Codes and composition of the prepared niosomal formulations ................................. 132 Table 4.2 Chemical structure, molecular weight, CMC and phase transition temperature of the investigated additives .................................................................................................................. 139 Table 4.3 Effect of selected membrane additives on D [4,3] and span values for niosomes prepared by the REV method ..................................................................................................... 145 Table 4.4 Contact angle and spreading coefficient measurements for the prepared niosomes .. 162 Table 4.5 Tukey’s pair wise comparison of viscosity values for the prepared niosomes at 25oC and 35oC ...................................................................................................................................... 166 Table 4.6 Tukey’s pair wise comparison of oxidation and photolysis for the prepared niosomes ..................................................................................................................................................... 169 Table 5.1 Ocular tissue and score for assessing ocular irritation using the Draize test............... 174 xvii List of Tables Table 5.2 HET-CAM scoring system .......................................................................................... 186 Table 5.3 Classification of cumulative scores in HET-CAM ...................................................... 186 Table 5.4 Average dimensions of human and cow corneas ........................................................ 188 Table 5.5 Bovine eye scoring system .......................................................................................... 190 Table 5.6 Classification of cumulative scores in bovine eye test ................................................ 190 Table 5.7 Summary of HET-CAM and bovine eye test interpretations ...................................... 202 Table 5.8 Histological corneal lesions scores for controls and the prepared niosomes at different time points .................................................................................................................................. 215 Table 5.9 In vitro release parameters for NTX from the prepared niosomes compared with NTX aqueous solution ......................................................................................................................... 218 Table 5.10 Tukey’s pair wise comparison of Q2h, Q6h and DE%-12 for NTX from the aqueous PBS solution and the prepared niosomal formulations ............................................................... 219 Table 5.11 In vitro release kinetic parameters for NTX from the prepared niosomes ................ 221 Table 5.12 Steady state flux, apparent permeability coefficient (Papp) and tL for NTX from the aqueous PBS solution and the prepared niosomes though excised cow corneas ....................... 224 Table 5.13 Tukey’s pair wise comparison of steady state flux, Papp and tL for NTX from the aqueous PBS solution and the prepared niosomal formulations ................................................. 224 Table 5.14 Effect of ageing on the volume diameter (D [4,3]) for the prepared niosomes over three months at three different temperatures ............................................................................... 229 xviii List of Abbreviations List of Abbreviations ao Optimal polar head group area of an amphiphile Abs Absorbance ANOVA Analysis of variance AUC Area under the curve o C Degree Celsius cal Calorie(s) CH Sodium cholate CMC Critical micelle concentration cps Centipoises Cryo-SEM Cryogenic scanning electron microscope C24 Solulan C24 D Distribution coefficient Da Dalton δ Chemical shift DCP Dicetyl phosphate DSC Differential scanning calorimetry D [4,3] Volume diameter EE% Entrapment efficiency % FAT Freeze and thaw g Gram(s) γ Surface tension H&E Haematoxylin and eosin stain HET-CAM Hen’s egg test chorioallanoic membrane h Hour(s) ICH International Conference on Harmonisation xix List of Abbreviations k o K Kilo Degree Kelvin KCl Potassium chloride KH 2PO4 Potassium dihydrogen phosphate λ max Wave length of maximum drug absorption lc critical chain length of an amphiphile LD% Loading efficiency % Lipid vesicles Liposomes and niosomes Liposomes Phospholipid vesicles LOD Limit of detection LOQ Limit of quantitation LUV Large unilamellar vesicles min Minute(s) µl Microlitre MLV Multilamellar vesicles mm Millimetre µm Micromitre µg Microgram(s) µmol Micromole(s) µM Micromolar mM Millimolar mol/mol Mole per mole Mol wt Molecular weight mPa.s MilliPascal-second = 1centipoise n Release exponent NaCl Sodium chloride xx List of Abbreviations Na2 HPO4 Disodium phosphate Niosomes Non-ionic-surfactant vesicles nm Nanometre NTX Naltrexone hydrochloride O/W Oil in water P Partition coefficient Papp Apparent permeability coefficient PBS Phosphate buffer saline PDA Photodiode array pH Negative log hydrogen ion concentration R2 Determination (regression) coefficient ® Registered trade mark REV Reverse phase evaporation rpm Rotation (s) per minute s Second(s) SUV Small unilamellar vesicles TFH Thin film hydration θ Contact angle tL Lag time t1/2 Time required for 50% drug degradation t90% Time at which drug decomposed to 90% of its original concentration U 1 unit of insulin V Volt v Volume of the hydrocarbon chains of an amphiphile v/v Volume per volume W/O Water in oil xxi List of Abbreviations w/v Weight per volume w/w Weight per weight xxii Chapter 1….Literature Review 1. General introduction: challenges in ocular delivery, controlled ocular drug delivery systems and new therapeutic approaches in diabetic keratopathy 1.1. Ocular bioavailability from ophthalmic dosage forms Topical ocular drug delivery is a commonly used and preferred route for treating disorders that affect the anterior segment of the eye (the cornea, conjunctiva, sclera, iris and lens) (Figure 1.1). This is because of the rapid and local effects, accessibility to the ocular tissue, relative safety and patient acceptability, as well as the relatively lower risk compared with systemic routes of drug administration (Macha et al., 2003). Figure 1.1 Gross anatomy of the anterior segment of the human eye However, there are many anatomical and physiological barriers forming part of the eye’s natural defence (Lang et al., 2002). These barriers and protective mechanisms prevent the administered ophthalmic formulations from residing on the eye surface for enough time to allow complete drug absorption. This prevents a significant portion of the instilled dose from being absorbed by the ocular tissue (ocular bioavailability). Therefore, the ocular bioavailability of topically applied drugs from simple solutions is often less than 1% (Urtti, 2006b). The protective barriers and constraints limiting ocular bioavailability of topically administered drugs are outlined in Figure 1.2 and are discussed in this chapter. 1 Chapter 1….Literature Review Figure 1.2 Factors limiting ocular bioavailability of an ophthalmic formulation following topical administration to the eye surface 1.1.1. Anatomical and physiological considerations The corneal and the conjuctival-scleral routes are the two major routes for ocular drug absorption from topically administered ophthalmic dosage forms (Ahmed & Patton, 1987, 1989). 1.1.1.1. Corneal route The cornea is the transparent avascular dome-shaped structure at the anterior part of the eye (Figure 1.1). The anterior corneal surface is covered by the tear film and the posterior surface is bathed by the aqueous humour. The surface area of the cornea is approximately 1.3 cm2, forming one-sixth of the surface area of the globe. The cornea is thinnest at the centre and its thickness gradually increases towards the periphery, with mean thickness of 0.5 mm and 0.7 mm respectively. Anatomically, it is divided into five layers (Klyce & Beuerman, 1998) (Figure 1.4): 2 Chapter 1….Literature Review 1. Epithelium 2. Bowman’s layer 3. Stroma 4. Descemet’s membrane 5. Endothelium Although it is not a part of the cornea, the precorneal tear film is intimately associated with the cornea both anatomically and functionally (Macha et al., 2003) (Figure 1.3). Precorneal Tear film The surface of the cornea must be kept moist to prevent damage to the corneal epithelium and to maintain the optical quality of the cornea. Moisture and smoothing are provided to the ocular surface by the preocular tear film in conjunction with the spreading function of the eyelids during blinking. The tear film is approximately 7 µm thick and 6.5 µl in volume. It consists of three structurally identifiable layers (Figure 1.3) (Ehlers & Hjortdal, 2006; Klyce & Beuerman, 1998): 1. A thin superficial lipid layer (0.1 to 0.5 µm) secreted by meibomian glands retards evaporation from the preocular tear film, which prevents drying between blinks. 2. A thick (approximately 6 µm) middle aqueous layer supplied by lacrimal glands, into which a mucin-rich glycocalyx extends. 3. A thin (0.02 to 0.05 µm) inner mucous layer composed mainly of glycoprotein and produced by the conjuctival goblet cells and the lacrimal gland. The mucin layer ensures the wetting of the corneal and conjunctival epithelia. 3 Chapter 1….Literature Review Figure 1.3 Structure of the precorneal tear film Figure 1.4 Photomicrograph of H & E stained bovine cornea showing five layers [epithelium (EP), Bowman’s membrane (BM), stroma (ST), Descemet’s membrane (DM) and endothelium (EN)] 4 Chapter 1….Literature Review Corneal epithelium The entire epithelium is about 4 to 6 cell layers thick and represents 10 % of the corneal thickness. The epithelial cells are non-keratinised, stratified squamous and divided morphologically into three layers (Klyce & Beuerman, 1998): The outermost layer comprises non-keratinised, stratified squamous epithelial cells. The innermost layer is a basal cell layer. The basal cells are the only epithelial cells that undergo mitosis. The daughter cells thus formed push anteriorly and change their shape conforming to the middle epithelial layer, which is called the wing cell layer. The wing cells are a transition stage between basal cells and superficial squamous cells. The superficial cells are polygonal in shape and 40-60 µm in diameter and have unique specialisations that maintain the tear film and the barriers that separate the extracellular space of the cornea from the tears. The surface membranes of superficial cells are distinctive in two respects: 1. They have microscopic projections (microvilli, reticulations and microplicae). 2. The outer leaflet is thickened and supports an extensive fibrillar glycocalyx, also called the buffy coat. The cells of all epithelial layers interdigitate and are separated by a 10-20 nm intercellular space. Important properties of the superficial cells include adherent junctions and junctional complexes formed with laterally adjacent cells. These complexes consist of tight junctions (zonulae occludens) that surround the entire cell and resist the flow of fluid through the epithelial surface. If the aqueous humour passes into the stroma due to a defect in the tight epithelial junction, it is trapped within the epithelium, resulting in epithelial oedema (Ehlers & Hjortdal, 2006). Basal lamina and bowman’s layer The basal lamina is an extracellular secretory product of the basal epithelial cells. It forms a scaffold for the organisation of the epithelium and separates the epithelium from the stroma. Bowman’s layer is an acellular region lying just under the basal epithelial membrane. It composed of randomly organised collagen fibrils that merge into more organised anterior stroma. Bowman’s layer is very thin and functions as a dome-shaped structure anchored to the limbus. Bowman’s layer is a rigid structure and it shapes the anterior corneal curvature. Once damaged, its architecture many not be restored, leading to abnormalities in the corneal thickness and optical properties that could result in permanent vision deficit (Klyce & Beuerman, 1998). 5 Chapter 1….Literature Review Corneal stroma The stroma constitutes approximately 90% of corneal thickness and it is composed of collagen fibrils in parallel arrays which make up the 300 to 500 lamellae of the stroma. The lamellae extend from limbus to limbus. Collagen constitutes approximately 71% of the dry weight of the cornea and is the structural macromolecule providing tissue transparency and mechanical resistance to intraocular pressure. Nerve axons and their Schwann cells are found in the anterior and middle third of the stroma. The stroma can be considered a comparatively open structure allowing diffusion of solute molecules weighing about 500 kilo Dalton (kDa). However, it acts as a diffusion barrier to all lipophilic drugs (Klyce & Beuerman, 1998). Descemet’s membrane Descemet’s membrane is the thick basal lamina (~5 µm thick) secreted by the endothelium. It is loosely attached to the stroma. Descemet’s membrane increases in thickness as an inevitable consequence of ageing. On the other hand, many diseases might damage the endothelial cells (e.g. interstitial keratitis) and consequently this damage can result in overproduction of Descemet’s membrane materials (Klyce & Beuerman, 1998). Also, wrinkling of the Descemet’s membrane is a diabetic complication of the anterior segment of the eye (Schultz et al., 1981 ). Corneal endothelium The corneal endothelium forms a single layer, of approximately 4 to 6 µm thick, on the posterior corneal surface. These cells play a key role in maintaining corneal transparency through their transporting, synthetic and secretory functions. The endothelium maintains tight apposition with neighbouring cells, preventing excessive seepage of aqueous humour into the stroma (Klyce & Beuerman, 1998). As previously outlined, the corneal route has exclusive tight junctional complexes of epithelial cells which endow the cornea with tight barriers against chemical and microbial insults. This makes the corneal epithelium a significant barrier to the penetration of many drugs. However, paracellular (small ions and hydrophilic molecules) and transcellular (hydrophobic) diffusion is reported. Most topically applied drugs permeate via the transcellular route (Lee & Li, 1989). The conjunctival-scleral route is thought to be an alternative portal gate for ocular permeation of hydrophilic large molecules such as peptides and proteins and its anatomical and physiological features are discussed in the following section. 6 Chapter 1….Literature Review 1.1.1.2. Conjunctival-scleral route The conjunctival-scleral route is the major pathway for polar and large-molecular weight drugs to reach the intraocular tissues (Ahmed & Patton, 1989). For example, high molecular weight drugs such as inulin (mol wt 5500) can gain access to the iris and ciliary body by diffusion via the conjuctival-sclera route. Further, drugs permeated through this route gain access directly to the posterior segment of the eye (choroid, vitreous humour and retina) (Ahmed & Patton, 1987). Conjunctiva Conjunctiva is a thin and vascularised mucus membrane. It covers the anterior surface of the eye globe with the exception of the cornea, where the conjunctiva ends at the limbus bordering the cornea. The conjunctiva consists of 2-3 layers of non-keratinised epithelial cells but conjunctival epithelial cells are not tight as corneal epithelial cells. The conjunctiva is permeable to molecules up to 20 kDa, in contrast to the cornea which is impermeable to molecules larger than 5 kDa (Greaves & Wilson, 1993). Anatomically, the conjunctiva is divided into three areas: conjunctiva covering the interior of eyelids, called palpebral conjunctiva; conjunctiva which is folded to cover the fornix, called fornix conjunctiva; and conjunctiva covering the anterior surface of the globe, called bulbar conjunctiva. The conjunctiva of the eyelid and globe forms a continuous surface area of 18 cm2, 17 times larger than the cornea1 surface area (Watsky et al., 1988). The bulbar conjunctiva, combined with the subsequent layer (sclera), represents the non-corneal barrier to drugs topically applied to the surface of the eye. Sclera The sclera (the white part of the eye) is the outermost layer that covers the anterior surface of the eye. It protects the sensitive inner parts of the eye. The sclera is a firm and resistant structure (0.5-1 mm thick) composed of the same collagen fibrils as the corneal stroma. However, the sclera’s collagen fibres are arranged in an irregular network rather than a lattice structure. This makes the sclera appear opaque compared with the transparent cornea. The sclera constitutes the posterior five-sixths of the globe, whereas the cornea comprises the remaining one-sixth (Macha et al., 2003). Generally, the sclera is significantly more permeable to solutes than the cornea. Solute size has a more pronounced effect on the sclera than the cornea whereas drug lipophilicty predominantly controls drug corneal permeability (Ahmed & Patton, 1989). The conjunctival-scleral route can be a promising avenue for intra-ocular drug delivery of hydrophilic and/or large molecules. These 7 Chapter 1….Literature Review molecules can benefit more from the conjunctival-scleral route than the corneal route (Fattal & Bochot, 2006; Geroski & Edelhauser, 2001). Therefore, large bio-organic compounds such as proteins and bioactive oligonucleotides drugs can be absorbed via this route and gain access to the intraocular tissues for safer and more convenient topical treatment of posterior segment diseases (e.g. diabetic retinopathy) than invasive intravitreous injection route. 1.1.2. Pharmacokinetic considerations Topical ocular pharmacokinetics studies the main pathways of drug administration, elimination and ocular bioavailability such as tears dynamics, non-productive absorption and enzymatic degradation (Urtti, 2006a). 1.1.2.1. Tears and tears dynamics The precorneal tear film structure, composition and function were previously discussed and shown in Figure 1.3. Human tears also contain around 0.6 to 2% w/v of proteins. The proteins in the tear fluid are mainly albumin, globulin, lysozyme and lactoferrin (Mikkelson et al., 1973a; Nishida, 2005). Emotional stress, irritation and inflammatory diseases greatly influence the protein content in the tear fluid. The relatively high concentration of proteins coupled with a relatively rapid tear turnover leads to a considerable loss in the drug activity. For example, the mitotic activity of pilocarpine hydrochloride was tested in the presence and absence of added rabbit serum albumin. A 3-fold increase in the mitotic activity was obtained in absence of the protein compared with that in the presence of 3% w/v albumin. The mitotic activity decreased gradually depending on the available protein concentration (Mikkelson et al., 1973a). The normal volume of tears, present in the human eye is approximately 7-10 µl. The average basal secretion is 1.2-1.5 µl/min and the physiological turnover rate is 0.1-0.15 min-1. Upon instillation of fluid into the eye, the drainage of instilled solution proceeds at a rapid rate until the volume is back to normal volume, about 7 µl. Thus as the volume of instilled dose increases, the volume of solution lost by spillage on the cheek and drainage increases (Chrai et al., 1974; Chrai et al., 1973). The drop size is responsible for considerable loss of the drug and hence affects the ocular bioavailability. Furthermore, this drainage contributes to non-productive absorption and leads to unwanted side effects. Therefore, it is recommended that the size of the eye drop should be reduced from 50-75 µl/drop to 5-10 µl/drop to maximise drug activity from ophthalmic ocular dosage forms (Chrai et al., 1973). In support of previous findings, it has been found that the ocular bioavailability, estimated as the area under the miosis–time curve, in rabbits did not show 8 Chapter 1….Literature Review any significant differences after the treatment using a 30 µl instillate of 1% w/v pilocarpine hydrochloride solution and the spray delivery of 5 µl volumes of the 1% w/v solution of pilocarpine hydrochloride. This study demonstrated that a 6-fold reduction in the administered volume achieved an equivalent miotic response (Martini et al., 1997). 1.1.2.2. Non-productive absorption Non-productive absorption is the absorption of topically applied ophthalmic drugs into the systemic circulation due to absorption from conjunctival blood vessels and dominantly from nasal mucosa due to the naso-lacrimal drainage (Chang & Lee, 1987; Shell, 1982). Most of the instilled dose (approx. 70%) is lost due to drainage via the naso-lacrimal duct within the first 15 to 30 s after instillation (Shell, 1982). Non-productive absorption following ocular drug administration not only leads to decreasing the ocular bioavailability, but is also seriously responsible for systemic side effects. Systemic side effects were reported with many drugs topically applied to the eye surface. For instance, topical ocular administration of atropine (German & Siddiqui, 1970), timolol (Kaila et al., 1986), betaxolol (Polansky & Alvarado, 1985), epinephrine (Anderson, 1980), oxymetazoline (Duzman et al., 1983), pilocarpine (Urtti et al., 1985), flurbiprofen (Tang-Liu et al., 1984) and cyclosporine (Mosteller et al., 1985) has been associated with systemic side effects. In addition, serious cardiovascular and pulmonary side effects were observed with topical ocular administration of timolol (a widely used anti-glaucomic drug) (Nelson et al., 1986). It has been reported that the nasal mucosa accounts for 70% of the timolol systemically absorbed (Chang & Lee, 1987). This is probably due to the consequence of rapid drainage of the instilled solution into the naso-lacrimal duct and then the nasal cavity. The conjunctival systemic absorption of ocularly applied drugs is increased by the relative leakiness of the membrane, rich blood flow, and large surface area (Urtti & Salminen, 1993). The nasal cavity has a large surface area of 150-200 cm2 and a total volume of 20 ml in humans. The nasal cavity surface area is 13 times that of the conjunctiva and almost 200 times that of the cornea (Lee et al., 2002; Pontiroli et al., 1989; Urtti & Salminen, 1993). The contact area between the instilled drug solution and nasal mucosa increases with increasing the instilled volume of the eye drop. The drug can gain access to the systemic circulation via absorption from nasal mucosa lining the mucosal cavity and/or via the gastrointestinal tract. Therefore, the absorption of the drug from the nasal region is reproducible and comparable to parenteral drug administration. This 9 Chapter 1….Literature Review efficient systemic absorption has been utilised as a non-invasive route of systemic drug delivery of proteins and peptides (Lee et al., 2002). Punctal occlusion and eyelid closure following topical solution instillation are the only measure recommended to the patients to minimise systemic drug absorption via the nasolacrimal apparatus (Kaila et al., 1986; Passo et al., 1984; Zimmerman et al., 1984). However, neither approach is a satisfactory long-term solution due to patient incompliance. The formulation approach could be a more effective solution. For example, both charged and neutral multilamellar liposomes have been shown to decrease the ocular drainage of [111In]indium chloride and [99Tc] sodium pertechnate compared with the buffer solution (Fitzgerald et al., 1987b). The two isotopes encapsulated in liposomes demonstrated prolonged precorneal residence time. The suspending medium drains away independently of the liposomes. As the size increases the nasolacrimal drainage decreases due to the larger liposomes restricted at the inner canthus region. In another study, they concluded that multilamellar liposomes have been found to reduce effectively the precorneal clearance rate and offer an advantage as a delivery system compared with small unilamellar vesicles, as the lower entrapment efficiency of the latter negate any advantages gained due to higher surface area (Fitzgerald et al., 1987a). 1.1.2.3. Enzymatic degradation A number of enzymes are secreted and detected in the precorneal tear film. These include carbonic anhydrases (Lonnerholm, 1974), esterases (Lee et al., 1982), peptidases (Stratford & Lee, 1985), ketone reductase (Lee et al., 1988) and glutathione-conjugating enzyme (Saneto et al., 1982). Depending on the nature of the applied drugs, these enzymes can adversely affect the biological activity of the applied active drugs or can activate topically applied prodrugs (Lee & Li, 1989). 1.1.3. Drug and formulations factors 1.1.3.1. Physicochemical properties of the drug The three main drug properties governing corneal drug absorption are: 1. Lipophilicity as reflected by n-octanol/buffer partition coefficient. 2. Dissociation constant (pKa) which determines the proportion of drug in its preferentially absorbed form at a given pH. 3. Drug molecular size. 10 Chapter 1….Literature Review Firstly, drug lipid solubility (lipophilicity) is an essential property for drug permeation via the lipidic epithelial corneal membrane (Schoenwald & Ward, 1978). Drug lipophilicity is related to its chemical structure and can be assessed by measuring the partition coefficient (P). According to drug partition coefficients, β-adrenoreceptor blockers are classified into three classes: very lipophilic (P ranged from 1640 to 14200), lipophilic (P ranged from 76 to 249) and hydrophilic (P ranged from 1.46 to 59) (Schoenwald & Huang, 1983). Lipophilic β-blockers, such as timolol, propranolol, atenolol, metoprolol and practolol, are able to lower the intraocular pressure and are used successfully as antiglaucomic agents, whereas other β-blockers, such as nadolol and sotatol, appear not to. This could be ascribed to the marked difference in lipid solubility and their ability to cross corneal barriers. The corneal epithelium is the rate-determining barrier for hydrophilic drugs whereas it is not for lipophilic drugs. Lipophilic drugs penetrate the cornea more rapidly (Huang et al., 1983). For instance, dipivalyl epinephrine ester (apparent P = 4.89) has been found to be about 100 times more effective than epinephrine (apparent P = 0.0081) in the management of glaucoma and about 100-400 times weaker than epinephrine in affecting the cardiovascular system in dogs and cats (McClure, 1975; Wei et al., 1978). This is due to enhancement of lipophilicity of epinephrine through the use of dipivalyl ester which can facilitate penetration through the lipoidal layers of the cornea (Lee & Li, 1989). Topically administered drugs, however, must possess adequate aqueous solubility in tear fluid in order to achieve an adequate penetration rate, to diffuse across the water-filled corneal stroma and to gain access to deeper ocular tissues (Schoenwald & Huang, 1983). In another report, the optimal P for corneal drug absorption was found to be in the range of 10-100 (Kaur & Smitha, 2002). Secondly, ionised drugs, such as weak acids and weak bases, can exist in both ionised and unionised forms at the pH of the lacrimal fluid. Unionised species usually penetrate more easily lipophilic barriers of the epithelial layer than ionised ones. Therefore, the degree of drug ionisation (expressed in pKa) in lacrimal fluid can determine how readily the drug passes epithelial corneal barriers. For example, the in vitro corneal permeability of pilocarpine base was 2-3 times greater than that of the ionized form (Francouer et al., 1983) and the miotic response of pilocarpine eye drops was greater at pH 7.0 than at pH 4.5 (Jarvinen et al., 1994). Timolol (weak base, pKa 9.2) concentrations in ocular tissues increase with increasing the pH of the instilled solution. pH values ranging from 6.2 to 7.5 are found to be optimum for ocular bioavailability (Kyyrönen & Urtti, 1990). Also, the charge of the drug molecule may influence its corneal 11 Chapter 1….Literature Review permeability. Cationic drugs are expected to permeate the cornea more easily than anionic drugs where the latter encounters electrostatic repulsion from the negatively charged mucin layer on the eye surface as well as the negatively charged pores present in the corneal epithelium (Jarvinen et al., 1995; Le Bourlais et al., 1998; Loftssona & Jarvinen, 1999). However, the permeability of the positively charged drugs can be hampered by ionic interactions with the negatively charged carboxylic groups of the tight junction proteins (Hornof et al., 2005; Palmgren et al., 2002). Thirdly, drug molecular size can also affect the amount of drug permeated through the tight junctions of the corneal epithelium. Drugs of medium molecular weight such as inulin (mol wt 5500) can cross the cornea solely by paracellular mechanism rather than transcellular and in smaller amounts than uncharged compounds of lower molecular weight (Lee et al., 1986). However, larger molecules such as insulin (mol wt ~6000) can only cross the cornea in the presence of permeation enhancers, such as saponin or Brij 78, which loosen the epithelial tight junctions (Pillion et al., 1991). 1.1.3.2. Formulation factors Some excipients were found to alter drugs’ ocular permeability, and consequently their ocular bioavailability from the used ophthalmic dosage forms. For instance, benzalkonium chloride is a cationic surfactant and is widely used as a preservative in ocular preparations. Benzalkonium chloride and other cationic surfactants have been found to enhance the ocular absorption of a number of drugs varying in molecular size and lipophilicity, including pilocarpine (Mikkelson et al., 1973b), carbachol (Smolen et al., 1973) and prednisolone (Green & Downs, 1974). This might be attributed to the amphiphilic nature of benzalkonium chloride, which imparts drug molecules with some penetration-enhancing properties through the corneal epithelium. Other formulation factors such as drop size, pH and the tonicity of the solution instilled into the surface of the eye must be physiologically compatible to minimise tissue irritation, reflex tearing and blinking, and consequently, this may lead to a substantial loss of the administered dose. As mentioned above, the cul-de-sac of the eye normally holds 7-10 µl, but it can accommodate up to 30 µl without overflowing. However, most commercially available eye droppers deliver a drop of approximately 50 µl. This leads to reflex blinking, drainage to naso-lacrimal duct and 12 Chapter 1….Literature Review spillage of a significant portion of the instilled dose to the cheek until the tear film returns to the normal volume (Chrai et al., 1974; Chrai et al., 1973; Martini et al., 1997). The normal pH of the tear fluid is 7.4. Increased blinking rate and lacrimation are associated with any disturbance of the physiological pH of the tear fluid. Therefore, the pH of the ophthalmic preparations should be formulated in pH 7 to 7.7 to minimise irritation to the eye surface (Ludwig, 2005). In some cases, however, the ideal pH for drug solubility and/or stability is outside this range. Therefore, it is recommended in this case to use the pH which can offer maximum stability, permeability and solubility of the administered drug with minimal buffer capacity, in order to allow the tear fluid to regain its normal pH rapidly (Ali & Lehmussaari, 2006; Van Ooteghem, 1993). Parallel to the ophthalmic formulations’ pH, adjustment of the osmolality of the instilled formulations is an issue. Lacrimal fluid contains various dissolved electrolytes (Na, K, Ca, Cl, and HCO3 ions) constituting about 310-350 mOsm/kg (Ludwig, 2005). Therefore, the instilled formulation must be isotonic with the tear fluid to minimise ocular irritation. The normal isoosmolar range is 270–330 mOsm/kg (Waymouth, 1970). Generally, hypotonic solutions are more tolerable than hypertonic solutions. Osmolality lower than 260 mOsm/kg or higher than 480 mOsm/kg is considered irritating to the eye (Lang et al., 2002; Ludwig, 2005). From the previous discussion, there seems to be a need for the development of an effective ocular delivery system for topically applied drugs, in order to overcome the aforementioned hurdles. The properties of the proposed ocular delivery system should be tailored to fit the major anatomical and physiological barriers of the eye. 13 Chapter 1….Literature Review 1.2. Drug delivery systems providing controlled and continuous ocular delivery Conventional ocular dosage forms such as simple solutions, suspensions and ointments account for approximately 90 % of currently marketed ophthalmic pharmaceuticals (Lang, 1995). This wide-spread use is due mainly to simplicity and patient acceptability. However, the ocular bioavailability of drugs topically applied from eye drops is typically less than 5%. This is due to the very short contact time with the ocular surface, normal tear turnover, conjunctival absorption and rapid drainage through the nasolacrimal duct. This prompts the clinician to recommend a frequent drug instillation at an extremely high concentration. This typical pulse-type dosing might pose safety concerns, by subjecting the ocular tissues to a relatively high local drug concentration and causing several systemic side effects. Therefore, conventional ocular dosage forms no longer constitute optimal therapy (Bochot et al., 1998; Urtti, 2006a). On the other hand, controlled release ocular drug delivery systems can offer many advantages compared with the conventional ones. These include: • Reducing the dose regimen with the goal of improving the patient compliance (Kaur et al., 2004; Kaur & Kanwar, 2002). • Decreasing the effects of washout when instilling multiple drops and hence decreasing systemic side effects (Kaur & Kanwar, 2002). • Minimising the preservative burden by decreasing the number of drops administered per day (Furrer et al., 2002a). Formulation of topical ocular delivery system with consistent bioavailability and minimal adverse effects is still a challenging endeavour facing the pharmaceutical formulator (Kaur et al., 2004; White & Byrne, 2010). The two options currently available to improve ocular drug bioavailability are firstly, prolonging precorneal residence time and minimising precorneal loss and secondly, maximising ocular permeability (du Toit et al., 2011; Kaur et al., 2004). There are many formulation approaches which have been investigated. These formulation approaches are discussed below. 14 Chapter 1….Literature Review 1.2.1. Ophthalmic inserts Ocular inserts are solid or semi-solid sterile preparations to be placed in the cul-de-sac of the eye or on the cornea, in order to prolong residence time (Gurtler & Gurny, 1995; Rathore & Nema, 2009). The size and shape of the inserts are tailored for ophthalmic application. They are composed of a polymeric support with or without drug(s). The drug is incorporated as dispersion or solution in the polymeric support. The inserts can be used for topical or systemic therapy (Gurtler & Gurny, 1995; Rathore & Nema, 2009). Ocular inserts are mostly inserted in the lower fornix and less frequently in the upper fornix or on the cornea. These solid ocular devices require minor surgery by which the polymeric system (into which the drug has been incorporated) is implanted inside the eye (Gurtler & Gurny, 1995). Based on solubility behaviour, ophthalmic inserts can be classified as insoluble, soluble and bioerodible inserts. 1.2.1.1. Insoluble inserts Insoluble inserts can be classified into three groups: diffusional, osmotic systems and contact lenses (Rathore & Nema, 2009). The first two systems include an insoluble membrane surrounding a drug reservoir to control the drug release rate. The reservoir contains a liquid, gel, colloid, semisolid or solid matrix containing a drug homogeneously or heterogeneously dispersed or dissolved therein. Drug-controlling membranes are composed of a wide range of polymers including hydrophobic, hydrophilic, organic, inorganic, naturally occurring or synthetic materials (Gurtler & Gurny, 1995). The third class includes contact lenses which are fabricated using insoluble ophthalmic devices. These classes are likely associated with a high rate of patient incompliance resulting from foreign body sensation due to insolubility of these ocular devices (Rathore & Nema, 2009). 1.2.1.2. Soluble inserts Ocular soluble inserts are completely soluble or biodegradable. This is likely to be an advantage over insoluble ocular inserts, because the former do not need to be removed from their site of application. Ocular soluble inserts can be formulated from natural polymers, synthetic or semisynthetic polymers (Gurtler & Gurny, 1995; Rathore & Nema, 2009). For example, a soluble insert loaded with gentamicin sulfate and dexamethasone phosphate was prepared (Baeyens et al., 1998). The system not only provided a sustained drug release but also showed a concomitant release of the two drugs for the first 10 h, followed by a sustained release of gentamicin over a period of 50 h. However, blurred vision was experienced by patients during treatment. In addition, bioerodible inserts are composed of a drug homogenously dispersed in monolithic 15 Chapter 1….Literature Review matrix devices to control the drug release rate. The main components used for the production of this type of inserts are bioerodible polymers. Bioerodible inserts-based polymers undergo hydrolysis of chemical bonds and hence dissolve over a prolonged period of time, in response to the environment in the eye. Further, a soluble bioadhesive ophthalmic drug insert (BODI) contained 25% w/w gentamicin sulphate was inserted into the lower cul de sac of dogs’ eyes. The clinical efficacy of the treatment of keratoconjunctivitis was compared with that of the aqueous gentamicin eye drops. The total clinical recovery outcomes from the BODI and gentamicin eye drops were obtained after 3 and 7 days respectively. The BODI significantly reduced the treatment period and improved the patient compliance when compared with the aqueous drug solution (Baeyens et al., 2002). Acyclovir water-soluble inserts were fabricated from polyvinyl alcohol and methylcellulose. Both the rate and acyclovir release profile were modified depending on the additives used. The prepared inserts were stored for 6 months at 25oC and the aged inserts demonstrated no change in the drug release profiles indicating good physical stability (El Gamal et al., 2008). 1.2.2. Collagen shields Collagen shields are ocular devices fabricated from porcine or bovine scleral tissues, which bear a collagen composition similar to that of the human cornea and sclera. These ocular devices should be hydrated before insertion to the eye (Lee, 1990; Willoughby et al., 2002). Typically, the drug is loaded into the collagen shield by soaking it in the drug solution for a period of time prior to application. Once in the eye, shields are hydrated by tear fluids, soften and form a clear, pliable, thin film approximately 0.1 mm in thickness with a diameter of 14.5 mm and a base curve of 9 mm that conforms to the corneal surface and is designed to dissolve slowly within 12, 24, or 72 h (Lee, 1990). Collagen shields have been utilised as ocular delivery systems for many pharmacological classes of drugs such as antibiotics, antifungal agents, steroids and immunosuppressive agents (Lee et al., 1992; Schwartz et al., 1990; Unterman et al., 1988; Vasantha et al., 1988; Willoughby et al., 2002). Collagen shields pre-soaked in tobramycin produced significantly higher concentrations of antibiotic in the cornea at one hour than subconjunctival injections of tobramycin (Unterman et al., 1988). The authors claimed that collagen shields containing antibiotics can serve as a vehicle for drug delivery and may prove superior to current methods for preoperative and postoperative antibiotic prophylaxis and the initial treatment of bacterial keratitis. 16 Chapter 1….Literature Review In addition, amphotericin B (an antifungal agent) was formulated in collagen shields. The prepared ocular devices were able to maintain ocular drug levels in the anterior segment of rabbit eyes, as well as frequent-drop delivery with the potential benefit of added convenience and compliance (Schwartz et al., 1990). Further, a combination of gentamicin and methypredinsolone, commonly used drugs in treatment of ocular problems, was formulated in a collagen shield as prolonged drug release ocular devices. However, severe corneal toxicity was reported due to sudden rapid drug release from the device (Lee et al., 1992). In other two reports, collagen shields pre-soaked in aqueous antibiotic solutions were studied for a potential topical ocular treatment for bacterial endophthalmitis (one of serious complications after intraocular surgery) (Hariprasad et al., 2005; Haugen et al., 2008). In one study, both 0.3% w/v gatifoxacin eye drops and collagen shield pre-soaked in 1% w/v gatifloxacin solution achieved significantly lower endophalmitis incidences than balanced salt solution controls in rabbits. However, no statistically significant difference was obtained between the presoaked collagen shields and the eye drops (Haugen et al., 2008). In the other study, collagen shields presoaked in 0.5% aqueous solution of moxifloxacin were studied as a topical ocular treatment option for prevention of bacterial endophthalmitis in 10 patients prior to ocular surgery (Hariprasad et al., 2005). The aqueous humour moxifloxacin levels estimated for collagen shields were lower than those for topical drops. However, the ability to leave the eye batch undisturbed after surgery and avoid patient manipulation of the eye post surgery, so as to instill topical drops could be a potential advantage for this route, especially in the immediate post-operative period (Hariprasad et al., 2005). 1.2.3. Polymeric gels Polymeric gels can be divided into two groups: preformed hydrogels and in situ forming gels (Le Bourlais et al., 1998). 1.2.3.1. Preformed or bio-adhesive hydrogels Bioadhesive hydrogels (hydrocolloids) are hydrophilic polymers which tend to swell, viscolise, gel and adhere upon contact with physiological mucus membranes by virtue of hydrophilic functional groups and structures. Bioadhesive polymers are capable of forming strong noncovalent bonds with the mucin coating mucus membranes and reside in place as long as the 17 Chapter 1….Literature Review mucin is present (Le Bourlais et al., 1998). Over the past three decades bioadhesive hydrogels have been utilised in topical ocular drug delivery to prolong the ocular residence time and minimise the frequency of the instilled drops (Robinson & Mlynek, 1995). For instance, betaxolol, a β-adrenoreceptor antagonist, polyacrylic acid (PAA) gel showed significantly higher bioavailability than a 0.5% w/v solution of betaxolol in rabbits (Weinreb & Jani, 1992). Also, acetazolamide (an antiglaucomic agent) formulated in carboxymethylcellulose hydrogel showed a prolonged effect when compared with the drug solution (Tous & El Nasser, 1992). Moreover, polycarbophil gel enhanced the ocular delivery of topically applied gentamicin in rabbits. The polymeric formulation increased the uptake of gentamicin by the bulbar conjunctiva twice, compared with an aqueous control formulation (Lehr et al., 1994). A three-fold increase of the precorneal residence time of tobramycin was achieved when formulated in a chitosan gel compared with the commercial drug solution of the drug (Ludwig, 2005). Xyloglucan [tamarind seed polysaccharide (TSP)] was tested as a novel vehicle for ophthalmic delivery of timolol (Burgalassi et al., 2000). The polymer used, in spite of a comparatively low viscosity, showed high timolol concentrations in the ocular tissues and a low systemic absorption compared with the commercially available product based on in situ gelling (Timoptic XE®). The results presented TSP as a potentially useful carrier for ophthalmic delivery systems. One of the most successful applications of hydrogels in ophthalmology is their use as tear substitutes due to their commercial availability. For example, celluloses (Lacril®), polyvinyl alcohol (Liquifilm®), polyacrylic acid (Lacrigel®, Lubrithal®, Gel-Larm®), and hyaluronic acid (Hy-Drop®) are frequently used as tear substitutes for the treatment of dry eye disease (Tsubota & Dogru, 2006; Zignani et al., 1995). Bioadhesive polymers, however, are associated with some pitfalls as an ocular drug delivery system. The drop size or volume of commercial ocular medication formulated using bioadhesive polymers is not as uniform as conventional ocular solutions. The amount delivered to the eye, consequently, is generally incorrect. Moreover, the presence of a viscous vehicle can cause blurred vision and formation of a veil in the corneal area leading to loss of eyesight (Winfield et al., 1990). 1.2.3.2. In situ activated gel forming systems The abovementioned problems related to the accuracy of the instilled dose and blurred vision from the preformed hydrogels could be overcome by in situ activated gel delivery systems. In situ activated gel-forming systems can be described as viscous liquids that upon exposure to 18 Chapter 1….Literature Review physiological conditions will shift to a gel phase (Le Bourlais et al., 1998). Three methods causing phase transition on the eye surface have been reported in the literature: a change in temperature (Miller & Donovan, 1982), pH (Gurny, 1981; Gurny et al., 1987) or electrolyte composition (Rozier et al., 1989). Other methods can induce in situ gel activation have been recently reported in the literature such as UV/visible light-irradiation (photopolymerisable hydrogels) and enzyme mediated gelation, however, they have not been investigated for ocular drug delivery (Van Tomme et al., 2008). Here are some applications of various methods used to induce in situ sol/gel transition on the surface of the eye. Poloxamers or pluronics are block co-polymers consisting of poly (oxyethylene) and poly (oxypropylene) units. They rapidly undergo thermal gelation when the temperature is raised to that of the ocular surface, while they remain liquid at refrigerator temperature. A mixture of pluronic F127 and pluronic F68 was prepared as an in situ gelling system for ocular delivery of sparfloxacin (antibacterial agent) for treatment of bacterial keratitis. The developed formulations provided sustained release of the drug over a 24-h period and better improvement in artificially induced bacterial conjunctivitis in rats' corneas (Nesseem, 2011). However, Poloxamers exhibit thermal gelation at high concentrations (usually between 20 and 30% w/v); Pluronic F127 has been found to be more damaging to the cornea than a physiological saline solution (Furrer et al., 2000). The relatively high polymer concentration (20-30% w/v) required for in situ thermal gelation can pose toxicological concerns (Miyazaki et al., 2001). Cellulose acetophthalate (CAP) latex is defined as a highly unstable system which coagulates when its native pH of 4.5 is raised by the tear fluid to pH 7.4(Gurny, 1981). This pH-induced coagulation of CAP occurs at a high polymer concentration (30% w/v) (Gurny, 1981; Kumar et al., 1994). Also, carbopol is a polyacrylic acid (PAA) polymer, which shows sol-gel transition in an aqueous solution as the pH is raised above its pKa of about 5.5. As the concentration of carbopol increases in the vehicle in order to improve its rheological properties, the acidity of the vehicle increases. Hence, the inherent acidity of the vehicle could induce ocular tissue irritation and induced lacrimation (Nanjawade et al., 2007). An example of the ion-activated gelation is gellan gum (an anionic polysaccharide) which exhibits phase transition with increased ionic strength. Also, it has been found that the extent of 19 Chapter 1….Literature Review gel-formation of gellan gum increases proportionally with the amount of mono- or divalent cations present in the tear fluid. Hence, the main triggering effect inducing phase transition is due to the generous availability of mono- and divalent cations associated with reflex tearing (Greaves et al., 1990). The dependency of the gel transition of some in situ gelling systems on the uptake of cations present in the tear film may be disrupted in the presence of dry eye syndromes (especially with diabetic patients). Moreover, the dependency on electrolytes for gel transition is likely to hinder the tonicity adjustment and consequently, the final product is not iso-osmotic with the tear fluid. 1.2.4. Colloidal delivery systems 1.2.4.1. Microemulsions Microemulsions (MEs) have been utilised as topical ocular drug delivery systems because of their inherent properties and structures (Vandamme, 2002). The use of MEs for ocular drug delivery is advantageous because the presence of surfactant and co-surfactant increases membrane permeability to the applied drug (Lawrence & Rees, 2000). Moreover, these systems offer the additional advantage of having low viscosity and the convenience of eye drops. Pilocarpine-based MEs have shown more sustained drug activity than the drug solution where twice daily instillations of these systems were equivalent to four instillations of conventional eye drops (Naveh et al., 1994). Recently, ME-based phase transition systems have been prepared and evaluated as ocular delivery systems using pilocarpine hydrochloride as a model water-soluble drug (Chan et al., 2007). These systems underwent phase transition from ME to liquid crystalline (LC) and to coarse emulsion (EM) with a change in viscosity depending on water content. ME and LC showed higher drug release retarding efficiency than EM. However, toxicity and tissue irritancy have been observed with microemulsion formulations. For example, ocular application of ME and LC disrupted the precorneal tear film and the tear evaporation rate increased compared with the aqueous solution. This is due to the use of relatively high concentrations of surfactants and co-surfactants within the formulations (Alany et al., 2006; Chan et al., 2008). 1.2.4.2. Nanosuspensions The ocular formulation of poorly water-soluble drugs is not an easy task. These drugs suffer from erratic ocular absorption and reduced availability (Kayser et al., 2005). This class of drugs can be formulated as a nanosuspension in an appropriate dispersing vehicle, in an attempt to reduce the 20 Chapter 1….Literature Review aforementioned problems. Drug compounds that form crystals with high energy content, rendering them insoluble in either organic (lipophilic) or hydrophilic media, are good candidates for formulation as a nanosuspension (Kayser et al., 2005). Thus, the use of nanosuspensions in ophthalmic pharmaceutical formulations is likely to be an attractive area offering a great possibility to overcome the inherent difficulties associated with the ocular delivery of poorly water-soluble drugs. For instance, methyl prednisolone was prepared in a nanosuspension. The prepared formulations showed localised and controlled ocular anti-inflammatory activity in rabbits (Adibkia et al., 2007). 1.2.4.3. Polymeric microspheres and nanoparticles Polymeric microspheres and nanoparticles are polymer-based particulate drug delivery systems. The difference between these systems is based on their size. Nanoparticles can be defined as particles with a diameter of less than 1 µm, whereas particles in the micrometre size (> 1 µm) are called microspheres (Ali & Lehmussaari, 2006). They can be fabricated from various biodegradable polymers such as natural and synthetic polymers. Drugs can either be dispersed in the polymeric matrix or physically bound to the surface. Polylactides (PLAs), polycyanoacrylate, chitosan, gelatin, sodium alginate and albumin were studied for efficient drug delivery to the ocular tissues (Sahoo et al., 2008). For instance, Polylactide-based nanoparticles (NPs) encapsulating two flourochromes (Rhodamine-6G and Nile red) were tested as drug delivery systems targeting retina and retinal pigment epithelium (RPE) and applied intravitreally (Bourges et al., 2003). NPs were able to give a high local concentration of the tested compounds at the RPE suggesting the feasibility of targeting the posterior segment of the eye by NPs. Moreover, chitosan is a cationic polysaccharide and is obtained by deacetylation of chitin. Chitosan has been shown to promote intraocular drug penetration with good ocular tolerability (Zambito & D i Colo, 2010). Chitosanbased nanoparticles have been prepared to enhance the ocular bioavailability of indomethacin, an anti-inflammatory drug used to minimise ocular inflammation after cataract surgeries. The prepared ocular carriers showed controlled and higher bioavailability than that of the drug solution (Badawi et al., 2008). 1.2.4.4. Solid lipid nanopraticles Solid lipid nanoparticles (SLNs) are used as a colloidal carrier system for controlled drug delivery. SLNs represent an alternative carrier system to emulsions and polymeric nanoparticles. 21 Chapter 1….Literature Review The main reasons for their development is more stable alternatives to liposomes and the combination of advantages from different carriers systems like liposomes and polmeric nanoparticles (Muller et al., 2000). SLNs have been evaluated for ocular controlled drug delivery (Cavalli et al., 2002; Cavalli et al., 1995; Gonzalez-Mira et al., 2011). For instance, tobramcin-loaded SLNs produced significantly higher bioavailability in aqueous humour when compared with an equal dose of tobramycin administered using standard commercial eye drops (Cavalli et al., 2002). Physical instability and encapsulating water-soluble drugs into the lipidic core are still unresolved issues. It has been found that SLN formulations gelled after a relatively short period of storage. Although this can be prevented by altering the composition and concentration of stabilising surfactants, these stabilising additives are hampered by toxicological considerations and incompatibilities with the incorporated drugs due to chemical degradation (Freitas & Muller, 1999). Recently, flurbiprofenloaded SLNs (FB-SLNs) have been prepared from stearic acid, miglyol 812 and castor oil. The prepared FB-SLNs were evaluated for ocular toxicity using the ocular Draize test. The tested formulations were found to be minimally irritant (Gonzalez-Mira et al., 2011). 1.2.4.5. Dendrimers Dendrimers are artificial nano-constructs and as the name suggests, have a tree-like structure. Their basic units are active chemical moieties, which construct around small molecules or the core using connectors and branching units. Poly(amidoamine) (PAMAM), poly-lysine, or poly(oropylenimine)-based dendrimers are amongst the preferred classes of these (Jain & Asthana, 2007; Tomalia et al., 1985). The final shape of the dendrimers is globular, with hollow internal cavities and a number of terminal groups. By virtue of the highly branched terminals, a number of drugs can be encapsulated or attached to the surface or peripheral group and hence, dendrimers can be utilised as a potential drug delivery system. For instance, many poorly soluble drugs can be properly encapsulated in the internal cavities of dendritic constructs (Cheng et al., 2008; Jain & Asthana, 2007; Quintana et al., 2002; Tomalia et al., 1985). PAMAM dendrimers were prepared for controlled ocular delivery of pilocarpine nitrate and tropicamide (Vandamme & Brobeck, 2005). The macromolecular carriers showed a prolonged ocular residence time compared with 0.2% w/v carbopol 980 NF solution. The residence time of the dendrimers was markedly dependent on the size and molecular weight of the prepared 22 Chapter 1….Literature Review carriers. The main drawback which hampers the wide use of dendrimers as a drug delivery system is the lack of sufficient evidence of their safety and toxicity (Jain & Asthana, 2007). Recent evidence has shown that fatal complication such as disseminated intravascular coagulopathy was reported after iv injection of small doses (> 10 mg/kg) of positively charged PAMAM dendrimers generations 3.5 to 7 in sizes of 50 nm in mice. Also, hemobilia (haemorrhage of the biliary tract) and a preferential toxicity on the intestinal wall in terms of haemorrhage and intraluminal bleeding were recorded with oral administration of the PAMAM dendrimers (Greish et al., 2010). 1.2.4.6. Cubosomes Cubosomes are submicron lipid particles of bicontinous cubic liquid crystalline phase. They have been proposed as a colloidal delivery system for lipophilic drugs for better solubilisation and a sustained drug release effect (Boyd, 2003; Spicer, 2005). Research on cubosomes, as drug carriers for ocular delivery, has been scarce (du Toit et al., 2011). Cubosomes were prepared from a mixture of monoolein (glycerol mono-oleate) and water by high pressure emulsification using poloxamer 407 as a dispersing agent (Gan et al., 2010). These cubosomes were formulated as an ocular delivery system for dexamethasone (DEX). Ex vivo permeation studies, using excised rabbit corneas showed that the apparent permeability coefficient calculated for the prepared DEX cubosomes was 3.5-4.5 times greater than the DEX aqueous solution. Also, the ocular bioavailability for DEX cubosomes was 1.8 times greater than that for the DEX aqueous solution. Histological examination of the excised rabbit corneas incubated in DEX cubosomes for 2 h did not show any harmful signs. These findings might present cubosomes as a new ocular delivery system for water-insoluble drugs (Gan et al., 2010). 1.2.4.7. Vesicular delivery systems The abovementioned ocular delivery devices can achieve more controlled and effective therapeutic actions with minimal systemic and ocular side effects associated with pulse-type conventional dosage forms. However, these systems are still not devoid of pitfalls, including: • Poor patient compliance during insertion, as in ocular inserts. • Tissue irritation and damage induced by penetration enhancers and co-surfactants commonly used to stabilise microemulsion and collagen shields (Alany et al., 2006; Kaur et al., 2004). 23 Chapter 1….Literature Review • Ocular toxicity of some polymers used to fabricate microspheres and nanoparticles such as albumin and polybutylcyanoacrylate (Zimmer & Kreuter, 1991). • Toxicological concerns due to relatively high polymer concentrations required for some in situ thermal gelation (Miyazaki et al., 2001). • Failure of drug delivery devices. For example, some in situ gelling systems rely on a generous supply of cations from the resident tears for sol-to-gel phase transition. Alteration of the composition of the tear fluid and/or the presence of dry eye symptoms (common with diabetic patients) (Akinci et al., 2007; Dogru et al., 2001; Kaiserman et al., 2005) might compromise the sol-to-gel phase transition of these systems. Hence, these systems might not be optimal for drug delivery in diabetic patients. Surfactant vesicles (liposomes and niosomes) could overcome the aforementioned hurdles (du Toit et al., 2011; El-Gazayerly & Hikal, 1997; Gregoriadis & Florence, 1993; Le Bourlias et al., 1995; Uchegbu & Vyas, 1998). They can offer a prolonged and controlled action at the corneal surface and a substantial increase of ocular bioavailability, by providing an intimate contact of the drug with the lipidic corneal barrier and preventing the metabolism of the drug by the enzyme present at the precorneal tear film (Azeem et al., 2009; Kaur et al., 2004). Moreover, lipid vesicles offer a promising approach to meet the need for an ophthalmic drug delivery system that has the convenience of being in the form of eye drops, but will localise and maintain drug activity at its site of action (Azeem et al., 2009; Kaur et al., 2004). Furthermore, plain liposomes have been used in a form of spray for topical ocular administration for treatment of dry eye symptoms. Two commercially available liposome products, clinically indicated for the management of dry eye, are ClaryMist® manufactured by Savant, UK and Ocusoft® manufactured by Ocusoft, USA. They are sprayed on the eyelids to improve the stability of the lipid layer of the tear film and treatment of dry eye symptoms (Craig et al., 2010). Liposomes Liposomes are closed bilayer lipid vesicles that can be produced basically from natural non-toxic phospholipids and a membrane stabilising lipid, most notably cholesterol. Liposomes offer many advantages as a unique drug delivery system. They are biodegradable and biocompatible; they can encapsulate both hydrophilic drug moieties in their aqueous milieu and lipophilic moieties between hydrocarbon chain bilayer membranes; and their properties can be tuned with lipid 24 Chapter 1….Literature Review composition, size, surface charge and the method of preparation (Kaur et al., 2004; Schaeffer & Krohn, 1982). Liposomes encapsulating gancyclovir and iododeoxyuridine (antiviral drugs) were prepared for more efficient treatment of herpes simplex virus eye infections (Norley et al., 1986). It was found that these delivery systems could act as improved vehicles for drug delivery in treatment of ocular herpes simplex virus infection. Also, antisense oligonucleotides which can be used efficiently to treat ocular diseases like cytomegalovirus retinitis can be encapsulated in the liposomes and efficiently targeted at the retina (Bochot et al., 2002). Moreover, liposomes encapsulating mitomycin C (MMC) were prepared in an attempt to minimise the toxicity of mitomycin on ocular tissues. Liposomes containing 0.2 mg/ml of MMC were tested against aqueous solutions and tamarind seed polysaccharide (TSP)-based viscous solutions. The results showed reduced cytotoxicity of MMC in TSP viscous vehicles and liposomes compared with the drug solution. Only liposomes encapsulating MMC vehicles produced both lower toxicity for a rabbit corneal epithelial cell line culture and a marked reduction of the corneal healing rate in vivo (Chetoni et al., 2007). Recently, fluconazole liposomes have been found to be more effective than the fluconazole aqueous solution in treating Candida keratitis (corneal inflammation due to fungal infection) in rabbits. The fluconazole liposome-treated group showed complete corneal healing for 86.4% of the infected rabbits compared with 50% with the fluconazole solution treated group. Liposomes offer more intimate contact with corneal barriers and can enhance corneal drug absorption (Habib et al., 2010). Niosomes Over the last three decades, liposomes of different compositions used for topical ophthalmic drug delivery have increased/prolonged the therapeutic effect while minimising toxic effects (Chetoni et al., 2007; du Toit et al., 2011; Gregoriadis & Florence, 1993; Kaur et al., 2004; Treblay et al., 1993). Vesicles consisting of one or more surfactant bilayers enclosing aqueous spaces are called nonionic surfactant vesicles or simply niosomes. Niosomes are considered of particular interest as they offer several advantages over liposomes: • Niosomes are more stable chemically (Kaur et al., 2004; Uchegbu & Florence, 1995). • Niosomes incur lower production cost due to the availability of starting materials (Sahin, 2006). 25 Chapter 1….Literature Review • Niosomes are biodegradable and non-immunogenic (Kaur et al., 2004). • Niosomes do not require expensive handling (storing in a freezer and preparation under nitrogen gas). Nevertheless, the formulation of niosomes can pose challenges due to possible physical instability and poor ability to encapsulate a considerable amount of water-soluble drug molecules (Essa, 2010). Niosomes have been evaluated as ocular vehicles for a wide range of therapeutic classes such as anticholinergic (cyclopentolate), anti glaucomic (acetazolamide and timolol maleate) and antibiotics (gentamicin) with minimal signs of ocular irritation as well (Abdelbary & El-gendy, 2008; Aggarwal et al., 2004; Saettone et al., 1996b; Vyas et al., 1998). Tween 20based niosomes significantly improved the ocular bioavailability of cyclopentolate, with respect to the reference buffer drug solution and micellar solution, suggesting that it can be used as an efficient vehicle for ocular drug delivery (Saettone et al., 1996b). Also, niosomes encapsulating timolol maleate (anti-gluacomic agent) were prepared using the reverse phase evaporation method and coated with 0.5% w/v chitosan solution. The timolol niosomes achieved a significantly (1.7 times) higher timolol concentration in the aqueous humour than that of the drug solution. This was attributed to better improvement of corneal penetration and prolonged precorneal residence time compared with the timolol aqueous solution (Kaur et al., 2010). Niosomes should have certain attributes in relation to their potential use as vehicles for ocular drug delivery. These include: • The size of the vesicle should be large enough to resist drainage by reflex tearing and eye blinking. For example, large multilamellar vesicles (MLV) were found to reside for a longer period on the ocular surface than small unilamellar liposomes. This was attributed to the ability of MLV to entrap a relatively higher amount of the drug and resist nasolacrimal drainage (Fitzgerald et al., 1987a; Hathout et al., 2007). The same argument applies to niosomes. Ideally, it has been reported that niosome sizes > 10 µm are suitable for ocular delivery (Sahin, 2006; Uchegbu & Vyas, 1998). • The shape of the prepared niosomes should show some irregularities to fit properly into the cul-de-sac of the eye and lodge on the eye surface (Uchegbu et al., 1992; Uchegbu et al., 1997; Uchegbu & Vyas, 1998). • Ocular niosomes should ideally be thermo-responsive in order to release drug content in a controlled, yet timely manner before being flushed by blinking and nasolacrimal drainage 26 Chapter 1….Literature Review (Uchegbu et al., 1997; Uchegbu & Vyas, 1998). Complete abolishment of gel/liquid transition of surfactant forming niosomes could produce niosomes with an extremely low release rate to water soluble drugs, which does not suit the short residence time of ocularly administered ophthalmic products. The abovementioned criteria have typically been attributed to discomes and polyhedral niosomes. These are modified vesicular structures and are different from conventional spherical niosomes (Azeem et al., 2009; Uchegbu & Florence, 1995; Uchegbu & Vyas, 1998). Discomes containing a low cholesterol concentration less than 30% mol/mol were prepared by incubating the preformed spherical niosomes in cholesteryl poly-24-oxyethylene ether (Solulan C24) at 75oC for 1 h. This resulted in the formation of large (11-60 µm) and multifaceted vesicular systems (Uchegbu et al., 1992). Since then, there has been only one report on discomes as an ocular delivery system for timolol maleate (Vyas et al., 1998). The prepared discomes have been found to entrap a relatively large quantity of timolol and improve the ocular bioavailability compared with a timolol maleate solution (Vyas et al., 1998). One possible reason for such scarcity is the need for a relatively high temperature during discome preparation which might affect the chemical stability of some thermo-labile therapeutic agents. On the other hand, spherical niosomes have been shown to possess more stable membranes than polyhedral ones due to the presence of cholesterol (Arunothayanun et al., 1999). In conclusion, ophthalmic niosomes could offer prolonged drug release and enhance drug permeation. Thus, the potential of niosomes as a vehicle for ocular delivery of naltrexone hydrochloride are studied in this thesis. Naltrexone hydrochloride has been found to be a promising treatment for diabetic keratopathy. Causes, prevalence and other treatment options of diabetic keratopathy are discussed in the following sections. 27 Chapter 1….Literature Review 1.3. Diabetic keratopathy and the available treatment options 1.3.1. Physiology of the cornea The physiological functions of the cornea are (Klyce & Beuerman, 1998; Nishida, 2005): • Refractive: it transmits and focuses light onto the retina. • Protective: it forms a mechanical and chemical barrier to the posterior segment of the eye. • Regulatory: it contains the intraocular pressure. 1.3.2. Diabetes mellitus and the cornea Diabetes mellitus has increasingly attracted the attention of scientists, especially since the WHO’s announcement that “the worldwide burden of diabetes in adults was around 173 million in the year 2002”. The incidence of diabetes has risen dramatically in recent decades and a twofold, or greater, increase is expected to occur in the coming decades (Wild et al., 2004). Diabetes mellitus has a serious impact on the tissues of the entire eye. For instance, diabetic retinopathy (damage of the retina from diabetes) and cataracts (opacity of the eye lens) are major ocular complications arising from diabetes. Diabetic retinopathy and cataracts are the leading causes of vision loss in the industrial world. These disorders account for approximately 6% of the US population and 17% of Americans ≥ 40 years, respectively (Frank, 2004; Kenny et al., 1995). However, in the last four decades, diabetic retinopathy has been investigated fully. Advances in treatment over the past 40 years have greatly reduced the risk of blindness from this disease (Frank, 2004; McLaughlin et al., 2010). On the other hand, diabetic effects on the anterior segment of the eye have been reported for many decades (Herse, 1988; Hyndiuk et al., 1977; McLaughlin et al., 2010; Schwartz, 1974). Observed diabetic abnormalities in the anterior segment of the eye include: • Decreased tear production (Herse, 1988). • Presence of glucose in tears (Gasset et al., 1968). • Increased risk of infection (Herse, 1988; Schultz et al., 1981 ). • Corneal epithelial fragility (Schultz et al., 1981 ). • Wrinkling of Descemet’s membrane (Henkind & Wise, 1961). 28 Chapter 1….Literature Review • Reduced corneal sensation (Ishida et al., 1984; Schwartz, 1974). • Delayed wound repair (Lambiase et al., 2000; Lambiase et al., 1998b). • Neurotrophic corneal ulcers (Hyndiuk et al., 1977). Corneal disorders secondary to diabetes (diabetic keratopathy) are increasingly recognized as a cause of the morbidity associated with diabetes (Cisarik-Fredenburg, 2001; Kaji, 2005). Indeed, diabetic keratopathy has been estimated to occur in 47-64% of diabetic patients during the course of their disease (Schultz et al., 1981 ) and diabetics have an increased risk of developing corneal epithelial fragility, epithelial defects, recurrent epithelial erosions, decreased sensitivity, abnormal wound healing, increased susceptibility to injury and non-healing or infected corneal ulceration. Ultimately such corneal complications may lead to visual impairment. Diabetic keratopathy is mainly attributed to corneal nerve tortuosity and degeneration as part of general neuropathy of diabetes mellitus (Rosenberg et al., 2000; Ruben, 1994; Schwartz, 1974). Nevertheless, the available treatment options for diabetic keratopathy are tear substitutes and antibiotics. Such treatments are symptomatic and conservative to keep the cornea hydrated with minimal microbial burden. There has not yet been a causative treatment developed for this serious diabetic condition. However, damage of the corneal nerves due to diabetes has been thought to be the main culprit (Feman, 2000; McLaughlin et al., 2010; Schultz et al., 1981 ). The following section sheds light on corneal nerve role and structure. 1.3.3. Corneal innervations Innervations of the cornea are essential for pain sensation and for tissue repair (Nishida, 2005). The cornea is the most densely innervated mammalian tissue (Figure 1.5). Human cornea has a nerve density of 300-600 times greater than that of the skin and 20-40 times greater than that of the tooth pulp (Muller et al., 1997). Most of the sensory nerves in the cornea are derived from the ciliary nerves of the ophthalmic branch of the trigeminal nerve (Nishida, 2005). Nerve bundles enter the corneal mid stroma at the periphery, each containing a few myelinated fibres and a number of finer unmyelinated fibres. The unmyelinated nerve endings are sensitive to touch, temperature and chemical stimulation (Muller et al., 1997). Because transparency is of prime importance, the cornea is avascular and receives necessary nutrients via diffusion from the tear film, the aqueous humour and from neurotrophins supplied by nerve fibres (Gobbels et al., 1989). Axons from the trigeminal ganglion and sympathetic ganglion terminate in delicate endings among the epithelial cells of the cornea. 29 Chapter 1….Literature Review Figure 1.5 Human corneal nerves penetration and distribution, modified from (Muller et al., 1997) and reproduced with permission of the copyright owner. Many ocular, neurological and systemic diseases can alter the typical corneal nerve architecture and impair corneal sensitivity (Table 1.1) (Lambiase et al., 1999). A major complication of that is the neurotrophic corneal ulcer. It is a degenerative corneal disease characterised by the impairment of corneal sensitivity associated with epithelial breakdown, a deficiency in the healing process leading to corneal ulceration, and subsequent vision loss. Neurotrophic corneal ulcers are classified into three stages (Mackie, 1995): Stage 1 is characterised by rose bengal staining of the inferior palpebral conjunctival surface, a decrease in tear breakup time, and punctate keratopathy. If these changes become chronic, superficial vascularisation, stromal scarring, epithelial hyperplasia and irregularity may develop. Stage 2 is characterised by epithelial breakdown where cells assume an oval or circular shape that is often localised in the superior half of the cornea. The area of epithelial deficit is usually 30 Chapter 1….Literature Review surrounded by loose epithelium that, in the absence of a healing process, becomes hazy, oedematous, and poorly adherent to Bowman’s layer. The edges of the defect become smooth and rolled as the defect ages without appreciable epithelial growth (Cavanagh et al., 1976). Stage 3 is characterised by stromal involvement with a corneal ulcer that may progress to melting and perforation. Secondary infection or topical treatment with corticosteroids increases the risk of perforation (Groos, 1997). 31 Chapter 1….Literature Review Table 1.1 Causes of decreased corneal sensitivity Category Disease Genetic Riley-Day syndrome (familial dysautonomia) Goldenhar-Gorlin syndrome Mobius syndrome Familial corneal hypoesthesia Systemic Diabetes Leprosy Vitamin A deficiency Central nervous system Neoplasm Aneurysm Ischaemia Neurosurgical procedures for acoustic neuroma or trigeminal neuralgia Other surgical injury to the trigeminal nerve Ocular Herpes infection (herpes simplex or herpes zoster) Chemical and physical burns Abuse of topical anaesthetics Topical toxicity (timolol, betoxolol, diclofenac sodium, sulfacetamide) Surgical or laser treatment to cornea Corneal dystrophies (lattice, granular) Chronic ocular surface injury or inflammation Contact lens wear 32 Chapter 1….Literature Review Over the past few years, there has been growing interest in exploring the mechanisms by which various growth factors and therapeutic agents promote healing of corneal epithelial cells in order to establish any significant correlations between topical administration and accelerated corneal healing (Kaji, 2005; Micera et al., 2004; Schultz et al., 1981 ; Woo et al., 2005). Before discussing the treatment options, different factors ameliorating altered corneal healing will be mentioned. 1.3.4. Mechanisms of corneal maintenance and healing The corneal epithelium is composed of a squamous epithelium of 5-7 layers of stratified cells attached to a basement membrane and is separated from the stroma by Bowman’s layer. Trauma, disease and corneal surgery, including the increasingly popular refractive laser surgery, all cause the cornea to enter a phase of wound healing. Upon wounding, the cornea, which has less immune protection than other tissues due to the lack of blood vessels, needs to re-epithelialise urgently in order to restore its barrier function. The healing of epithelial wounds can be divided into several distinct but continuous phases: sliding of superficial cells to cover the denuded surface, cell proliferation, and stratification for re-establishment of multicellular layers (Lu et al., 2001). Although several layers participate in cell migration, the majority of the defect is initially covered by the sheet-like movement of a monolayer of epithelial cells, followed by a landslide-like mass movement of the epithelium (Suzuki et al., 2003). Based on the XYZ hypothesis for corneal epithelial maintenance put forward by (Thoft & Friend, 1983), researchers have assumed that limbal stem cells and transient amplifying cells are induced after acute wounding and that these cells always migrate centripetally to repair the wound. However, it has been recently reported that after corneal wounding, the capacity for epithelial cell proliferation and migration appears to be as active in the central cornea as in the peripheral/limbal areas (Chang et al., 2008). Indeed, central and peripheral epithelial recovery remains equal even after ablation of the limbus. Thus corneal epithelial recovery may be independent of limbal stem cells (the mother cells of the Y component in the previous hypothesis) in a human corneal organotypic model. Therefore, central human corneal epithelial cells are fully capable of corneal epithelial regeneration, at least in the first 12 h after wounding. This observation may lead to new insights into the pathogenesis and therapy of corneal epithelial diseases. 33 Chapter 1….Literature Review Corneal epithelial cell migration is modulated by various humoral and extracellular matrix (ECM) proteins (Lu et al., 2001). Growth factors, such as insulin-like growth factor (IGF)-1 (Nagano et al., 2003), epidermal growth factor (EGF)(Maldonado & Furcht, 1995a), interleukin 6 (Nishida et al., 1992), fibroblast growth factor-2 (FGF-2) (David et al., 1995), transforming growth factor-β (TGF-β) (Saika et al., 2004), keratinocyte growth factor (KGF), and hepatocyte growth factor (HGF) (Sharma et al., 2003; Wilson et al., 1993) have been shown to stimulate corneal epithelial cell migration both in vivo and in vitro. Glycoproteins of the ECM, such as fibronectin (Fn), laminin (Ln), and collagen IV, also facilitate cell migration (Maldonado & Furcht, 1995b; Ohji et al., 1993). However, the precise relationship between these cytokines and ECM protein production in the corneal epithelium is not well understood. Once the wound area is covered the cells are further induced to proliferate to restore the stratified nature of the epithelium. Corneal wounding induces an increased release of several growth modulating cytokines and many of these cytokines have been linked to the regulation of epithelial cell proliferation (Lu et al., 2001). The presence of TGF-β in the later stages prevents over proliferation of the epithelial cells (Joyce & Zieske, 1997). Whilst the above data seems to indicate specific roles for specific cytokines in corneal wound healing, the interaction between growth factors and cells, and growth factor to growth factor, inevitably means that the right combination of factors needs to be expressed with spatial and temporal coordination to ensure correct wound healing. 1.3.5. Treatment of diabetic keratopathy 1.3.5.1. Standard treatment The standard treatment of neurotrophic corneal ulcers consists of a number of components including: maximising preservative free topical lubricants, minimising evaporative tear loss, use of topical antibiotics, protecting the corneal surface with a bandage contact lens, covering the ocular surface via lid closure by patch, tarsorrhaphy or induced ptosis, and more permanent solutions such as construction of a conjunctival flap. However, even in combination these measures may be ineffective, and the outcome is often severe impairment or loss of vision (Lambiase et al., 1998b). 34 Chapter 1….Literature Review 1.3.5.2. Growth factors A number of growth factors has been identified in the corneal epithelium and their gene expression unravelled (Hongo et al., 1992; Kiritoshi et al., 1991; Schultz et al., 1990). The role of growth factors in maintaining the normal structure and function of the cornea, and in corneal epithelial healing has emerged over the past three decades (Schultz et al., 1990; Tripathi et al., 1990). Autologous serum has been shown to be beneficial in epithelial wound healing both in vitro and in clinical studies (Schulze et al., 2006). Although the exact mechanism for these effects is not entirely clear, growth factors such as EGF and TGF may play a role. A brief overview of some of the key growth factors involved in corneal epithelial healing follows. Insulin-like growth factor-1 (IGF-1) IGF-1 is a multifunctional regulatory peptide that shares structural homology with proinsulin. It has been shown to mediate proliferation, differentiation, and survival effects, depending on the target cell and the presence of other hormones and growth factors (Gockerman et al., 1995). IGF1 is secreted by the liver. Like other growth factors, it is present in high concentrations in serum, produced locally, and may act in an autocrine or paracrine manner (McAvoy & Chamberlain, 1990). Accumulated evidence indicates that IGF-1 promotes cell motility in a variety of normal and malignant cell types. However, the effect of IGF-1 on corneal epithelial cell motility and migration is controversial (Nishida et al., 1996). It has been reported that substance P (SP) and IGF-1 synergistically stimulated corneal epithelial migration in an organ culture of the cornea. Accordingly, it was found that addition of SP or IGF-1 separately did not affect epithelial cell migration, yet co-treatment with SP and IGF-1 significantly stimulated epithelial cell migration (Nakamura et al., 1997). Also, it was noted that IGF-1 significantly increased migration and expression of laminin-5 in cultured human corneal epithelial cells (Lee et al., 2006). In the clinical setting, the administration of eye drops containing both SP and IGF-1 has been shown to be an effective treatment in the case of a child with neurotrophic and anhidrotic keratopathy (Brown et al., 1997). Moreover, eye drops containing peptides based on SP and IGF1 have been reported to be effective in the prevention of superficial punctate keratopathy in diabetic patients after cataract surgery and have also successfully induced rapid epithelial resurfacing in patients with persistent corneal epithelial defects (Yamada et al., 2008). 35 Chapter 1….Literature Review Insulin Insulin, an anabolic peptide hormone closely related to IGF, is another growth factor implicated in wound repair, but its role is not well documented (McAvoy & Chamberlain, 1990). Insulin stimulates hepatotactic migration of human epidermal keratinocytes (Benoliel et al., 1987) and topical insulin therapy has been shown to aid the healing of ulcerations (Rosenthal, 1968; Van Ort & Gerber, 1976) and burns (Pierre et al., 1998). More interestingly, it has been shown that insulin is present in human tear fluid, and receptors to insulin have been detected on the human ocular surface (Rocha et al., 2002), cornea (Naeser, 1997) and neuronal and vascular tissues of the retina (Im et al., 1986; Thomopoulos & Pessac, 1979). The functions of insulin receptors within structures of the eye are not yet known, but diabetes is the major cause of blindness in people of working age and is often associated with disorders of the corneal epithelium (Rosenberg et al., 2000; Zagon et al., 2002a). Recently, intensive systemic therapy with insulin which establishes normoglycemia in rats with diabetes was found to prevent the delay in wound healing of the ocular surface epithelium in poorly controlled diabetic animals (Zagon et al., 2006c). Normoglycemia in diabetic animals, induced by systemic treatment with insulin, restores decreased levels of DNA synthesis in the ocular surface epithelium to normal values when examined three weeks after wounding (Zagon et al., 2006c). Also, it has been reported that treatment with topical insulin significantly accelerates wound healing in diabetic rats compared with the untreated diabetic group. However, topical insulin had no effect on corneal re-epithelialisation of corneal wounds in healthy rats (nondiabetic) (Zagon et al., 2007a). These results were attributed to the ability of topical insulin to restore the decreased levels of DNA synthesis in basal epithelial cells to normal values, seen 48 h after wounding. These observations present a new and promising therapeutic indication for insulin in the treatment of corneal ulcers in diabetics. Topical application of insulin to the surface of the eye will inevitably result in some systemic absorption. It has been shown previously monitoring serum glucose levels up to 14 h of topical exposures to 1 U of insulin had no effect on plasma glucose values in diabetic or healthy rats (Zagon et al., 2007a). However, the insulin used in this study was compounded as a simple solution. This is different from the commercially available insulin dosage forms. Commercial insulin contains pharmaceutical additives that are added to 36 Chapter 1….Literature Review maintain and prolong the physical and microbial stability of the formulation. These may include benzalkonium chloride (preservative) and sodium editate (antioxidant and chelating agent). These additives have well-established penetration enhancing effects. This might increase systemic insulin absorption and as such might have an effect on blood glucose levels. Therefore, the aforementioned therapeutic benefits might be offset by possible systemic absorption of insulin via the ocular route, hence potentiating the effect of co-administered hypoglycaemic drugs (Lee et al., 2002). Table 2.1 presents the major attributes of insulin. Table 1.2 Major attributes of insulin Advantages Successfully enhanced corneal healing rate in diabetic rodent Disadvantages Limited clinical trials (unknown mechanism) Insulin (6 kDa protein) cannot easily cross intact corneal barrier Long term safety on ocular tissue is controversial Nerve growth factor (NGF) NGF is a non-covalently linked dimer consisting of two 118-residue polypeptides, each of which contains three intramolecular disulphide bridges. NGF is a prototypical member of the neurotrophin family of growth factors that promote survival and growth of sympathetic and sensory neurons and differentiation of neurons in the central nervous system (Levi-Montalcini, 1987). Its biological action is not restricted to cells of neuronal origin but extends to cells of the immune system. NGF receptors (TrKA) were identified on the human corneal surface (Lambiase et al., 1998a). NGF is expressed constitutively in normal human and rat corneas. Published data supports the possibility that NGF modulates ocular inflammation and corneal epithelial proliferation and differentiation through its receptors. Also, recent evidence has demonstrated that NGF is identified in the cornea, conjunctiva, tear fluid, and lacrimal glands (Ghinelli et al., 2003; Ríos et al., 2007; Touhami et al., 2002; You et al., 2000). 37 Chapter 1….Literature Review Many reports describe the relationship between NGF and the ocular surface health and wellbeing: one in vitro study showed that NGF induces proliferation and differentiation of rabbit corneal epithelial cells (Kruse & Tseng, 1993). A clinical report showed an increase in NGF plasma levels in vernal keratoconjunctivitis (VKC) with a direct correlation between mast cell conjunctival infiltration and NGF levels (Lambiase et al., 1995). Topical murine NGF eye drops was applied to 14 eyes with non-infectious corneal ulcers caused by essential neurotrophic keratitis (5 eyes), chemical burns (3 eyes), abuse of topical anaesthetics (2 eyes), orbital tumour surgery (1eye), surgery of acoustic neuroma (1 eye), penetrating keratoplasty of unknown reason (1 eye) and a lamellar keratoplast for a herpatic vascularized scar (1 eye) (Lambiase et al., 1998b). Remarkably, considering the diverse aetiology, all patients had complete corneal healing after 10 days to 6 weeks of NGF treatment. However, all patients experienced mild to moderate conjunctival hyperaemia accompanied by pain, photophobia, and superficial or deep corneal neovascularisation occurred in 9 of the 14 treated eyes. All ocular symptoms disappeared once the corneal ulcers had completely healed. In another study, it has been reported that murine NGF (dosage: one drop of 200 µg/ml NGF solution every 2 h for 2 days, followed by one drop six times daily until the ulcer showed signs of healing, followed by one drop four times daily of 200 µg/ml NGF solution) improved corneal sensitivity and promoted corneal epithelial healing in 45 eyes with both moderate and severe neurotrophic keratitis within 12 to 42 days of NGF treatment (Bonini et al., 2000). Visual functions such as corneal sensitivity and visual acuity were improved, and healing of the corneal ulcers was promoted in NGF-treated eyes. Interestingly, except in a few cases, no relapse of the disease occurred during the follow-up period. However, transient adverse effects such as hyperaemia and pain in the eye were observed on instigation of NGF treatment. Supporting to that, a report has suggested NGF as a future treatment in several pathologies of both the ocular surface and the retina (Micera et al., 2004). Therefore, NGF has attracted the attention of pharmaceutical scientists to develop specialised drug delivery systems aimed at inhibiting NGF’s degradation, prolonging its biological half-life and enhancing its biological activity (Pfister et al., 2008; Sakiyama-Elbert & Hubbell, 2000; Xie 38 Chapter 1….Literature Review et al., 2005). All of the aforementioned evidence augments the clinical interest in NGF as a potential therapeutic agent to promote corneal wound healing. However, most of the experimental evidence in relation to the clinical efficacy of NGF has been obtained with the 2.5S murine NGF which has several limitations (Table 1.3). Several attempts have been made to produce recombinant human nerve growth factor (rhNGF) using different micro-organisms, including Saccharomyces cerevisiae (Kanaya et al., 1989; Nishizawa et al., 1993) and Escherichia coli (Negro et al., 1992; Rattenholl et al., 2001) as well as insect (Allen et al., 2001; Barnett et al., 1991) and mammalian cells (Iwane et al., 1990). However, most of these studies have been carried out in vitro, and the available evidence in vivo indicates that the action of rhNGFs in human peripheral neuropathies was not comparable with the effect of murine NGF (Apfel, 2002). The technology of rhNGF preparation is not complex; however the biological activity of β-NGF relies on the formation of three disulfide bonds and a cysteine knot within two β-chains of 120 amino acids each after cleavage of pro-peptide sequences from a larger precursor molecule (Edwards et al., 1988; Shooter, 2001; Ullrich et al., 1983). Recently, it has been reported that the production (on a laboratory scale) of large amounts of rhNGF was shown to be effective both in vitro and most importantly in vivo (Colangelo et al., 2005). Table 1.3 Major attributes of murine NGF Advantages Successfully used in many preclinical and clinical trials Disadvantages High cost Limited sources and insufficient quantity (adult male mouse submandibular glands, snake venom) Allodynia (hyperalgesia induced upon topical administration) 39 Chapter 1….Literature Review Hepatocyte growth factor (HGF), epidermal growth factor (EGF), keratinocyte growth factor (KGF), transforming growth factor-β (TGF-β) HGF, EGF, KGF, and TGF-β are expressed in keratinocytes, corneal endothelial cells and upregulated in rabbit keratocytes after wound healing (Li & Tseng, 1996; Wilson et al., 1993). However, conflicting results have been reported in human trials, such as those described for EGF (Daniele et al., 1992; Hosotani et al., 1995). Moreover, TGF-β has been implicated as a potent stimulant of the scarring process in the eye. Corneal scarring is a major problem in influencing results of photorefractive surgery, giving rise to symptoms of haze and resulting in a reduction of the best corrected visual acuity (Jester et al., 1997; Mayers et al., 1997). Opioid growth factor (OGF) Endogenous opioids and their receptors were originally thought to be related to neural modulation (Akil et al., 1984). Although there are numerous endogenous opioids and respective receptors, only the native opioid peptide, [Met5]enkephalin, is classified as an opioid growth factor (OGF). OGF is a pentapeptide with the sequence Tyr-Gly-Gly-Phe-Met. OGF has been determined to serve a growth regulatory role and its binding receptor is not homologous to classic opioid receptors. OGFs act as growth regulators in many normal and malignant tissues (Zagon & McLaughlin, 1991b, 1993a). It has been identified in both eukaryotes and prokaryotes as a potent inhibitor of cell replication (Zagon & McLaughlin, 1991a, 1991b, 1993a). In vertebrates, OGF is also thought to be important in cellular maturation and survival (Meriney et al., 1991; Zagon & McLaughlin, 1993a). OGF has been documented as being produced in an autocrine and/or paracrine, manner (Zagon et al., 1994a). OGF activity is mediated by the zeta (ζ)-opioid receptor (Zagon & McLaughlin, 1993a, 1993b). Endogenous opioid systems influence proliferation in a wide variety of cells and tissues both in vivo and in vitro. These include developing neurons and galia of the mammalian nervous system (including the retina), glial cells in culture, myocardial and epicardial cells in neonatal heart, neuronal and non-neuronal cancer cells, and cellular renewal in the epithelium of the adult tongue (McLaughlin, 1994; Zagon & McLaughlin, 1991a, 1991b, 1993a; Zagon et al., 1994b). OGF has been detected in the bovine corneal epithelium (Tinsley et al., 1990; Tinsley et al., 1989). OGF and the ζ-receptor are also present in the normal rabbit corneal epithelium (Zagon et al., 1995). 40 Chapter 1….Literature Review A study was performed to show the physiological role of endogenous opioid systems (opioid peptides and receptors) in the mammalian corneal epithelium in vitro using rabbit corneal explants (Zagon et al., 1995). The following strategies were adopted; firstly, naltrexone (opioid antagonist) was administered to explant cultures of rabbit corneal epithelium to examine whether this approach modulates epithelial outgrowth; secondly, OGF was administered to cultures of corneal explants to investigate the effects of this opioid peptide on the extent and organization of epithelial outgrowth. Immunocytochemistry experiments using antibodies specific to OGF and ζreceptor were performed to confirm that OGF systems are present in normal corneal epithelium in culture. The study demonstrated that OGF and ζ-receptor are present in the normal corneal epithelium and, more surprisingly, OGF exerted an inhibitory effect on cellular proliferation and growth in the corneal epithelium. This tonically active growth regulatory inhibitory pathway is involved in cell replication. OGF targets cyclin-dependent kinase inhibitors (P16 and/or P21) (McLaughlin et al., 2010). The regulation of cells proliferation by OGF-OGF receptor (OGFr), as aforementioned, has been termed OGF-OGFr axis for cellular renewal and proliferation (McLaughlin et al., 2010). Fluorescence imaging of OGFr by green fluorescent protein-labelled OGFr revealed that OGFr is assembled into outer nuclear envelope. This complex is shuttled to the nucleus. The OGF-OGFr complex unregulated p16 and/or p21 leading to a marked delay in the G1-S phase of the cell cycle. Naltrexone (a long acting potent opioid antagonist) administration blocks OGF-OGFr interactions. Hence, this increases cell proliferation (Cheng et al., 2009). Blockade of opioid receptors by naltrexone (a long acting potent opioid antagonist) for seven days significantly increased the extent of proliferation and outgrowth of the corneal epithelium in vitro. This study suggested, for the first time, an inhibitory role for an endogenous growth factor. Moreover, the study demonstrated that blocking of the corresponding receptor by naltrexone significantly increased DNA synthesis, cell migration and proliferation of corneal epithelium. Other reports have demonstrated that topical or systemic application of naltrexone markedly accelerates DNA synthesis and wound healing of corneas in humans (Zagon et al., 2000), rats (Zagon et al., 1998b) and rabbits (Zagon et al., 1998a) as well as in diabetic rodents (Zagon et al., 2002a). Indeed, topical application of naltrexone healed corneal epithelial wounds in uncontrolled Type 1 diabetic rodents at rates equal to or surpassing that in normal controls (Klocek et al., 41 Chapter 1….Literature Review 2007). However, concomitant application of topical naltrexone and topical insulin does not have an additive or synergistic effect on corneal re-epithelialisation (Klocek et al., 2009). In another study, the same group used a gene gun to introduce sense and antisense complementary DNA (cDNA) for OGFr to establish that the effects of OGF on corneal reepithelialization are specifically dependent on the OGFr. This study establishes, for the first time, that the OGF-OGFr axis is an autocrine loop that serves as a crucial regulatory system for corneal epithelial wound healing. The plasmid pcDNA3.1-OGFr, carrying the rat OGFrcomplementary DNA in both the sense and antisense orientations, and control, were delivered by gene gun to the rat cornea. After 24 h, corneas were abraded and re-epithelialisation was monitored by fluorescein photography. Corneal cells exposed to sense OGFr exhibited corneal defects that were up to 52% larger than the control group. On the other hand, corneas subjected to the antisense constructs of OGFr exhibited defects that were up to 56% smaller than the control group (Zagon et al., 2006b). Naltrexone has been found to enter human cells by passive diffusion. Fluorescent-tagged naltrexone was found to accumulate in the cytoplasm of the human head and neck squamous cell carcinoma cell line as little as 60 s after incubation. Neither fluorescein dye nor fluoresceintagged dextrans were found to enter the cells (Cheng et al., 2009). In the same study, untagged naltrexone and fluorescent-tagged naltrexone were found with an increase of up to 70% in cell proliferation markers compared with the vehicle-treated group (Cheng et al., 2009). Table 1.4 summarises the major features of naltrexone. 42 Chapter 1….Literature Review Table 1.4 Major attributes of naltrexone Advantages Successfully reversed most signs of diabetic keratopathy (corneal sensation, delayed wound repair, dry eye) Well-defined mode of action Synthetic chemical entity (commercially available) Safe on ocular tissues Disadvantages Extremely bitter taste (nasolacrimal drainage) 1.3.5.3. Substance P Substance P (SP) (11-amino acid polypeptide) is a neurotransmitter of the trigeminal nerve and a member of the tachykinin family. SP is present in physiologically relevant concentrations in the normal cornea (Nishida et al., 1996). SP concentrations in tears are thought to reflect the neuropeptide levels in ocular tissues. In patients with unilateral corneal hypesthesia, substance P concentrations in tears from the affected eye were significantly lower than those in tears from the unaffected eye (Yamada et al., 2000). Topical application of SP alone has no detectable effect on the rate of corneal epithelial wound closure in rabbits (Kingsley & Marfurt, 1997). However, SP in combination with IGF-1 has been demonstrated to facilitate synergistically corneal epithelial migration in rats with trigeminal denervation (Nagano et al., 2003) and in organ cultured corneas (Nishida et al., 1996). As discussed above, this combination has been shown to be effective in healing corneal epithelium in humans in the clinical setting. Unfortunately, SP is not available commercially. 1.3.5.4. Aldose reductase inhibitors Aldose reductase (AR) is the rate-limiting enzyme in the polyol pathway in which glucose is converted to sorbitol (Herse, 1988). AR activity is implicated in diabetic corneal epithelial disease primarily because of the effect that AR inhibitors have on these documented diabetic epithelial abnormalities (Herse, 1988). The AR inhibitor was applied topically to treat a non43 Chapter 1….Literature Review healing corneal ulcer in a patient with diabetes. The corneal ulcer healing was dramatically improved (Ohashi et al., 1988). Topical CT-112 (an AR inhibitor) was used to treat two patients with diabetes with non-resolving corneal epithelial lesions (Ohashi et al., 1988). Not only did the lesions resolve, but they returned when the CT112 therapy was discontinued and resolved again when CT-112 therapy was reinstituted. The patients remained lesion-free on long-term maintenance therapy with CT112. Also, topical CT-112 was applied to a group of galactosaemic rats, vehicle alone to a second group of galactosaemic rats and vehicle alone to a third group of normal rats (Yokoi et al., 1997). After 3 weeks, fluorescein uptake (as an indicator of corneal permeability) was measured by fluorophotometry in the three groups. The galactosaemic rats that received vehicle alone displayed a significantly higher sodium fluorescein uptake than the normal rats and the CT-112 treated galactosaemic rats showed no significant difference in the fluorescein uptake between the vehicle-treated normal rats and the CT-112-treated galactosaemic rats. Treatment with AR inhibitors has been shown to reduce corneal changes after they occurred, and limit the development of these corneal changes when compared to untreated controls in both animals and humans. The mechanism by which increased AR activity contributes to diabetic corneal pathology remains unknown (Herse, 1988). 1.3.5.5. Miscellaneous Nicergoline is an ergoline derivative known to cross the blood-brain barrier and is now widely and safely used to treat cognitive impairment from stroke and degenerative dementia. In vivo studies have shown that nicergoline treatment induces significant increase in the NGF levels within the frontal region of the brain and supports cholinergic neurons, increasing the content of NGF and brain-derived neurotrophic factor in the brain of aged rats (Giardino et al., 2002; Nishio et al., 1998). Recently, it has been found that oral nicergoline (10 mg/kg per day) for 2 weeks increased the rate of corneal wound healing in 50 rat eyes (Kim et al., 2009). The promoting effect of nicergoline in corneal wound healing is attributed to increased NGF in corneas and/or lacrimal glands. However, the exact ocular tissue responsible for increasing NGF levels as a response to nicergoline has not yet been identified. Recently, oral treatment of nicergoline at least for 4 weeks has been found to restore corneal sensation in rats after photorefractive keratectomy (Kim et al., 2011). 44 Chapter 1….Literature Review 1.4. Thesis aims and structure The cornea is the most densely innervated mammalian tissue and the integrity of corneal nerves is of paramount importance for the refractive and protective functions of the cornea. With the advent of modern refractive surgeries and the increasing number of diabetic patients worldwide, the cornea is routinely subjected to injury and this damage may cause transient or chronic neurotrophic deficits (corneal scarring). If this damage does not receive proper medical attention loss of vision can ensue. The corneas of diabetic patients show significant and characteristic signs known collectively as diabetic keratopathy. These characteristic diabetic signs include: delayed wound repair, reduced tear production, reduced corneal sensation and neurotrophic corneal ulcers. The prevalence of diabetic keratopathy is approximately 47% to 64% during the course of diabetes mellitus (Schultz et al., 1981 ). The available treatment options are topical antibiotics, tear substitutes and eye patches. This treatment is in most cases insufficient because it does not treat the underlying causes (McLaughlin et al., 2010). Naltrexone (a potent opioid antagonist) has recently been found to reverse the signs and symptoms of diabetic keratopathy. Naltrexone successfully accelerates corneal wound healing, both systemically and topically, in normal and diabetic rats, rabbits and humans (McLaughlin et al., 2010; Zagon et al., 2009; Zagon et al., 1997; Zagon et al., 1998a, 1998b; Zagon et al., 2007b). Recently, it has been found that naltrexone can restore impaired corneal sensation in diabetic rats (McLaughlin et al., 2010; Zagon et al., 2009). These effects are attributed to blocking opioid growth factor receptors in corneal tissues and consequently accelerate DNA synthesis in the cornea and re-epithelialisation. Whilst there is substantial clinical evidence on the effectiveness of naltrexone in diabetic keratopathy, there are hardly any reports on its formulation or incorporation into an appropriate ocular delivery system. Naltrexone has an extremely bitter taste. This bitterness can be tasted upon ocular administration of the drug from simple solution due to rapid and extensive nasolacrimal drainage. There are also limited reports on the chemical stability of naltrexone in aqueous solution. Parallel to that, frequent instillation of a simple drug solution to the eye is associated with pulse-type entry and may prompt patient incompliance. 45 Chapter 1….Literature Review As previously outlined, niosomes lend themselves to potential ocular drug delivery systems. They have the convenience of being delivered as eye drops, the ability to prolong precorneal drug residence, a localised effect, complete biodegradability and minimal irritancy. This work aims to formulate and characterise niosomal formulations for the ocular delivery of naltrexone hydrochloride. To achieve this aim, this thesis will explore the following objectives: • Develop and validate an analytical method for accurate and selective determination of naltrexone hydrochloride in the prepared formulations. In addition, forced degradation studies will be conducted under different stress conditions to demonstrate that the developed method is a stability-indicating method. • Study the chemical stability of naltrexone hydrochloride in aqueous solutions. • Study the fundamental physicochemical properties of naltrexone hydrochloride such as melting behaviour, aqueous solubility and lipophilicity (octanol-buffer partition coefficient determination). • Perform formulation studies on niosomes that includes: o Investigating the effect of surfactant type (sorbitan esters and ployoxyethylene alkyl ethers) and cholesterol (membrane stabiliser) on vesicle sizes, entrapment efficiency and vesicle membrane fluidity (gel/liquid phase transition temperatures). Also, the effect of total lipid and initial drug concentration on the entrapment efficiency of the prepared niosomes. o Determining whether some selected membrane additives could alter the physical properties such as drug entrapment efficiency, morphology, size and gel/liquid transition temperatures of the prepared niosomes with emphasis on topical application to the eye surface. o Determining the physical and chemical properties of the prepared niosomes such as contact angle, spreading coefficient, viscosity measurements and influence of vesicular encapsulation on oxidative degradation of naltrexone hydrochloride. • Evaluate the prepared niosomes for ocular naltrexone hydrochloride delivery; ocular toxicity using suitable in vitro models; in vitro drug release; ex vivo drug permeation through excised bovine corneas; and physical stability of the selected formulations at different temperatures. 46 Chapter 2….Preformulation of Naltrexone hydrochloride 2. Preformulation studies of naltrexone hydrochloride 2.1. Introduction Preformulation characteristics are certain fundamental physical and chemical properties of a drug that can eventually help in selecting the suitable formulation approach to develop an effective, stable and economic dosage form (Wells & Aulton, 2007). The selection of particular preformulation properties to study depends on the proposed dosage form. For example, powder flow, compressibility, intrinsic solubility and dissolution studies are essential for developing solid dosage forms. Generally, fundamental preformulation experiments include: development of a suitable analytical method; aqueous solubility studies; melting point measurement; partition coefficient and dissociation constant determination. Studying these preformulation characteristics for the drug under investigation is essential especially at the early stages of designing and developing a new dosage form. Up to the early seventies, the available compounds for treatment of narcotic addiction were cyclazocine and naloxone. These two drugs proved to be excellent opioid receptors antagonists but both had drawbacks (Fink et al., 1968; Freedman et al., 1968; Renault, 1981). Cyclazocine was limited because of dysphoric side effects due to weak agonistic activity (not pure antagonist). Naloxone is a pure antagonist and thus free of side effects, but lack potency when administered orally (Blumberg & Dayton, 1973; Freedman et al., 1968; Renault, 1981). Naltrexone has been developed to fulfil the need for a long-acting, potent and pure antagonist with minimal unwanted side effects. These requirements are met by naltrexone; it is about twice as active as naloxone as an antagonist in mice. Furthermore, it is longer acting than naloxone in human (Blumberg & Dayton, 1973). Currently, naltrexone is a widely used opioid antagonist for treatment of heroin dependence (Renault, 1981) and alcoholism (Sweetman, 2007). Naltrexone is marketed commercially as hydrochloride salt. Naltrexone hydrochloride is administered orally either in liquid (extemporaneously prepared) or solid (tablets and capsules) form. The capsule form is preferable, because naltrexone has an extremely bitter taste (Renault, 1981). 47 Chapter 2….Preformulation of Naltrexone hydrochloride 2.1.1. Physical properties of naltrexone hydrochloride Naltrexone hydrochloride (NTX) is a hydrochloride salt of naltrexone. It is chemically named Ncyclopropylmethyl-14-hydroxydihydromorphinone. NTX is a white crystalline solid, soluble in water. NTX has an empirical formula of C20H23NO4. HCl (mol wt 377.87) and a melting point of 275oC (crystals from methanol)(O’Neil et al., 2006). It has an extremely bitter taste (Crabtree, 1984); has a low partition coefficient (P), a log P (octanol/water) value of 1.9 (Galichet et al., 2006); it is a weak basic drug, a pKa value of 8.38 at 20oC (Kaufman et al., 1975). Figure 2.1 shows a schematic of the chemical structure of NTX. Figure 2.1 Chemical structure of NTX 2.1.2. Pharmacological uses and dosage for systemic administration of NTX NTX is widely used in treatment of alcohol dependence and rescue medications to reverse the side effects of opioid agonists (respiratory depression overdose) (Goodman et al., 2007; Kastrup et al., 2009). Off-labelled uses of NTX include treatment of obesity, severe pruritus, psychosis and Parkinson’s disease (Goodman et al., 2007). The usual dosage of NTX is 50 mg once daily or 350 mg per week in three divided doses (Crabtree, 1984; Goodman et al., 2007). 2.1.3. Pharmacokinetics of NTX Absorption: NTX is a pure opioid receptor antagonist. Although it is absorbed well orally, naltrexone is subject to significant first-pass metabolism. Its oral bioavailability ranges from 5% to 40% (Kastrup et al., 2009). 48 Chapter 2….Preformulation of Naltrexone hydrochloride Distribution: The volume of distribution for NTX following iv injection is estimated to be 1350 litres. In vitro tests of human plasma show NTX to be 21% bound to plasma proteins over the therapeutic dose range (Kastrup et al., 2009). Metabolism: The systemic clearance (after iv administration) of naltrexone is 3.5 ml/min. The major metabolic site is the liver and the major metabolite is 6-β-naltrexol. The primary metabolic pathway is glucuronide conjugation (Misra, 1981). Excretion: Urinary excretion is the major excretion route of NTX and its metabolites (> 95%). Less than 5% of an orally administered dose is found in faeces after 48 h (Crabtree, 1984; Misra, 1981). 2.1.4. Adverse reactions of systemic administration of NTX Combined reporting of adverse events from oral and injectable formulations includes: Cardiovascular syncope (13%), headache (25%), insomnia (14%), anxiety (12%), nausea (33%), vomiting (14%), abdominal pain and cramps (11%) (Fuller & Sajatovic, 2009; Kastrup et al., 2009). Contrary to expectation, injectable forms of NTX seem to induce more adverse reactions than oral forms, leading to a higher rate of treatment discontinuation by patients. Amongst these adverse reactions associated with injectable NTX administrations are pain, bruising, induration, angioedema and nodules at the injection site (Roozen et al., 2007). 2.1.5. Warnings on the systemic administration of NTX The following warnings are reported with oral administration of NTX (Kastrup et al., 2009): • NTX has the capacity to cause hepatocellular injury when given in excessive doses. • NTX is contraindicated in acute hepatitis or liver failure, and its use in patients with active liver disease must be carefully considered in light of its hepatotoxic effects. • Patients should be warned of the risk of hepatic injury and advised to stop the use of NTX and seek medical attention if they experience symptoms of acute hepatitis. • NTX and its primary metabolite are excreted primarily in urine and caution is recommended in administering the drug to patients with renal impairment. 49 Chapter 2….Preformulation of Naltrexone hydrochloride 2.1.6. Naltrexone as a new ophthalmic pharmaceutical Topical ocular application of NTX in doses of up to 0.4 mg/ml has been found to accelerate markedly the wound healing of cornea in humans (Zagon et al., 2000), rats (Zagon et al., 1998b) and rabbits (Zagon et al., 1998a), as well as diabetic rodents (Zagon et al., 2002a). Moreover, topical application of NTX can enhance diabetic corneal epithelial healing without causing morphologic abnormalities in the reassembly of adhesion structures (Zagon et al., 2007b). Furthermore, topical treatment with NTX has been found to normalise tear production and corneal sensitivity in type 1 diabetic rats (Zagon et al., 2009). The mechanism of NTX for accelerating corneal wound healing is the blockade of opioid growth factor (OGF) interaction with the OGF receptor (Zagon et al., 2002b). Consequently, it can enhance DNA synthesis and corneal epithelialisation. However, the exact mechanism for normalising tear production and restoring corneal sensation in diabetes mellitus is still unknown (Zagon et al., 2009). It is worth mentioning that the corneal safety of topically applied NTX has been studied (Zagon et al., 2006a). The results showed that concentrations of naltrexone ranging from 10-3-10-7 M are not toxic when applied topically to the cornea. 2.1.7. Chemical stability of NTX There are few reports on the chemical stability of NTX and its major degradation pathways in aqueous solutions (Fawcett et al., 1997; Gupta, 2008). Table 2.1 summarises some of the reported HPLC methods for NTX analysis. The physicochemical stability of NTX oral liquid formulations prepared from commercially available tablets and powder at two different concentrations (1 mg/ml and 5 mg/ml) was studied (Fawcett et al., 1997). A vehicle of pH 3.5 was used and contained ascorbic acid (0.5%), sodium benzoate (0.1%), glycerol (20%) and distilled water (to 100%). The prepared formulations were stored in the dark at three temperatures 4, 25, and 70oC and monitored for chemical and physical stability of NTX up to 90 days. The samples were assayed for NTX concentration to assess decomposition over 90 days. The results showed that decomposition of both NTX from drug powder and tablet stored at 4 and 25oC was not significant over 90 days. However, the percentage of NTX remaining after storage for 90 days at 70oC was 51.1 % and 64.1% for NTX liquid formulations prepared from tablet and powder respectively. In 50 Chapter 2….Preformulation of Naltrexone hydrochloride terms of physical stability, powder formulations stored at 4oC developed a pale yellow colour after 90 days. Both solutions prepared from powder and tablets stored at 25 and 70oC developed a pale yellow colour by day 30 and day 2 respectively. In another study, the chemical stability of a NTX injection (extemporaneously prepared) was studied. NTX was dissolved in sufficient sterile water for injection under aseptic conditions to form a concentration of 1.4 mg/ml. The solution was stable for at least 42 days when stored in clear glass vials at room temperature. Also, the physical appearance of the injection did not change during the study period (Gupta, 2008). Having a morphinan structure, oxidation could be the main culprit for the loss of potency and degradation of NTX. Furthermore, oxidation is the main degradation pathway for phenolic compounds such as opioids and catecholamines (Florence & Attwood, 1998a; Lachman et al., 1986). 51 Chapter 2….Preformulation of Naltrexone hydrochloride Table 2.1 Summary for HPLC conditions of NTX determination Author Chromatographic conditions Column (Fawcett et al., 1997) (Iyer et al., 2007) (Gupta, 2008) C18: SupelcosilTM LC-18-DB (250 mm × 4.6mm, 5µm) C18: SupelcosilTM LC-18-DB (150 mm × 4.6mm, 5µm) C18: Beckman UltrasphereTM LC-18-DB (150 mm × 4.6mm, 5µm) Mobile phase A : 0.06% triethylamine solution in Phosphate buffer (40mM) pH 5 (50%v/v) B: Acetonitrile (50% v/v) A : 0.1% of 85% phosphoric acid and 0.1% glacial acetic acid solution (90% v/v) pH 2.6 B: Acetonitrile (10% v/v) Injection volume (µl) Elution type Temperature (oC) Flow rate (ml/min) Detection Retention time (min) Linearity range (µg/ml) 10 Isocratic Ambient 1 UV at 214 nm 6 10-200 A : 0.06% triethylamine solution in Phosphate buffer (40mM) pH 4.75 (88 parts)B: Acetonitrile (12 parts), A and B were premixed together 5 Isocratic 50 1.25 UV at 204 nm 4.5 0.16-20 80 Isocratic Ambient 1.5 UV at 285 nm 4 140-300 52 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.1.8. Oxidation Oxidation is one of the most common causes of the chemical degradation of a considerable number of pharmaceutical compounds (Florence & Attwood, 1998a; Lachman et al., 1986). For instance, the degradation of steroids, vitamins, antibiotics and morphine is due to oxidation reactions (Lachman et al., 1986). Oxidation involves the removal of electrons, radicals or electropositive atoms (Florence & Attwood, 1998a). The most common form of oxidation occurring in pharmaceutical dosage forms is autoxidation, which is the reaction between molecular oxygen and organic compounds under mild conditions (Johnson & Gu, 1988). Apart from the loss of drug potency due to autoxidation, other instability problems can occur such as discolouration, changes in dissolution rate, precipitation, loss of preservative efficacy, evolving of off-odours and sometimes generation of toxic compounds (Johnson & Gu, 1988). It is well accepted that most liquid-phase autoxidations are free-radical chain reactions. A free radical is an atom or molecule with one or more unpaired valence electrons. Three well-distinct steps are identified in autoxidation reactions (Johnson & Gu, 1988): Initiation: equation 2.1 This stage is initiated once free radicals are generated. This inevitably takes place over a certain period of time but the generation of free radicals can be accelerated by the action of heat, light and traces of heavy metals (Florence & Attwood, 1998a; Johnson & Gu, 1988). The length of the initiation is called the induction period (Stewart & Tucker, 1984). Propagation: equation 2.2 In the propagation stage, the generated free radicals in the initiation stage react with molecular oxygen to yield peroxy radicals. Therefore, the oxygen concentration is of importance in the autoxidation process, especially when determining the reaction rate of oxidation at different temperatures due to the effect of temperature on the oxygen content (concentration of oxygen) of the liquid (Lachman et al., 1986); the higher the temperature, the lower the oxygen concentration. The hydroperoxide reacts with a new available free radical to propagate the reaction. 53 Chapter 2….Preformulation of Naltrexone Hydrochloride Termination: equation 2.3 The oxidation reaction proceeds until the free radicals are scavenged by inhibitors or halted by side reactions which eventually break the chain reactions. Figure 2.2 is a schematic showing a typical example of autoxidation reactions of morphine (M) with atmospheric oxygen in aqueous solution to morphine N-oxide (MNO) (Yeh & Lach, 1961). 54 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.2 Schematic of morphine autoxidation in aqueous solutions 55 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.1.9. Arrhenius plot and predicting drug stability Temperature is a key factor in determining the velocity of the degradation reaction of pharmaceutical products. It is necessary to evaluate the temperature dependency of the reaction. This allows prediction of the stability of the product (Lachman et al., 1986). The speed of many degradation reactions doubles for each 10o rise in temperature (Lachman et al., 1986). The most satisfactory method for quantitative determination of the influence of temperature on the reaction velocity is that proposed by the Arrhenius equation which will be described later in the method section in this chapter. The most important parameter which can be calculated from the Arrhenius equation is the heat of activation (∆Ea), which represents the energy the reaction molecules must acquire to undergo reaction. Plotting reaction rate constant values obtained at several elevated temperatures versus the inverse of absolute temperature gives rise to what is known as the Arrhenius plot. Since the plot is linear, the prediction of stability at shelf temperature is possible by extrapolating the curve to the lower temperature (Lachman et al., 1986). The use of the Arrhenius equation to predict drug stability at shelf temperature is most useful when the reaction is too slow to be monitored conveniently and when∆Ea is relatively high. For example, for a reaction with a ∆Ea value of 25 kcal/mol, a temperature increase from 25 to 45 oC brings about a 14-fold increase in the reaction rate constant (Guillory & Poust, 2002). 2.1.9.1. Limitations of the Arrenius plot for predicting oxidative degradation There are some issues related to accelerated thermal degradation methods hindering the ability of the Arrhenius equation to predict the shelf-life of dosage forms at lower temperatures (Lachman et al., 1986; Waterman et al., 2002): • Higher temperatures may evaporate solvents, thus producing unequal moisture concentration. • At higher temperatures, there is less oxygen solubility, thus hindering the predictability of the room temperature stability of drugs sensitive to oxidation. • The permeability of oxygen through packaging is temperature dependent. • If the generation of radicals is rate limiting, the kinetics can show autocatalysis, the rate of reaction increases as the radical concentration increases. • Instability may be related to the amount of peroxide impurities present in a particular lot of excipient. This is related to age of the excipient and the conditions under which the material was stored. 56 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.2. Chapter aims Designing a new ophthalmic liquid dosage form for NTX compels performing preformulation studies for NTX. This chapter aims at carrying out preformulation studies of NTX which can help select subsequent approaches during the formulation development. Specific objectives include: • Development and validation of a sensitive and standard HPLC method for NTX analysis in the prepared formulation. • Studying various stress conditions which can affect the chemical stability of NTX in accordance to International Conference on Harmonization (ICH) guidelines. • Determining the following fundamental physicochemical properties of NTX: o Melting behaviour. o Spectrometric fingerprints (infrared, mass and 1H-NMR spectra). o Aqueous solubility. o Ionisation constants (pKa values) at different temperatures. o Lipid solubility (distribution and partition coefficients). 57 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.3. Materials and methods 2.3.1. Materials NTX was purchased from Mallinckrodt Inc., St. Louis, MO, USA; n-octanol was purchased from Schalau Chemie, Sentmenat, Spain. All other solvents and buffer salts were of analytical grade and used as received. 2.3.2. Methods 2.3.2.1. Melting behaviour study of NTX using differential scanning calorimetry (DSC) A small amount (typically 2 mg) of NTX powder was weighed in an aluminium pan and covered with an aluminium lid and hermitically sealed using a pan press (Thermal Science, USA). The temperature of the pan gradually increased from 50 to 300oC at a rate of 10oC/min using a differential scanning calorimeter (Q 1000 series, Thermal Analysis, USA) precalibrated with indium. The purging gas was nitrogen at a flow rate of 45 ml/min. 2.3.2.2. Spectrometric identification of NTX Mass spectroscopy (MS) A sample weight (approx. 1 mg) was dissolved in 1 ml methanol and injected directly into the ionisation chamber. The atmospheric pressure chemical ionisation (APCI) technique was used to ionise the sample and generate ions using a LC-M spectrometer (Surveyor MSQ, ThermoFinnigan, New Jersey, USA). Fourier transform infrared (FT-IR) spectroscopy A FT-IR spectrometer (Bruker Miracle Micro ATR, Tensor 37, GmbH, Germany) was used to record the FT-IR spectrum. A clean diamond window was used to measure the background spectrum. A sufficient amount (approx. 1mg) of NTX was placed to form a thin film covering the diamond window. The data was acquired and analysed using OPUS software (OPUS 6.5, Germany).The FTIR spectrum was recorded at spectral resolution of 2 cm-1 with an average of 120 scans. Proton nuclear magnetic resonance (1H-NMR) spectroscopy 1 H-NMR spectrum of NTX was determined in deuterated dimethyl sulfoxide (CD3)2SO on a 1H- NMR spectrometer (Bruker Avance 400 Spectrometer, GmbH, Germany) at 400 MHz. Chemical shifts are recorded in ppm. 58 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.3.2.3. Analytical methods Ultraviolet (UV) spectrophotometry An amount of NTX (10 mg) was accurately weighed and dissolved in 10 ml of phosphate buffer saline (PBS) pH 7.4 to produce a concentration of 1mg/ml. Aliquots of 0.3-1ml of the stock solution were transferred in triplicate to a series of 10-ml volumetric flasks and diluted to 10 ml using PBS to obtain final concentrations of 30-100 µg/ml. Then, absorbance values of these dilutions were measured at 281.5 nm using a UV-spectrophotometer (Thermospectronic, Helios Gamma, England). A plot of absorbance vs. concentrations (µg/ml) was constructed. High performance liquid chromatography (HPLC) method development A modification of the method described by (Fawcett et al., 1997) was performed. In brief, an HPLC system (Agilent 1200, Agilent Corporation, Germany) comprising a quaternary pump, an automatic sampler and a photodiode array (PDA) detector was used with data acquisition by ChemStation® software (Agilent Corporation, Germany). The chromatographic separation was achieved using a Supelcosil C18 column (5 µm; 250 mm×4.6 mm, Supelco Corporation, PA, USA) maintained at 25◦C. The mobile phase was prepared and premixed. Seventy parts (by volume) of monobasic potassium dihydrogen phosphate solution (40 mM) pH 4 were mixed with thirty parts (by volume) of acetonitrile; and the isocratic flow rate was 0.85 ml/min. A water: acetonitrile (70:30 v/v) mixture was used as a rinse solution for the injector, and the injection volume was fixed at 5 µl. Detection was carried out using a wavelength of 214 nm. The peak purity for NTX was also determined. Validation of the HPLC analytical method The developed method was validated according to International Conference on Harmonisation (ICH) guidelines (Shah et al., 1992). Approximately 10 mg of NTX were accurately weighed, transferred into a 10 ml volumetric flask and dissolved in triple-distilled water obtained in-house by reverse osmosis and commercially known as Milli-Q water (MilliQ, Millipore, USA) to yield a stock solution of 1mg/ml. From the stock solution, a serial dilution was performed using the water:acetonitrile (70:30 v/v) mixture to yield a standard calibration curve with a concentration range from 5 to 50 µg/ml. 59 Chapter 2….Preformulation of Naltrexone Hydrochloride Precision and accuracy Intra-day and inter-day variabilities were determined by repeated injections of quality control (QC) samples. The QC samples were prepared at 8, 18 and 40 µg/ml representing low, middle and high controls respectively. Accuracy was assessed by comparing the predicted concentrations of the QC samples with the nominal 8, 18 and 40 µg/ml concentrations. Limit of detection (LOD) and limit of quantitation (LOQ) LOD is the lowest amount of analyte in a sample which can be detected but not necessarily quantitated as an exact value, calculated using equation 2.1. LOD = 3σ S equation 2.1 Where σ is the standard deviation of the intercept, and S is the slope of the calibration curve LOQ is the lowest amount of analyte in a sample, which can be quantitatively determined in suitable precision and accuracy calculated using equation 2.2. LOQ = 10σ S equation 2.2 Specificity (peak purity determination) The HPLC-PDA detector with the Agilent ChemStation software allowed on-line data acquisition of UV-spectra between 190-400 nm during peak elution. The PDA detector provides more information on sample composition than single-wavelength detection (Krull & Szulc, 1997; Sinha et al., 2007). The UV spectra were obtained at four different points across the NTX peak; two points before the peak apex (leading front), one point at the apex and one point after the apex (tailing front). The peak purity was assessed by examining the similarity of the UV spectra obtained at the five points. If an impurity or degradation product coelutes with the NTX peak, the five UV spectra obtained across the peak are different. The peak purity analysis was carried out for all NTX samples throughout the study. 2.3.2.4. Forced degradation studies NTX was subjected to common stress conditions according to ICH guidelines to force NTX degradation. The stress conditions were acid (0.1 M HCl) at 60oC, base (0.1 M NaOH) at 60oC, oxidative (2% v/v H2O2) at room temperature, thermal (PBS pH 7.4) at 60oC and photolysis (exposure to artificial daylight illumination of 10,000 lux) at 40oC for 24 h (Alsante et al., 2003; 60 Chapter 2….Preformulation of Naltrexone Hydrochloride Reynolds et al., 2002). The samples were allowed to cool, neutralise (acid and base samples), be filtered and be diluted before injection into the chromatographic system. All experiments were performed in triplicate. 2.3.2.5. Degradation product identification using electron spray ionisation-mass spectroscopy (ESI-MS) To identify peaks of the resultant degradation products, samples were run on a Bruker MicroOTOFQ mass spectrometer (Bruker Daltronics, Bremen, Germany) coupled with a KD Scientific syringe pump (KD Scientific, Holliston, Massachusetts, USA). Samples were diluted to 1-10 µg/ml and were introduced by direct infusion at 180 µl/h into an ESI source in a positive ionisation mode (capillary voltage -4500V) using 99.999% nitrogen gas as drying nebuliser gas (180oC, 5 L/min and pressure of 0.5 bar). External mass calibration was performed using sodium formate clusters and an enhanced cubic fit for the mass curve. Data was acquired for 3 min over a 50 to 1000 m/z range. Spectra were averaged over the 3 min and processed using Compass software 1.3 (Bruker Daltonics, Bremen, Germany). 2.3.2.6. pH-degradation rate studies NTX solutions were prepared at a concentration of 0.4 mg/ml in phosphate buffers of pH 4.5-9.5. The buffer solutions were prepared by dissolving a suitable amount of NaH2PO4 or Na2HPO4 to make up final concentrations of 40 mM with MilliQ water. The pH of the final solutions was adjusted using H3PO4 (1M) or NaOH (1M) solutions to the desired pH values at room temperature using a pH-meter (SevenEasy, Mettler-Toledo, Schwerzenbach, Switzerland) precalibrated using standard buffers at pH 4 and pH 7 (BDH, England). The prepared NTX solutions at pH 4.5-9.5 were incubated in a stability chamber (BINDER, Tuttlingen, Germany) at 60oC. 2.3.2.7. Accelerated stability and chemical kinetic studies Screw-capped glass vials containing NTX solutions (0.4mg/ml) in PBS pH 7.4 were stored at 40, 50, 60 and 70oC to force degradation of NTX. Aliquots were withdrawn at suitable time intervals depending on the decomposition rate, and analysed immediately using the HPLC method described above. The observed first order degradation rate constant (kobs) was calculated from the slope of log C versus time (t) plot according to equation 2.3 (Lachman et al., 1986). 61 Chapter 2….Preformulation of Naltrexone Hydrochloride kobst C log = − 2.303 Co equation 2.3 Where C is the percentage of the drug remaining at time (t) and Co is percentage of the drug at t equal to zero. Plotting log kobs at different temperatures versus the reciprocal of absolute temperature (1/T) in degrees absolute (oK) gives rise to what is known as the Arrhenius plot and the plots were then fitted to the Arrhenius equation (equation 2.4) (Lachman et al., 1986). log Kobs = log A − ∆Ea 2.303RT equation 2.4 Where A the Arrhenius factor or the frequency factor (a measure of the frequency of collisions between the reacting molecules),∆E a is the activation energy and R is the gas constant (1.987 cal.mol-1. degree-1). The kobs values at lower temperatures can be estimated using equation 2.5. log ∆Ea T 2 − T 1 kobs 2 = kobs1 2.303R T 2T 1 equation 2.5 2.3.2.8. Solubility studies of NTX Equilibrium solubility studies of NTX were performed in both MilliQ water and PBS (NaCl 137 mM, Na2HPO4 10 mM, KH2PO4 1.47 mM, KCl 2.68 mM) pH 7.4 at 25oC ± 0.5oC. An excess amount of NTX was dispersed in 5 ml of either PBS solution or water in 10 ml-screw-capped glass vials and shaked at 200 strokes/min using a thermally-controlled shaking water bath (Heto SBD 50, Jouan Nordic, Denmark) for 24 h and left for another 24 h to attain equilibrium. The samples were withdrawn through 0.22 µm filter membranes, properly diluted and analysed spectrophotometrically at 281.5 nm. 2.3.2.9. Determination of n- octanol/PBS distribution coefficient and partition coefficient of NTX The distribution of NTX between equal volumes of octanol (oil phase) and PBS at pH 7.4 (aqueous phase) was determined at three different temperatures 25, 35 and 37oC. Both n-octanol and PBS were mutually saturated over overnight for at least 12 h at the corresponding temperature before use. A specified amount of NTX was dissolved in presaturated PBS with the oil phase to make a final concentration of 1 mM. Aliquots of 2 ml of the NTX solution were 62 Chapter 2….Preformulation of Naltrexone Hydrochloride mixed with equal volumes of the oil phase in screw-capped glass vials and vortexed for 2 min. The glass vials were left for a period of 12 h which was sufficient to allow the drug to reach distribution equilibrium between the two phases (preliminary experiments showed that equilibrium was reached at this time). After equilibration, a Pasteur pipette was used to separate the two phases, followed by centrifugation to remove any of the remaining immiscible octanol phase. The drug concentration in the oil phase was determined by subtracting the drug concentration in the aqueous phase before and after distribution using the HPLC method described above. The distribution coefficient (D) was calculated using equation 2.6 (Florence & Attwood, 1998d; Wu et al., 2005): D= Co Cinitial − CE = Cw CE equation 2.6 Where Co is the concentration of NTX in n-octanol, Cw is the total concentration of NTX in PBS, and Cinitial and CE are the concentration of NTX in PBS initially and at equilibrium respectively. The true partition coefficient (P) was calculated from D using equation 2.7 (Schoenwald & Huang, 1983): P= (CB )o = D1 + 1 (CB )w anti log( pH − pKa) equation 2.7 Where (CB)O and (CB)W are the concentration of NTX free base in octanol and water respectively; pH at 7.4. No activity corrections were used as the concentration of the drug was significantly less than 0.01 M (Leo et al., 1971). To calculate the pKa value at 25 and 35 oC, equation 2.8 was used (Albert & Serjeant, 1971): − d ( pKa ) pKa − 0.9 = dT T equation 2.8 Where the left side of the equation is the temperature coefficient which is defined as the rate of decrease of pKa as the temperature is raised by 1o, pKa is the dissociation constant at T in degrees absolute (oK). 63 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.3.3. Statistical analysis A one-way analysis of variance (ANOVA) followed by Tukey’s pairwise comparisons at the 5% significance level was used to test statistical significance differences between the test solutions for forced degradation studies, distribution coefficient and partition coefficient at different temperatures. These were performed using GraphPad Software Version 3.05, San Diego California, USA. 64 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.4. Results and discussion 2.4.1. DSC study Figure 2.3 shows the DSC thermogram of NTX. A broad thermal event appeared at 60-110 oC. This could be due to surface moisture evaporation. A second endothermic peak was observed at 220oC. This was due to NTX melting. These results were in a good agreement with those reported elsewhere (Fuentes et al., 1997). According to the report, NTX showed endothermic peaks at 102 and 217 oC. These peaks were attributed to dehydration and melting of NTX respectively. This result is in contrary to that of the reference melting point of NTX which is 275oC (Galichet et al., 2006). The marked difference between the observed melting point and the reference one is attributed to crystallisation from two different solvents, resulting in two different crystal forms. The reference melting point was of the drug crystallised from methanol (O’Neil et al., 2006). However, the NTX powder purchased from Mallinckrodt Inc, St. Louis, MO, USA is crystallised from another solvent system (undisclosed by the company). Therefore, spectrometric studies were carried out to confirm the identity of the NTX powder used. Figure 2.3 DSC thermogram of NTX 65 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.4.2. Spectrometric analyses of NTX powder 2.4.2.1. Mass spectroscopy The mass spectra were determined on an LC-MS system, using atmospheric pressure chemical ionisation (APCI) to ionise the sample (Figure 2.4). Under these detection conditions, the sample ionises by interacting with the solvent (in this case methanol). The detector can detect positive ions (lower trace) and negative ions (upper trace) at the same time. To form positive ions the sample picks up a hydrogen ion: [M + H]+ so the mass is one atomic mass unit (AMU) more than the molecular weight (the lower trace, Figure 2.4) To form negative ions the sample loses a hydrogen ion: [M – H]- so the mass is one AMU less than the molecular weight (the upper trace, Figure 2.4). Figure 2.4 APCI mass spectra of NTX 2.4.2.2. FTIR-spectroscopy Figure 2.5 shows the FT-IR spectrum of NTX powder. The spectrum shows the characteristic peaks of NTX assigned at wave numbers 1717, 1660, 1504, 1314 and 1274 cm−1(Galichet et al., 2006). A high intensity peak at 1717 cm-1 is due to strong C=O of saturated cyclic aliphatic ketone stretching absorption (Silverstein et al., 2005). The C=C stretching band of C=C-O-H occurs at 1660 cm-1. The C=C aromatic ring stretching appears at 1504 cm-1, O-H bending absorption at 66 Chapter 2….Preformulation of Naltrexone Hydrochloride -1 1314 cm , and C-O-C asymmetric stretching peak observed at 1274 cm-1. A strong peak at 900 cm-1 is due to aromatic C-H bending vibration. Wavenumber (cm-1) Figure 2.5 FT-IR spectrum of NTX 2.4.2.3. 1H-NMR spectroscopy Figure 2.6 and Table 2.2 show the 1H-NMR spectrum of NTX and chemical shifts (δ) in ppm for the assigned hydrogens. The spectrum shows that all characteristic protons, which are chemically attached to characteristic functional groups of NTX such as alcoholic, phenolic and ammonium ion salt, were assigned. Conclusively, the data collected from the spectrometric analyses confirm the integrity of the chemical structure of NTX. 67 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.6 1H-NMR spectrum of NTX in CD3SO, 400 MHz 68 Chapter 2….Preformulation of Naltrexone Hydrochloride Table 2.2 Assignments of 1H-NMR spectra for NTX δ (ppm) 0.3 - 0.7 1.05 1.4 2 2.15 2.45 2.7 3 3.3 4 5 6.55 6.7 7 9 9.45 Assignment H19, H19’, H20, H20’ H18 H8’, H15’ H8 H7’ H16’ H15 H7, H10’, H16, H17’ H10, H17 H9 H5 H1 H2 H14 (OH Alcoholic) H (+NH) H3 (OH Phenolic) Integral 4H 1H 2H 1H 1H 1H 1H 4H 2H 1H 1H 1H 1H 1H 1H 1H 69 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.4.3. UV-spectrophotometric assay The normal absorption spectrum of NTX was carried out in PBS pH 7.4 and showed maximum absorption at 281.5 nm (λmax) (Figure 2.7). Under the experimental conditions described above, the graph obtained by plotting absorbance (A) at 281.5 nm against concentration of NTX, in the range of 30 to100 μg/ml, shows a linear relationship (Figure 2.8). Figure 2.7 UV-Spectrum of NTX (100 µg/ml) in PBS pH 7.4 70 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.8 Standard UV-spectrophotometric calibration curve of NTX in PBS at pH 7.4 (Results expressed as mean, n = 5) 2.4.4. HPLC method validation The system suitability parameters were found to be within acceptable limits. An analytical run time of 8 min was optimised for each sample. The retention time of NTX under the experimental conditions was 4.16 min and well resolved from the formulation excipients used (Figure 2.9 A and B). The following validation parameters were estimated and evaluated according to the ICH guidelines (ICH., 1996) (Shah et al., 1992): 2.4.4.1. Linearity The standard calibration curve was linear in the range of 5–50 µg/ml (R2 > 0.999). Percentage (%) recovery ± % relative standard deviation (RSD) ranged from 98.4 ± 3.84 to 103.53 ± 1.59 (Table 2.3). 71 Chapter 2….Preformulation of Naltrexone Hydrochloride Table 2.3 Reverse predicted concentrations, % recovery and regression coefficient (R2) Set number Nominal concentration (µg/ml) R2 Slope Intercept 5 10 15 20 25 30 35 50 1 5.34 9.68 14.50 19.91 25.44 30.95 34.13 50.00 0.9984 13.481 9.5682 2 4.95 10.11 14.96 19.43 25.15 30.26 35.48 49.65 0.9995 13.58 5.127 3 5.07 10.31 15.35 19.77 25.14 29.55 33.93 50.86 0.9984 13.210 3.0877 4 5.29 10.02 14.48 19.71 25.67 29.14 35.88 49.80 0.9983 13.324 7.0479 5 5.21 9.77 14.51 20.22 26.46 29.22 36.00 50.32 0.9975 13.280 7.6631 Mean 5.17 9.98 14.76 19.80 25.57 29.83 34.75 50.13 0.9996 13.375 6.4988 % RSD 1.59 4.39 3.84 0.73 1.60 3.40 3.18 0.81 % Recovery 103.53 99.80 98.40 99.00 102.00 99.43 99.28 100.26 2.4.4.2. Limit of detection (LOD) and limit of quantitation (LOQ) LOD and LOQ were estimated to be 0.55 µg/ml and 1.85µg/ml respectively. 2.4.4.3. Precision and accuracy The precision and accuracy of the method were reported as the relative standard deviation (%RSD) and % recovery of the QC samples (Table 2.4). The inter-day and intra-day precision for the three samples was found to be between 0.64 to 2.73%, and 0.14 to 1.16% respectively. These values comply with the acceptance criteria of the ICH guidelines (Shah et al., 1992). The purity of NTX peaks was checked all over the study using a PDA detector and Agilent ChemStation software through examining the UV spectra acquired of four predetermined points on the eluted drug peak from beginning to end by PDA detector. Figure 2.10 is a representative example showing four superimposable NTX spectra acquired during NTX elution. The examined points were 2 upslope points, a base point and a downslope point for purity determination. The UV spectra obtained across the peak were superimposable indicating a single component and a pure peak (Krull & Szulc, 1997; Wiberg et al., 2004). 72 Chapter 2….Preformulation of Naltrexone Hydrochloride Table 2.4 Precision and accuracy data of the QC samples (Results were expressed as mean values, n = 5) Mean % RSD % Recovery 8 7.92 1.16 99.00 Nominal concentration (µg/ml) Intra-day Inter-day 18 40 8 18 40 18.20 38.36 7.95 18.60 40.10 0.14 0.71 2.73 1.40 0.64 101.00 96.00 99.40 103.00 100.24 Figure 2.9 Representative HPLC chromatograms of standard NTX (0.4 mg/ml) solution (A) and NTX resolved from Span 60: cholesterol: DCP-niosomal formulation (B) 73 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.10 PDA- spectra of the NTX peak acquired from 4 different positions throughout the eluted peak for peak purity assessment 2.4.5. Forced degradation studies Forced degradation studies were performed in order to demonstrate the specificity of the developed method, to ensure that the method is stability-indicating and to provide insights on the stability of the drug under common stress conditions (Alsante et al., 2003; Reynolds et al., 2002). The results of % recovery of NTX from the various stress conditions used after 24 h exposure are shown in Figure 2.11. The chromatograms of NTX resolved after exposure to base-heat, acidheat, heat, oxidation and artificial daylight illumination at 40oC for 24 h are shown also in Figure 2.12 and Figures 2.14 -2.17 respectively. 74 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.11 Percentage (%) remaining of NTX after 24 h exposure to various stress conditions (Results are expressed as mean values ± SD, n = 3) Approximately 50% of the potency of NTX was lost after exposure to strong alkaline conditions for 24 h at 60oC and discolouration of the solution to yellowish brown. Figure 2.12 and Figure 2.13 show the HPLC chromatograms and the different spectra of both NTX and the main degradation products of NTX after exposure to a strong base and heat. Figure 2.12 shows the chromatogram of NTX after base-catalysed degradation at 60oC for 24 h. The chromatogram shows three peaks at 3.13, 3.39 and 3.8 min in addition to the parent drug at 4.15 min. The peak resolved at 3.13 min was due to medium or salt-related impurities, as seen in the blank chromatogram (Figure 2.12), whereas the peaks at 3.39 and 3.8 min were due to NTX degradation. The chromatogram shows a resolution factor of 1.8 for NTX and its degradation product at 3.8 min. Although the resolution of peaks of different heights should be ≥ 2 (Krull & Szulc, 1997), this can be accepted because the smallest peak was eluted first (Gleditsch & Waaler, 2001). The similarity of the parent compound and its degradation products makes them difficult to separate (Gleditsch & Waaler, 2001; Krull & Szulc, 1997). For example, a resolution factor of morphine and morphine-N-oxide is found to be 1.6. This is due to the similarity of hydrophilicity of morphine and morphine-N-oxide (Gleditsch & Waaler, 2001). 75 Chapter 2….Preformulation of Naltrexone Hydrochloride Exposure of NTX to a strong acid and heat showed moderate degradation (Figure 2.11). The extent of NTX degradation at pH 1.2 was significantly less (P < 0.01) than that at pH 10. Also, the extent of degradation (16%) at pH 1.2 was significantly lower (P < 0.01) than that (23.5%) at pH 7.4. These results demonstrate that the chemical stability of NTX is pH-dependent. Assuming that the main degradation reaction is oxidation, it is generally more difficult to remove an electron from a drug when it is more positively charged. Therefore, drug stability against oxidation is often greater under lower pH conditions. An acidic pH promotes protonation of the basic drug. On the other hand, higher pH conditions deprotonate the basic drug. The basic conditions make the basic drug more susceptible to oxidation than the acidic conditions (Waterman et al., 2002). For example, NTX is completely ionised in pH 1.2 whereas NTX, with a pKa value of 8.13, is predominantly unionised at pH 10. This explains why the degradation of NTX decreased in the following pH order pH 10 > pH 7.4 > pH 1.2. Furthermore, NTX underwent extensive degradation when subjected to the oxidising agent (H2O2) (Figure 2.16). The percentage of NTX remaining was found to be 16%. The major degradation product was eluted at 4.3 min (Figure 2.16), whereas it was eluted at 3.8 min in the heat and base stress test. These findings suggest that the product of oxidation due to H2O2 is different from that produced in the heat-base degradation. H2O2 is often non-predictive of molecular oxygen reactions because it does not involve the radical chain reaction common with the oxygen-based processes (Waterman et al., 2002). However, H2O2 is a common oxidising agent and is recommended by ICH to induce drug oxidation (Alsante et al., 2003; Reynolds et al., 2002). The order of the degradation of NTX under the various stress conditions is as follows: oxidation > base > heat > photolysis > acid. 76 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.12 HPLC chromatograms of the blank solution and aquous NTX solution (0.4 mg/ml) pH 10 after 24 h exposure to heat at 60oC 77 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.13 PDA spectra of NTX eluted at 4.16 min (A), the degradation product resolved at 3.7 min (B) and the second degradation product resolved at 3.3 min (C) after exposure to pH 10 at 60 oC 78 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.14 HPLC chromatograms of the blank solution and aquous NTX solution (0.4 mg/ml) of pH 1.2 after 24 h exposure to heat at 60oC Figure 2.15 HPLC chromatograms of the blank solution and aquous NTX solution (0.4 mg/ml) of PBS pH 7.4 after 24 h exposure to heat at 60oC 79 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.16 HPLC chromatograms of the blank solution and aquous NTX solution (0.4 mg/ml) of PBS pH 7.4 after 24 h exposure to H2O2 (2% v/v) Figure 2.17 HPLC chromatograms of the blank and aquous NTX solution (0.4 mg/ml) of PBS pH 7.4 after 24 h exposure to artificial daylight illumination (10,000 lux) at 40oC 80 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.4.6. Degradation products identification using ESI-MS NTX degradation was indicated by appearance of additional peaks at 3.39 min and 3.8 min as well as discolouration of NTX solutions. These peaks increased in area as a function of temperature and time and the intensity of the discolouration obviously increased, as well. Heat and base-catalysed degradation solutions were introduced to ESI-MS. The mass spectra were generated to determine the molecular weight of the major degradation products (Figure 2.18). The results showed that major ion peaks appeared at 342, 358 and 374 m/z. These peaks were attributed to the positively charged molecular ion of naltrexone [M+1]; and other metabolites (degradation products) at [M+16] and [M+32]. The metabolite at [M+16] was called product I and attributed to 10-hydroxy naltrexone (Figures 2.18 and 2.19). According to the United States pharmacopeia (USP), 10-hydroxy naltrexone is eluted at around 0.3 min earlier than the naltrexone peak (USP32/NF27, 2009). The metabolite at [M+32] was attributed to N-oxide-10hydroxy naltrexone, as nitrogen has a lone pair of electrons. This makes it a nucleophilic centre for oxidative reaction. An N-oxide product is a common oxidation product for morphinan-bearing compounds (Gleditsch & Waaler, 2001; Yeh & Lach, 1961). Figure 2.19 shows the proposed degradation pathway of NTX. The N-oxide position in product II was evidence based on a published report on a similar compound where N-oxide morphine is the oxidation product of morphine (Yeh & Lach, 1961). Other m/z ratio peaks at 364, 380 and 396 were observed and related to the sodium and potassium adducts of naltrexone and sodium adduct of sodium 10hydroxy-naltrexone respectively. Figure 2.18 ESI mass spectra of product I and Product II of NTX oxidation 81 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.19 Schematic of the proposed oxidative degradation pathway for NTX 82 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.4.7. pH-degradation rate of NTX Figure 2.20 shows semi-log degradation plots of NTX in phosphate buffers over pH 4.5-9.5 at 60oC. The semi-logarithmic plots of % NTX remaining versus time showed good linearity; and the correlation coefficient (R2) ranged from 0.98 to 0.995. This indicates that the degradation of NTX followed first-order kinetics. A pH-rate profile of a sigmoid or S- shape was obtained over a pH range of 4.5 to 9.5 at 60oC (Figure 2.21). The S-shaped pH-rate profile is most likely to be associated with autoxidation reactions (Johnson & Gu, 1988). Additionally, a yellowish-brown colour was observed with the degraded samples and the intensity of colour increased as the percentage of degradation increased. Discolouration of the NTX solution is another symptom which supports the autoxidation process (Johnson & Gu, 1988). This indicates that autoxidation reactions are involved in the decomposition of NTX. The degradation occurred more rapidly in basic than acidic solutions (Figure 2.21). For example, the observed rate constants (kobs) were 0.02 and 0.172 day-1 at pH 4.5 and 8.5 respectively; and the estimated half-life (t1/2) values were found to be 35 and 4 days at pH 4.5 and 8.5 respectively at 60oC. These results are attributed to the pH-dependency of oxidation reactions in aqueous solutions. Figure 2.20 First-order degradation plots for NTX at 60oC in phosphate buffers of various pH values (Results are expressed as mean values ± SD, n=3) 83 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.21 pH-rate (kobs, day-1) profile for NTX at 60oC The removal of electrons from a positively charged compound is more difficult than a neutral one. This causes an increase of the oxidation potential with a concurrent increase in stability against oxidation (Lachman et al., 1986). A basic compound undergoes accelerated decomposition in alkaline conditions where the drug is dominantly unionised. Another factor which can enhance the oxidation reaction is the hydroxyl ion concentration through a catalytic effect (Florence & Attwood, 1998a; Johnson & Gu, 1988; Waterman et al., 2002). For instance, the autoxidation rate for the fully protonated form of NTX at pH 4.5 is approximately 17 times slower than that for the fully unprotonated form of NTX at pH 9.5 at 60oC. Two steeps were observed at pH ranges of 6.5-7.5 and 8.5 and 9.5. The state of NTX ionisation apparently influences the oxidation reaction rate. Below pH 6.5, NTX molecules are dominantly in an ionised state, hence the reaction rate proceeds at a low rate. The steep at pH 6.5-7.5 is likely due to availability of 50% unionised NTX species at pH 7.5, which is equal to pKa of NTX, at 60oC. The second steep at pH higher than 8.5, NTX molecules are dominantly in an unionised state. Further, NTX is degraded by two oxidation reactions (Figure 2.19); these could explain the two steeps in the pH-oxidation profile of NTX. 84 Chapter 2….Preformulation of Naltrexone Hydrochloride The pH at which NTX shows a maximum stability (pH 4.5), unfortunately, is not suitable for ocular administration. This is because it is well below the physiological pH of the resident tears (pH 7.4) leading to possible precorneal tear film breakdown, reflex lacrimation and blinking. Consequently, the administered dose is rapidly diluted or removed from the surface of the eye (Lang et al., 2002). Additionally, NTX is dominant in the ionised form (the least absorbable form) at the acidic conditions. NTX penetration via the lipophilic epithelial corneal membrane is expected to be minimal at the pH of maximal stability. Therefore, ophthalmic solution should be ideally buffered to a pH of 7.4 which is the normal physiological pH of the tear fluid. The reason for this is that the product would be comfortable and offer optimum therapeutic activity (Lang et al., 2002). 2.4.8. Accelerated stability and chemical kinetic studies The effect of temperature on NTX stability was studied in PBS at pH 7.4. NTX was subjected to forced degradation at different temperatures 40, 50, 60 and 70oC. Semi-log plots of the % NTX remaining versus time showed linear relationships indicating first order kinetics (Figure 2.22). Table 2.5 shows the calculated and predicted kobs at different temperatures. The results show the higher the temperature, the greater the value of kobs (Table 2.5). There was also an inverse relationship between the temperature and t1/2 and t90% (time of 10% degradation). The kobs values at the assigned temperature were plotted against the reciprocal of the absolute temperature according to the Arrhenius equation. The activation energy (Ea) was estimated to be 15.28 kcal/mol. The Ea value reported for morphine sulphate (a morphinone-based drug) is 23 kcal/mol (Kennon, 1964). From the literature, the Ea values of degradation reactions for most of the drugs in aqueous solutions are found to be in the range of 10-30 kcal/mol (Lachman et al., 1986). The stability of NTX in ambient and cold storage conditions was predicted by calculating kobs values at 25 and 4oC respectively (Table 2.5). These findings show that NTX is less chemically stable than morphine sulphate. Hence, encapsulating NTX in bilayer membranes in lipid-based vesicles could enhance its chemical stability, because this reduces the direct interaction of NTX with light and other initiators of oxidative catalysis. The photo-degradability of tretinoin encapsulated by niosomes was lower than the free tretinoin solution (Manconi et al., 2003). Therefore, this is an incentive to study the effect of encapsulating NTX in niosomes on the chemical stability. 85 Chapter 2….Preformulation of Naltrexone Hydrochloride Figure 2.22 First-order degradation plots for NTX in PBS pH 7.4 at various temperatures (Results are expressed as mean values ± SD, n=3) Table 2.5 First–order kinetic parameters and correlation coefficient of NTX degradation in aqueous solution of PBS pH 7.4 at different temperatures Temperature (oC) 4 25 40 50 60 70 R2 0.97 0.99 0.96 0.97 kobs (day-1) 0.001* 0.008* 0.02 0.086 0.14 0.188 t1/2 (days) 693 87.00 35.00 8.00 5.00 3.50 t90% (days) 104.00 13.00 5.20 1.21 0.74 0.55 * Average kobs values at 4oC and 25oC were predicted using the Arrhenius equation. The other Kobs values were experimentally determined. 86 Chapter 2….Preformulation of Naltrexone Hydrochloride 2.4.9. Solubility studies of NTX Thermal analysis using DSC showed that the NTX powder used had a melting peak at 220oC suggesting that it is a different crystalline form for NTX. The object of this study was to determine the aqueous solubility of the available NTX crystalline form. The results showed that the respective solubility of NTX in water and PBS pH 7.4 was > 100 mg/ml and 60 mg/ml. These results indicated that the available NTX form is freely soluble in water, and there is no noticeable difference between the solubility of the NTX form used and the solubility of the reference NTX. The lower solubility of NTX in PBS pH 7.4 compared to that in water is attributed to the lower ionisation (PKa of NTX is 8.26 at 25oC) at pH 7.4 than water (pH 6) where ionisation percentage of NTX in pH 7.4 and 6 is 87.87% and 99.45% respectively. Additionally, the ionic strength of the buffer system might also contribute to lowering NTX solubility compared with pure water. This negative effect of ionic strength on NTX solubility is due to the salting-out effect (Miyazaki et al., 1981). Furthermore, the availability of chloride ions in the PBS might decrease the solubility of the hydrochloride salt due to the common ion effect on the solubility product (Florence & Attwood, 1998d). 2.4.10. Determination of n-octanol/PBS distribution coefficient (D) and partition coefficient (P) of NTX For ionisable drugs, the distribution coefficient (D) is the ratio between drug concentration (unionised species) in oil phase and drug concentration in the aqueous phase (both free base and acid salt). The true partition coefficient (P) for ionisable drugs is the ratio between concentration of unionised species in the oil phase (octanol) and concentration of unionised species in the aqueous phase. The aqueous phase used was a PBS solution with pH 7.4 and osmolality of 278 ± 5 mOSm/kg to mimic the pH and osmolality of the tear fluid (Lang et al., 2002). The D values were determined at three different temperatures 25oC (ambient temperature), 35oC (ocular surface temperature) (Schoenwald & Huang, 1983) and 37oC (body temperature). The D values at the three different temperatures were converted through equation 2.4 to The P values to determine lipohilicity of NTX. The reported NTX pKa values are 8.38 and 8.13 at 20oC and 37oC respectively (Kaufman et al., 1975). Nitrogenous bases are highly temperature sensitive where they become weaker as the temperature is raised. The pKa of NTX decreases when the temperature increases. Equation 2.11 governs the temperature variation for the ionisation of monoacidic bases. This equation has been well supported by experimental evidence (Albert & Serjeant, 1971). The average temperature coefficient was calculated using equation 2.11 and was 87 Chapter 2….Preformulation of Naltrexone Hydrochloride estimated to be 0.024. This value was multiplied by 5 and the product was subtracted from the pKa value at 20oC to give the value of pKa at 25oC (Table 2.6). To calculate pKa at 35oC, the average temperature coefficient was multiplied by 2 and the product was added to pKa at 37oC (Table 2.6). Table 2.6 n-octanol-buffer distribution coefficient (D) and partition coefficient (P) and calculated pKa at different temperatures (Results are expressed as mean values ± SD, n =3) Temperature (oC) 25 35 37 D 4.28 ± 0.25 6.05 ± 0.24 9.10 ± 0.07 pKa 8.26 8.18 8.13 P 35.30 ± 2.00 42.50 ± 1.70 58.00 ± 0.51 Log P 1.54 ± 0.02 1.61 ± 0.001 1.76 ± 0.002 In addition, Table 2.6 presents D and P values at the tested temperatures. The D values at 25, 35 and 37 oC were converted to P values to judge lipophilicity of NTX. The results showed that both D and P values increased significantly (P < 0.001) as a function of the studied temperatures. For example, 1.4- and 2-fold increases in the D values were obtained as the temperature is raised from 25oC to 35oC and 37oC respectively. These results demonstrate the dependency of the D value on the temperature. The temperature dependency of D value has been reported with most of opioid agonists and antagonists (Kaufman et al., 1975). This is partly attributed to decreasing in the ionisation constant as the temperature increases. Hence, the amount of unionized species of NTX increases and diffuses to the oil phase, and the concentration of unionised species increases in the octanol phase. The P value of NTX was found to be 42.5 at 35oC. Based on β-adrenergic blockers classification on lipophilicity, NTX can be considered as a hydrophilic drug and hence, the corneal epithelium is expected to be the rate-limiting barrier for ocular absoprtion (Schoenwald & Huang, 1983). This is an additional incentive to consider niosomes (surfactant/lipid-based system) for the ocular delivery of NTX. 88 Chapter 2….Preformulation of Naltrexone Hydrochloride Conclusion • The available crystal form of NTX is freely soluble in water and soluble in PBS. • NTX undergoes chemical decomposition upon melting. • The validated HPLC method was rapid, precise, sensitive and stability indicating. • NTX is vulnerable to autoxidation as demonstrated by S-shaped pH-degradation rate constant profile, discolouration of NTX solutions and ESI-mass spectroscopy. • NTX shows temperature-dependent lipid solubility as shown from the distribution coefficient/partition coefficients at different temperatures. • According to P values, NTX is a hydrophilic agent and its permeation through the cornea is likely to be dependent on the epithelium barrier. From the preformulation study, there are incentives to consider surfactant-based vesicles (niosomes) as a delivery system for NTX in order to enhance NTX uptake by the lipophilic corneal epithelium and provide NTX with a shield of bilayer membranes to protect it from oxidation. Therefore, the next chapter is concerned with preparing and characterising NTX niosomes. 89 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3. Preparation of niosomes using biocompatible surfactants: optimisation of surfactants and cholesterol level 3.1. Introduction 3.1.1. Niosomes Niosomes are non-ionic surfactant based vesicles. They were originally developed as an alternative controlled drug delivery system to liposomes, in order to overcome the problems associated with sterilisation, large-scale production and stability (Azmin et al., 1985, 1986; El Maghraby & Williams, 2009). The hydration of a film, comprising a mixture of a single or double-alkyl chain and cholesterol, leads to the formation of vesicular dispersion. These dispersions were termed niosomes (Baillie et al., 1985). Basically, these vesicles do not form spontaneously. Thermodynamically stable vesicles form only in the presence of proper mixtures of surfactants and a membrane stabilising agent (cholesterol), at a temperature above the gel/liquid transition of the main lipid forming vesicles (Azmin et al., 1985, 1986; Nefise, 2007). The first niosome formulations were developed and patented by L’Oreal in 1975 (Nefise, 2007). Niosomes were first utilised as a drug delivery system for anticancer drugs (Azmin et al., 1985, 1986). The developed niosomes were capable of altering the pharmacokinetic profile, organ distribution and metabolism of methotrexate in mice (Azmin et al., 1985, 1986). Niosomes are versatile in structures, morphology and sizes; they can entrap hydrophilic drugs in aqueous compartments or lipophilic drugs by partitioning of these molecules into bilayer domains (Figure 3.1). 90 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.1 Locations of entrapped hydrophilic, lipophilic and amphipathic drugs in niosomes Furthermore, they can be formulated to be unilamellar, oligolamellar or multilamellar (Figure 3.2). Niosomes also possess great stability, are cost-effective, and provide a simple methodology for routine and large-scale production (Baillie et al., 1985; Uchegbu & Florence, 1995; Uchegbu & Vyas, 1998). Figure 3.2 Major types of niosomes, MLV (multilamellar vesicles), OLV (oligolamellar vesicles) and ULV (unilamellar vesicles). Small circles (o) represent polar head group, sticks (–) represents apolar tails of single-chain surfactant molecules and a bilayer membrane represent a circular double surfactant molecules layer oriented in continuous tails-to-tails and polar heads lining the inner and outer circle 91 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3.1.1.1. Self-assembly of lipid molecules The principle of opposing forces Formerly, the self-assembly of amphiphilic molecules, composed of hydrophilic and lipophilic portions, in an aqueous environment into well-defined structures, such as micelles, open bilayers and vesicles, was solely ascribed to favourable hydrophobic and hydrophilic interactions in the aggregated phase (Tanford, 1973). The major forces derive from the hydrophobic interactions between the hydrocarbon tails, which induces the molecules to associate; and the hydrophilic nature of the head group, which imposes the opposite requirement, that they remain in contact with water (the theory of opposing forces) (Tanford, 1973). These two interactions compete giving rise to the idea of two ‘opposing forces’ acting mainly in the interfacial region: one tends to decrease and the other tends to increase the interfacial area per molecule (the head group area) exposed to the aqueous phase. While this notion is true and can be accepted thermodynamically, it does not explain all the possible structures into which lipids can aggregate. For instance, some form micelles (e.g. lysolecithin), and some form bilayers (e.g. lecithin), while others form only inverted micellar structures or precipitate out of solution (Israelachvili et al., 1980). The principle of geometrical or packing properties The role of molecular geometry or the ‘packing’ properties of amphiphiles (surfactants), as shown in Figure 3.3, in determining these structures was recognised by (Israelachvili et al., 1976). Accordingly, an optimal polar head group area ao, an optimal volume of the hydrocarbon chains v and a maximum effective length that the hydrocarbon chains can assume (critical chain length) lc will determine the self-assembly of the amphiphiles into organised lipid aggregates. The value of the dimensionless ‘packing parameter’ v/ ao lc will determine whether a certain amphiphile or lipid structure will form spherical micelles (v/ ao lc < 1/3); non-spherical micelles (1/3 < v/ ao lc < 1/2); bilayers (1/2 < v/ ao lc <1); an inverted micellar structure (v/ ao lc >1); or even precipitate out of solution (e.g. cholesterol), since cholesterol molecules possess a headgroup too small to pack into organised structures (Israelachvili et al., 1980; Israelachvili & Mitchell, 1975; Israelachvili et al., 1976). Apart from the selection of amphiphiles with proper geometrical properties, the formation of niosomes involves some input of energy (physical agitation and heat), as closed vesicles do not form spontaneously (Baillie et al., 1985; Sahin, 2006). 92 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.3 Schematic of a single-chain surfactant, ao = hydrophilic head group area, v = hydrophobic chain volume, lc = hydrophobic chain length In pharmaceutical research, the hydrophilic-lipophilic balance (HLB) system, which is a measure of the relative contributions of the hydrophilic and lipophilic regions of the surfactant molecules, is more commonly used as an indicator to niosomes formation. Surfactants possess HLB values of 4 to 8 have been found to be optimal for the formation of niosomes (Girigoswami et al., 2006). Surfactants of high HLB values (9.6-16.7) such as polysorbates (Tween®) tend to form micelles or open lamellar structures (Manosroi et al., 2003). Other surfactants cannot assemble into niosomes due to their geometry (packing) properties, such as Span 80 (HLB of 4.3). This is because of the kink in the hydrocarbon chain due to double bond at C9 (Yoshioka et al., 1994). 3.1.1.2. Raw Materials used in manufacturing niosomes Niosome forming lipids (Surfactants) Unlike naturally occurring phospholipids (the principal liposome-forming lipids) which have double alkyl chains, most widely used niosome forming surfactants are single chain, synthetic, non-ionic surfactants (Florence & Baillie, 1989; Uchegbu & Florence, 1995). Alkyl ethers, alkyl esters and alkyl amides, as well as fatty acid and amino acid compounds, can form vesicles (Uchegbu & Florence, 1995). Recently, niosomes consist of Pluronic L64 and P105 (copolymers of ethylene oxide and propylene oxides) have been prepared and used as a transdermal delivery system for sulfadiazine sodium (a model hydrophilic drug) (Muzzalupo et al., 2011). However, both alkyl ethers and alkyl esters have a wider application as niosomes drug delivery carriers, due to their commercial availability and low toxicity. 93 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Alkyl ethers Alkyl ethers can be divided broadly into two subclasses, based on their hydrophilic head group: alkyl ethers whose heads consist of glycerol subunits or larger sugar molecules, and alkyl ethers whose hydrophilic head groups consist of repeat ethylene oxide subunits (Uchegbu & Florence, 1995). Alkyl glycerol ethers, most notably hexadecyl diglycerol ether (C16G2), have been reported as niosome-forming surfactants and have been used in many drug delivery applications such as experimental cancer chemotherapy (Baillie et al., 1985; Uchegbu & Florence, 1995). These surfactants were used to alter the pharmacokinetics of methotrexate (Azmin et al., 1985) and doxorubicin (Uchegbu et al., 1996a). Alkyl glycerol ethers were also used to encapsulate sodium stibogluconate to control leishmaniasis (Baillie et al., 1986). The second group of alkyl ether surfactnats, in which the hydrophilic region consists of repeat oxyethylene units, was used to encapsulate insulin for oral delivery, in order to prevent the inactivation of insulin by the gastric juice (Pardakhty et al., 2007). Alkyl esters Alkyl esters, such as sorbitan fatty acid esters (Span®) and polyoxyethylene sorbitan fatty acid esters (Tween®), are widely used in cosmetics and foodstuffs, as well as oral, parenteral and topical pharmaceutical formulations. They are regarded as non-toxic and non-irritant materials (Lawrence, 2003). These surfactants have been utilised in manufacturing niosomes as a drug delivery system, including oral (Varshosaz et al., 2003), trandermal (Balakrishnana et al., 2009; Ibrahim et al., 2005) and ocular delivery (Abdelbary & El-gendy, 2008; Aggarwal et al., 2004; Kaur et al., 2008). Cell culture toxicity studies have shown that niosomes composed of ester-type surfactants are less toxic than ether-type niosomes. This could be attributed to the possible enzymatic degradation of ester bonds (Hofland et al., 1992). Cholesterol (bilayer membrane stabiliser) Some water-soluble surfactants, such as Tween 20 and Tween 61, could not form vesicles under normal conditions; these surfactants prefer to form micelles, in a manner identical to lyolecithin (Florence & Baillie, 1989; Manosroi et al., 2003). These surfactants might form vesicles when mixed with cholesterol. Cholesterol molecules operate cooperatively to form bilayer vesicles. Cholesterol is considered as lipid, but it is a comparatively rigid molecule and lacks the accommodating ability of an acyl-chained lipid. In particular, cholesterol does not form bilayers and its shape can be modelled as an inverted cone (Figure 3.4) (Israelachvili et al., 1980). 94 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties However, cholesterol has been called the ‘mortar’ of bilayers because, by virtue of its molecular shape and solubility properties, it fills in empty spaces among the amphiphiles, thereby anchoring them more strongly into the bilayer structure. Figure 3.4 Chemical structure of cholesterol Cholesterol in niosomes is reported to increase the membrane stability, decrease the fluidity of the membrane and alter membrane permeability (Uchegbu & Florence, 1995; Uchegbu & Vyas, 1998). Other excipients, such as dicetyl phosphate and stearyl amine, were used in niosomes technology to impart negative and positive charges respectively (Yoshioka et al., 1994). In such cases, the operational packing parameter is estimated for all the components. 3.1.1.3. Vesicle preparation Vesicle formation requires an input of energy, in the form of either gentle agitation or probe sonication. Usually, the surfactant/lipid mixtures are hydrated in the presence of the drug in an aqueous solution at a temperature above the gel/liquid transition temperature of the surfactant. The unentrapped drug can be separated by ultracentrifugation, low-pressure gel filtration chromatography or exhaustive dialysis. Since vesicle size is an important determinant of drug biodistribution, it may be reduced by extrusion via nucleopore filters, microfluidisation or highpressure homogenisation. The different techniques used in surfactant/lipid hydration and drug loading are summarised as follows. Hydration and drug loading techniques Below are the commonly used laboratory methods of niosome preparation and drug loading identified in the literature: 95 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 1. The injection of an organic solution of surfactant/lipid into a hot aqueous solution of the drug to be encapsulated, which is heated above the boiling point of the organic solvent (ether injection) (Baillie et al., 1985). 2. The formation of a surfactant/lipid film by evaporation of an organic solution of surfactant/lipid. This lipid film is then hydrated with a solution of the drug (Azmin et al., 1985; Baillie et al., 1985). This method is commonly known as thin film hydration (TFH) or the conventional method. The niosomes generated by THF are multilamellar vesicles (MLV). The TFH method was previously described by (Bangham et al., 1965) for the preparation of liposomes. 3. The formation of a W/O emulsion from an organic solution of surfactant/lipid and an aqueous solution of the drug. The organic solvent is then slowly evaporated to leave niosomes dispersed in the aqueous phase. In some cases, a gel results which must be further hydrated to yield niosomes [reverse phase evaporation (REV)] (Kiwada et al., 1985) (Szoka & Papahadjopoulos, 1978). The vesicles generated by the REV method are called unilamellar or oligolamellar vesicles and are relatively more uniform in size compared with the TFH method (Szoka & Papahadjopoulos, 1978). 4. The injection of melted surfactant/lipid into a highly agitated heated aqueous phase in which presumably the drug is dissolved (Wallach & Philippot, 1993) or the addition of a warmed aqueous phase dissolving the drug to a mixture of melted lipids and hydrophobic drug (Niemac et al., 1995). This method does not require the use of organic solvents, which are expensive, hazardous and difficult to completely remove. 5. The homogenisation of a surfactant/lipid mixture, followed by bubbling of nitrogen gas through this mixture (Talsma et al., 1994). The bubbling unit consisted of a round bottom flask with three necks, positioned in a water bath to control temperature. A water-cooled reflux column was connected to the central neck of the flask to avoid loss of water during the experiment. A thermometer for temperature monitoring was positioned in the second neck. The third opening was used for the tubing. A continuous stream of gas bubbles was generated at the bottom of the flask by forcing nitrogen gas through a sintered glass filter with a surface area of 1 cm2 by a slight over pressure. Prior to the gas bubbling step, the lipid materials are added to the aqueous solution and then mixed for 30 min with a homogenizer. The prepared vesicle dispersion was in the size range of 0.2 µm to 0.5 µm. 96 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Omitting the homogenisation step has been found not to affect the particle size; however, a longer ‘bubbling time’ was required. 6. The supercritical carbon dioxide (scCO2) technique has been used as a more environmentally friendly alternative. The carbon dioxide gas can be in a fluid state (scCO2) under a temperature of 31.1oC and pressure of 78.8 bars. The scCO2 has good solvating properties (Manosroi et al., 2008a). Niosomes composed of different molar ratios of Tween 61: cholesterol mixtures were successfully prepared using the scCO2 method and encapsulated glucose (a model water soluble agent). The scCO2 niosomes demonstrated higher drug entrapment efficiency than the niosomes prepared using the TFH method. However, The TFH niosomes showed superior stability and dispersibility compared with the scCO2 niosomes (Manosroi et al., 2008a). Niosome size The particle size of niosomes can determine their behaviour in the body (Uchegbu & Vyas, 1998). It has been reported that particles smaller than 100 nm easily penetrate tumour tissues (Mayer et al., 1989). This is because these nano-particles can penetrate the discontinuous capillaries and fenestra of tumour tissues (Nakano et al., 2008). A size reduction step is occasionally incorporated into the niosome production procedure, subsequent to the initial hydration step. This is in order to improve the biodistribution of systemically injected niosomes and to target the administered niosomes at the affected area (Azmin et al., 1986; Uchegbu et al., 1995; Uchegbu & Vyas, 1998). Niosome sizes > 10 µm are suitable for ophthalmic drug delivery (Sahin, 2006). This allows them to encapsulate a considerable amount of therapeutic drug molecules, resist removal by lacrimation and prolong ocular residence time. Micro-size niosomes are not easily washed out by lacrimal fluid compared with nano-size niosomes (Uchegbu & Vyas, 1998). For instance, large niosomes (discomes) sized 12 - 60 µm were prepared and used for ocular delivery of timolol maleate (water soluble drug). These were found to entrap a relatively larger amount of timolol maleate (Vyas et al., 1998). The prepared discomes produced a sustained drug effect. They were found to lower significantly intraocular pressure (IOP), compared to an aqueous solution. The systemic drug absorption measured from discomes was of a negligible level (Vyas et al., 1998). Figure 3.5 demonstrates a schematic of suitable niosome sizes for particular routes of administration as adopted from (Uchegbu & Vyas, 1998). 97 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.5 Suitable niosome sizes for particular routes of administration Generally, size reduction (sonication or extrusion methods) adversely affects the entrapment efficiency of the vesicles to the active agents. This could be ascribed to their small inner volume and drug leakage from small and intermediate-sized unilamellar vesicles (up to 100 nm) (Uchegbu & Florence, 1995; Uchegbu & Vyas, 1998; Zhigaltsev et al., 2001). A reduction in vesicle size may be achieved by a number of methods: • Probe sonication can produce niosomal vesicle sizes in the range of 100 to 140 nm (Azmin et al., 1985; Baillie et al., 1985). Although it is more effective than bath sonication, probe sonication suffers from high energy input to the solution, leading to a sharp local increase in the temperature of the solution and shedding of titanium particles into the solution. This necessitates an additional centrifugation step to remove the metallic particles from the niosomal dispersion. • Extrusion through polycarbonate (Nucleopore™) filters using liposome extruders. • A combination of sonication and filtration has been utilised to prepare doxorubicin niosomes in the 200 nm size range (Uchegbu et al., 1995). • High pressure homogenisation also yields vesicles below 100 nm in diameter, although drug loading is ultimately sacrificed to achieve this small size (Uchegbu & Vyas, 1998). 98 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3.1.2. Non-conventional niosomes 3.1.2.1. Elastic niosomes Over the past three decades, researchers in the field of liposomes have introduced a new class of liposomes. These have significant elasticity and ultradeformability, compared with conventional liposomes. These elastic or ultradeformable liposomes, Transferosomes®, demonstrated the ability to penetrate intact skin while carrying therapeutic concentrations of drugs (Cevc & Blume, 1992). They consist of phospholipids and an edge activator; an edge activator is often a single chain surfactant that can destabilise the lipid bilayers of the vesicles, and increase the deformability of the bilayers (Cevc et al., 1996). Parallel to this development, other classes of vesicles were produced; vesicles with ethanol (ethosomes) to provide flexibility were developed (Touitou et al., 1997). Ethosomes embodying a high concentration (20–45%) of ethanol demonstrated enhanced skin delivery and were reported to penetrate intact skin layers (Touitou et al., 2000). For more detail about these two delivery systems, the reader can refer to references (El Maghraby & Williams, 2009; Elsayed et al., 2007). When it comes to the literature on niosomes, there is scarcely any detailing on non-conventional forms of niosomes. To our knowledge, only one study has reported on elastic niosomes (Manosroi et al., 2008b). Tween 61 and Span 60 niosomes entrapping diclofenac diethyammonium and embodying 0-25% ethanol have been developed. The prepared niosomes were called elastic niosomes. This is because the prepared niosomes registered deformability index values of 13.76 and 3.44 times higher than conventional niosomes loaded and non-loaded with the drug respectively (Manosroi et al., 2008b). 3.1.2.2. Proniosomes One approach to minimise the water content in liposomes and niosomes dispersions is to develop a vesicle proconcentrate, called proliposomes or proniosomes. These contain all the vesicular components with ethanol and only traces of water. These systems are gel-like preparations that are believed to generate vesicles used for application to the skin (El Maghraby & Williams, 2009). Proniosomes are mainly composed of non-ionic surfactants, phospholipids, alcohols and water, with or without cholesterol (Mokhtar et al., 2008). The effects of formulation composition, surfactant and alcohol type, and preparation method on the transdermal permeation profile of various proniosomes have been studied (Fang et al., 2001; 99 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Vora et al., 1998). The effects of non-ionic surfactants on the levonorgestrel permeation profile showed that the flux value was the highest for Span 80 and lowest for Span 60. No significant difference was observed in the skin permeation profile of formulations containing Span 40 and Span 60, due to their higher phase transition temperature, which is responsible for their lower permeability. Conversely, opposite effects in the case of estradiol were reported (Fang et al., 2001). Span 60 showed the highest flux on the estradiol permeation profile. Proniosomes from estradiol and levonorgestrel formulations differed in their content of cholesterol, and type and content of alcohol. The effect of different alcohols on the drug permeation profile was found to be as follows: isopropanol > butanol > propanol > ethanol. Proniosomes could become a useful dosage form for transdermal drug delivery. This is due to their simple and scaling-up manufacturing and the ability of proniosomes to modulate drug delivery across the skin (El Maghraby & Williams, 2009). 3.1.3. Niosomes as drug delivery systems Over the past three decades, niosomes have been successfully utilised as a drug delivery system to solve some major biopharmaceutical problems. Poor bioavailability of slightly soluble drugs can be improved, as niosomes can accommodate hydrophobic molecules in their bilayer domains (Nasr et al., 2008; Suwakul et al., 2006). Niosomes can also minimise the side effects of some potent drugs by localising the pharmacological effects to the target tissues (Baillie et al., 1986; Baillie et al., 1985; Guinedi et al., 2005; Vyas et al., 1998). Niosomes can serve as a non-toxic penetration enhancer for transdermal drug delivery (Schreier & Bouwstra, 1994; Solanki et al., 2010). Moreover, niosomes can improve the chemical stability of photosensitive drugs by entrapping drug molecules in the aqueous internal milieu (Manconi et al., 2003). Below are discussed the most common applications of niosomes as drug delivery systems. 3.1.3.1. Anticancer delivery Methotrexate-loaded niosomes were first utilised as a more stable and economic alternative drug delivery system to liposomes, in order to improve the therapeutic index of the anticancer drug by localizing the cytotoxic effects to target cells (Baillie et al., 1985). The anti-tumour activity of vincristine also increased in S-180 sarcoma and Ehrlich-ascites-bearing mice when encapsulated in niosomes (Parthasarathi et al., 1994). Moreover, Span 60 bleomycin niosomes increased the anti-tumour action of bleomycin in the abovementioned two tumour models (Naresh & Udupa, 1996). Cytotoxic side effects of doxorubicin decreased when encapsulated in C16G2 niosomes. Doxorubicin copolymer-loaded niosomes were not hemolytic in vitro at the doses selected for in 100 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties vivo use. C16G2 niosomes also showed localised and sustained effects on liver and spleen (Uchegbu & Duncan, 1997). Furthermore, Span 40- and Span 60-based niosomal delivery systems for 5-flurouracil showed a sustained and higher local concentration of the drug in the liver and kidney of the rats, compared with the injectable drug solution (Namdeo & Jain, 1999). 3.1.3.2. Anti-infective delivery Sodium stibogluconate is a drug used in the treatment of visceral leshmniasis (a protozoan infection of the reticuloendothelial system). The drug’s pharmacokinetics were equally altered by encapsulation into liposomes and niosomes (Hunter et al., 1988). It was found that niosomal sodium stibogluconate was significantly more active in reducing parasite burden than the free drug, and the effect observed after multiple dosing suggesting that niosomal formulations act as a depot within the liver. Rifampicin, an anti-tuberculosis drug, encapsulated in Span 85 niosomes was found to accumulate in the lungs of mice (Jain and Vyas, 1995b). Therefore, it can offer the possibility of improving anti-tuberculosis therapy (Jain & Vyas, 1995). Moreover, niosomes composed of Span 60: cholesterol: dicetyl phosphate at 4:2:1molar ratio successfully encapsulated ribavirin (an antiviral agent used for treatment of hepatitis C). The concentration of ribavirin in the rat liver obtained from the ribavirin niosomes was 6-fold higher than that obtained from the ribavirin solution after an intraperitoneal injection of a single dose equivalent to 30 mg/kg of the drug. This demonstrates effective liver targeting properties for niosomes (Hashim et al., 2010). 3.1.3.3. Macromolecules and vaccine delivery Span 60 niosomes loaded with tetanus toxoid (TT) were prepared by the REV method. The prepared niosomes were coated with a modified polysaccharide O-palmitoyl mannan (OPM), in order to protect the niosomes from enzymatic degradation in the gastrointestinal tract and to enhance their affinity toward the antigen presenting cells of Peyer’s patches. The results were compared with alum-adsorbed TT following oral and intramuscular administration, and it was observed that OPM-coated niosomes produced higher IgG levels compared with plain uncoated niosomes and alum-adsorbed TT upon oral administration. Oral niosomes also elicited a significant mucosal immune response (sIgA levels in mucosal secretions). The developed niosomes also elicited a combined serum IgG2a/IgG1 response suggesting that they were capable of eliciting both humoral and cellular response. The proposed system was simple, stable, and cost-effective and may be clinically acceptable. The study emphasised the potential of OPM101 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties coated niosomes as an oral vaccine delivery carrier (Jain & Vyas, 2006). Moreover, niosomes loaded with insulin were prepared using the TFH method. The Span 60 niosomes protected insulin against proteolytic enzymes, and showed good stability in the presence of sodium desoxycholate (Varshosaz et al., 2003). 3.1.3.4. Transdermal delivery Niosomes were studied as a transdermal drug delivery systems and their ability to enhance drug permeation through the intact stratum corneum was also investigated (Reddy & Udupa, 1993; Schreier & Bouwstra, 1994; Van Hal et al., 1996). For instance, small niosomes (100 nm) have been visualised by freeze fracture electron microscopy technique between the first and second layer of human corneocytes 48 h after incubation with niosomes (Junginger et al., 1991). The penetration of this upper layer by niosomes appeared feasible (Junginger et al., 1991). However, no transport of niosomes took place across the whole skin layer (Junginger et al., 1991). In vitro studies on the transdermal penetration of oestradiol using high-phase transition sucrose ester niosomes or C18EO7 niosomes and low phase transition poly-7-oxyethylene alkyl ether (C12EO7) niosomes revealed that the latter were better transdermal carriers (Van Hal et al., 1996). C12EO7 micelles were ineffective as drug carriers in this study. The higher flexibility of these bilayers is believed to be responsible for the improved transdermal penetration (Van Hal et al., 1996). Also, it was found that reducing the cholesterol content in niosomes increased the transdermal delivery of oestradiol. These can be attributed to increases in the flexability of the vesicles (Van Hal et al., 1996). Niosomes were developed for the transdermal drug delivery of nimesulide (non-steroidal antiinflammatory drug). The prepared niosomes were evaluated in vivo in rats using the carrageeninduced rat paw oedema method. This investigation conclusively demonstrated that the niosomes successfully prolonged the drug release into the skin and across the skin (Shahiwala & Misra, 2002). Furthermore, permeation of ketorolac (a potent nonsteroidal anti-inflammatory drug) across excised rabbit skin from various proniosome gel formulations was investigated using Franz 102 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties diffusion cells. The prepared proniosomes significantly improved drug permeation and reduced lag time (Ibrahim et al., 2005). 3.1.3.5. Ophthalmic delivery For the ideal properties and application of niosomes as an ocular drug delivery system refer to Niosmes at Section 1.2.4.7 in Chapter 1. Furthermore, the physical stability of niosomes will be discussed in Chapter 5. 3.2. Chapter aims The drug under investigation (NTX) is water-soluble, and as general rule of thumb, hydrophilic drugs are less preferentially up taken by niosomes than lipophilic ones. The aim of this chapter is to study the effect of the main constituting lipids (surfactants and cholesterol) on niosomes formation and their physical properties. Two classes of surfactants are used; sorbitan esters (Span®) and polyoxyethylene alkyl ethers (Brij®). These surfactants are commercially available and widely used in the pharmaceutical, food and cosmetic industries (Lawrence, 2003; Yu, 2003). The surfactants used are selected for their versatility in chemical structures and gel/liquid transition temperatures. They have different polar head groups and hydrocarbon chains of various lengths and degrees of saturation. The specific objectives of this chapter were to: • Prepare niosomes by the conventional thin film hydration method using two classes of surfactants, which include sorbitan esters (Span 20, Span 40, Span 60 and Span 80) and polyoxyethylene alkyl ethers (Brij 52 and Brij 72). • Image niosomes macroscopically and microscopically using different techniques such as light/polarised light microscopy and cryogenic scanning electron microscopy. • Optimise cholesterol (a bilayer membrane stabiliser) levels in the prepared niosomes by studying the effect of cholesterol concentrations on: o Vesicle size. o Membrane fluidity (gel/liquid phase transition temperature). o Drug entrapment efficiency and loading efficiency. • Study the effect of total lipid content (surfactant/cholesterol) and the initial amount of NTX on entrapment efficiency and drug loading. 103 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3.3. Materials and methods 3.3.1. Materials NTX was purchased from the same source as mentioned in Chapter 2. Span 20, Span 40, Span 60, Span 80, Brij 52 and Brij 72 were purchased from Sigma-Aldrich, St. Louis, USA. All other solvents and buffer salts were of analytical grade and used as received. 3.3.2. Methods 3.3.2.1. Niosome preparation Thin film hydration method (TFH) Niosomes were prepared by adopting the TFH method (Azmin et al., 1985). Briefly, a specified amount (300 µmol) of lipids (surfactant: cholesterol) at a molar ratio of 1:0, 9:1, 8:2, 7:3, 6:4, and 1:1 (Table 3.1) was dissolved in a chloroform: methanol (2:1 v/v) mixture and rotary evaporated (Heidolph, Laborota 4000, GmbH, Germany) to form a thin lipid film. The dried lipid film was purged with a stream of nitrogen for 5 min in order to remove any residual traces of the organic solvent. The lipid film was then hydrated with either 5 ml of PBS pH 7.4 or NTX solution (1mg/ml) in PBS at 200 rpm for 1 h. The selected hydration time at 1h was based on preliminary studies. Longer hydration time did not show further improvement in terms of NTX loading. The resultant niosomal suspension was set aside for at least 2 h at room temperature, to allow the vesicle’s membrane to anneal. The formed niosomes were stored in a fridge for subsequent analyses. 104 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Table 3.1 Composition of the prepared niosomes using various surfactant: cholesterol molar ratios Molar ratio Span 20 1:1 6:4 7:3 8:2 9:1 Span 40 1:1 6:4 7:3 8:2 9:1 Span 60 1:1 6:4 7:3 8:2 9:1 Span 80 1:1 6:4 7:3 8:2 9:1 Brij 52 1:1 6:4 7:3 8:2 9:1 Brij 72 1:1 6:4 7:3 8:2 9:1 Surfactant (g/mol) 346.00 52.00 62.28 72.66 83.00 93.42 403.00 60.45 72.54 84.63 96.72 108.81 431.00 64.65 77.58 90.50 103.44 116.37 429.00 64.35 77.22 90.10 102.96 115.83 330.00 49.50 59.40 69.30 79.20 89.10 358.00 53.70 64.44 75.19 85.92 96.66 Cholesterol (g/mol) 386.65 58.00 46.40 35.00 23.20 11.60 386.65 58.00 46.40 35.00 23.20 11.60 386.65 58.00 46.40 35.00 23.20 11.60 386.65 58.00 46.40 35.00 23.20 11.60 386.65 58.00 46.40 35.00 23.20 11.60 386.65 58.00 46.40 35.00 23.20 11.60 Total lipid weight (mg) Equivalent no. of moles (µmol) 110.00 108.68 107.46 106.20 105.02 300 300 300 300 300 118.45 118.94 119.43 119.92 120.41 300 300 300 300 300 122.65 123.98 125.50 126.64 127.97 300 300 300 300 300 122.35 123.62 125.10 126.16 127.43 300 300 300 300 300 107.50 105.80 104.30 102.40 100.70 300 300 300 300 300 111.70 110.84 110.19 109.12 108.26 300 300 300 300 300 105 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3.3.2.2. Physical properties Niosome imaging Plane and polarised light microscopy To assure vesicle formation, all lipid dispersions were examined using a light microscope equipped with a cross polariser. Briefly, a drop of the prepared lipid dispersion was placed on a microscope glass slide, covered with a glass cover slip and examined using a polarised light microscope (Leica DMR, GmbH, Germany). The samples were examined for vesicle formation. Large multilamellar vesicles were recognised by their characteristic Maltese cross textures (Manosroi et al., 2003; Weder & Zumbuehl, 1984). Cryogenic scanning electron microscopy (Cryo-SEM) The morphology and lamellarity of the prepared niosomes were studied using cryo-SEM. A drop of the prepared niosomal suspension was sandwiched between two fracture rivets (internal diameter = 1.58 mm). This assembly was held by a gold-plated copper sample holder. The whole assembly was connected to a transfer rod, slushed in liquid nitrogen under a vacuum and then transferred into a Cryo-unit (Gattan Alto 2500, England). A fracture was created by removing the upper rivet using a scalpel. The fractured sample was heated up to -85oC for 30 min to sublime the surface moisture. Subsequently, the sample was gold coated and transferred to SEM (Philips XL30S FEG, Netherlands) for imaging at a temperature -120oC and voltage of 5 kV. Entrapment efficiency percentage (EE %) and drug loading percentage (DL%) The EE% and DL% determination methods used relied on destruction of the bilayer membranes and subsequent quantification of the released material (Abdelbary & El-gendy, 2008; Nasr et al., 2008). Ultracentrifugation was adopted to determine the encapsulated amount of NTX in niosomes. In brief, NTX-loaded niosomes were separated from the non-entrapped NTX molecules by diluting 2 ml of niosome dispersion to 30 ml with PBS, prior to centrifugation at 60,000 g (Sorvall Discovery 100S, USA) for 30 min. The formed niosomal pellets were washed with an additional 30 ml of PBS and centrifuged for a further 30 min. The niosomal pellets were re-suspended in 2 ml of PBS and 2ml of isopropanol were added (1:1 v/v), followed by 5 ml of PBS. The dispersion was re-centrifuged and the supernatant was quantitatively assayed for NTX content using high performance liquid chromatography (HPLC). The HPLC system used (Agilent 1200, Germany) was equipped with a C18-reverse phase column (Supelcosil LC-18DB 25 cm x 4.6 mm, 5 µm). The mobile phase comprised an acetonitrile: phosphate buffer pH 4 (40 mM) 106 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 30:70 system, and detection was achieved using a PDA detector at 214 nm. The drug entrapment efficiency was expressed as a percentage (EE %) and calculated using equation 3.1(Nasr et al., 2008): A X 100 Ao EE % = equation 3.1 Where A is the amount of NTX entrapped by niosomes and Ao is the initial amount of NTX used. DL % was calculated using equation 3.2 (Wang et al., 2004). DL% = A X 100 W equation 3.2 Where A is the amount of NTX entrapped by niosomes and W is the total weight of lipid used initially. Niosome size measurements The size of the prepared vesicles was determined by a laser diffraction technique (Mastersizer 2000, Malvern Instruments, UK). The samples were properly diluted with PBS and measured at 25 oC. Niosomes size was expressed in terms of volume diameter (D [4,3]) and the measurements were done in triplicate and the average values were used. Membrane fluidity (gel/liquid transition temperature) studies of niosomes using differential scanning calorimetry (DSC) Niosomal dispersions containing 100 mg/ml of the total surfactant/lipid mixtures were prepared using the TFH method, as described in Section 3.3.2.1. The niosomal dispersions comprised 1:0, 9:1, 8:2, 7:3, 6:4, and 1:1 Span 60: cholesterol molar ratio. A small amount (approx. 5 mg) of the prepared niosomes was accurately weighed in an aluminium pan, covered with an aluminium lid and hermitically sealed using a pan press (Thermal Science, USA). Another pan containing an equivalent amount of water was sealed and used as a reference cell. The temperature of the pans was raised from 4 to 80oC at a rate of 5oC/min using a differential scanning calorimeter connected to an automatic sampler (Q 1000, Tzero series, Thermal Analysis, USA). Nitrogen gas was purged at a flow rate of 45 ml/min. 107 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3.3.3. Statistical analysis Statistical significant differences between formulations were analysed, using one-way analysis of variance (ANOVA) and student t test at the 5% significance level, by GraphPad Software Version 3.05, San Diego California, USA. 3.4. Results and discussion 3.4.1. Niosome formation and imaging using plane /polarised microscopy The following surfactants were investigated for their ability to form niosomes in combination with different concentrations of cholesterol (0% to 50 % mol/mol) using the TFH method: sorbitan esters (Span 20, Span 40, Span 60 and Span 80) and polyoxyethylene alkyl ethers (Brij 52 and Brij 72). Table 3.2 shows the chemical structures, gel-liquid transition temperatures (Tm) and hydrophilic-lipophilic balance (HLB) values of the surfactants used, and summarises the ability of each surfactant to form niosomes alone and the minimal level of cholesterol required to form niosomes. The formed niosomes were visualised under a polarised light microscope, and Figure 3.6 shows representative micrographs of niosomes. Different textures, such as uniform spheres and Maltese crosses, are characteristic under the normal and polarised light microscope (Mura et al., 2007). Span 40, Span 60, Brij 52 and Brij 72 formed niosomes without cholesterol. On the contrary, Span 20 and Span 80 did not self-assemble into niosomes unless 10 to 20% mol/mol of cholesterol was incorporated. These findings can be attributed to the geometrical packing parameters of the surfactants used. The geometrical packing parameters depend on the optimal polar head group area ao, the volume of the hydrocarbon chains v and the maximum effective length that the hydrocarbon chains can assume (critical chain length) lc. Accordingly, the value of the dimensionless ‘packing parameter’ v/ ao lc will determine whether a certain amphiphile will form niosomes (1/2 < v/ ao lc < 1), or other surfactant aggregates, such as spherical micelles ( v/ ao lc < 1/3), non-spherical micelles (1/3 < v/ ao lc < 1/2), or inverted micellar structures (v/ ao lc > 1), or even precipitate out of solution (Israelachvili et al., 1980; Israelachvili & Mitchell, 1975). The niosomal dispersions formed and contained 0% cholesterol exhibited extensive gelling at the room temperature. The formed gel was difficult to separate from the non-entrapped drug molecules. Additionally, all niosomal dispersions prepared from Brij 52 showed phase separation and creaming upon storage overnight (Figure 3.7A). Therefore, all Brij 52 niosomes were excluded from any further investigation. 108 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Table 3.2 Chemical structure, phase transition temperature, HLB and vesicle-forming ability of the used surfactants a Phase transition temperature (Tm) of the unhydrated samples. b Hydrophilic- lipophilic balance (HLB) values were given by suppliers. * Vesicle formation was determined by visualisation of Maltese cross textures under polarised light 109 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.6 Representative micrographs of Brij 72:cholesterol 7:3 mol/mol-based niosomes under plane light (left) and polarised light microscope (right) A 1:1 6:4 7:3 B 1:1 6:4 7:3 Figure 3.7 Niosomal dispersions containing various molar ratios of Brij 52:cholesterol (A) and Span 60:cholesterol (B); and showing phase separation of the Brij 52 niosomes (A) and poor lipid hydration and non-vesicle aggregates at higher contents of cholesterol in Span 60 niosomes (B) 110 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3.4.2. Niosome imaging using Cryo-SEM The microstructure and lamellarity of the prepared niosomes were studied using Cryo-SEM (Figure 3.8). The micrographs captured revealed spherical-shaped niosomes and typical concentric multilayers, organised in an onion-like or rose-like morphology, indicating multilamellarity of the prepared niosomes. Niosomes of rough surfaces were obvious due to the sample preparation technique used which is based on snap freezing in liquid nitrogen and surface fracture formation by a scalpel. It is worth noting that Cryo-SEM has been reported to be more reliable than freeze-fracture transmission electron microscopy (FF-TEM) (Egelhaaf et al., 2003; Perrie et al., 2007). The steps involved in FF-TEM are likely to promote the formation of artefacts during sample preparation and replica creation (Egelhaaf et al., 2003; Perrie et al., 2007). Figure 3.8 Cryo-SEM micrographs of Span 60: cholesterol 7:3 mol/mol (left) and Brij 72: cholesterol 7:3 mol/mol (right) niosomes 3.4.3. Niosome size measurements Table 3.3 shows the D [4,3] of the prepared niosomes composed of various surfactants and contained various cholesterol levels (30%-50% mol/mol). The mean D [4,3] values of the prepared niosomes ranged from 7.0 ± 1.0 to 14.6 ± 0.8 µm (Table 3.3). The size measurements were repeatable and reproducible from batch to batch (acceptable deviations from the average values, Table 3.3). 111 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Table 3.3 Effect of surfactant type and cholesterol level on the D [4,3] and EE % of the prepared niosomes (Results are expressed as mean values ± SD, n = 3) Span 20 Span 40 Span 60 Span 80 Brij 72 Cholesterol level (% mol/mol) 30% 40% 50% D [4,3] (µm) 14.2 ± 0.6 14.5 ± 0.65 14.6 ± 0.8 12.7± 1.0 13.0±1.50 13.4±1.20 7.3 ± 0.62 7.6 ± 0.75 7.9 ±0.90 7.0 ± 1.00 7.4 ± 0.88 7.5 ± 1.20 9.0 ± 0.82 9.5 ± 0.88 9.8 ± 1.0 Cholesterol level (% mol/mol) 30% 40% 50% EE% (w/w) 3.5 ± 0.23 3.0 ±0.15 2.5 ±0.23 9.0 ± 0.67 7.0 ± 0.62 6.0 ±0.54 13.5 ± 0.56 9.0 ± 0.45 7.0 ±0.67 3.2 ± 0.22 4.0 ± 0.35 2.8 ± 0.25 7.0 ± 0.55 5.0 ± 0.47 4.5 ± 0.32 The increases in the D [4,3] values were insignificant (P > 0.05), when cholesterol level ranged from 30% to 50% mol/mol. Similar results were reported elsewhere (Essa, 2010). Span 40: cholesterol-based niosomes showed insignificant increase in sizes when cholesterol ratio increased from 2:1 to 1:1 mol/mol. Cholesterol is a rigid molecule with an inverted cone shape. It can be intercalated between the fluid hydrocarbon chains of the amphiphile when hydrated at a temperature above the gel/liquid transition temperature. It can also produce larger vesicles (Essa, 2010; Uchegbu & Florence, 1995). Cholesterol is a membrane stabiliser and can be accommodated and work cooperatively with the surfactant bilayer membranes. However, this accommodating characteristic is limited to a certain cholesterol level which is apparently equal to 30% mol/mol. The ability of the investigated surfactants to accommodate cholesterol was obviously decreased when cooled below the melting point, where the hydrocarbon chains are fully extended (all-trans) and exist in a gel state. Figure 3.7B shows plain niosomal dispersions containing different levels of cholesterol after storage at the room temperature overnight. Such storage allows the vesicle’s membrane to anneal and correct any membrane defect. Niosomal dispersions containing cholesterol > 30% mol/mol were poorly hydrated and excess cholesterol precipitated out (Figure 3.7B). This may be attributed to the compromised ability of the hydrocarbon chains of the surfactant used to accommodate cholesterol (Grant et al., 2001; Israelachvili et al., 1980). Table 3.3 also showed that the D [4,3] values of the prepared niosomes were significantly (P < 0.05) dependent on the HLB of the used surfactants. This effect was more clearly illustrated in 112 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.9. The D [4,3] values of Span 60-based niosomes were significantly smaller than the Span 40-based ones. Lipid aggregation and self-assembly in an aqueous medium is mainly controlled by two opposing electrostatic forces. These are the hydrophobic attraction forces between the hydrocarbon tail chains and the electrostatic repulsion of the polar head groups (Tanford, 1973). The surface free energy increases with any decrease in hydrophobicity (Yoshioka et al., 1994). Hence, surfactants of higher hydrophilicity (higher HLB) should yield larger vesicles. Figure 3.9 Effect of HLB on Span-based niosomes composed of surfactant:cholesterol at 7:3 molar ratio (Results are expressed as mean values ± SD, n = 3) * Significant effect (P < 0.05) ** No significant effect (P > 0.05) 113 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3.4.4. NTX EE% 3.4.4.1. Effect of surfactant type on NTX EE% The EE% values for the prepared niosomes ranged from 2.5% ± 0.23 to 13.5% ± 0.56 (Table 3.3). Span 40 and Span 60 established the highest EE%, whereas the lowest EE% values were related to Span 20 and Span 80. The main factors affecting the entrapment of small water-soluble drug molecules in the bilayer vesicles (niosomes) are the permeability of the biomelcular membranes and the structure continuity of the hydrocarbon chain of the bilayer-forming surfactant (Grant et al., 2001). Span 40 and Span 60 have saturated acyl chains [palmityl (C-16) and stearyl (C-18) chains respectively] which are in a gel state at ambient conditions (Grant et al., 2001). Such features render these vesicles less likely to leak the encapsulated water-soluble drug molecules (NTX). In contrast, Span 20 and Span 80 are fluid at room temperature and, in the presence of cholesterol, are disorganised due to the trans-gauche conformations of their acyl chains. The introduction of double bond at C9 position of the oleyl chain of Span 80 creates a twist or imperfection of the niosomal bilayer membranes (kinks). These kinks create vacancies in the hydrophobic chains and renders Span 80-based niosomes more leaky and permeable to NTX than even Span 20 (C12)-based niosomes. These findings explain the lowest EE% achieved by Span 80-based niosomes. The EE% values for the prepared niosomes can be summarised as follows: Span 60 > Span 40 > Brij 72 > Span 20 > Span 80. These results support the hypothesis that the higher the transition temperature of the surfactant is, the higher the corresponding EE% will be for water-soluble solutes (Yoshioka et al., 1994). This effect is clearly illustrated in Figure 3.10. 114 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.10 Effect of gel/liquid transition temperature on EE% for niosomes composed of surfactant:cholesterol at 7:3 molar ratio (Results are expressed as mean values ± SD, n = 3) * Significant effect (P < 0.05) 3.4.4.2. Effect of cholesterol on NTX EE% The literature details controversial results concerning the effect of cholesterol on EE%. Some studies have shown that cholesterol has no effect on EE% (Uchegbu & Florence, 1995). Other studies demonstrated that EE% increases with increasing cholesterol concentrations from 0% to 50% mol/mol (Yoshioka et al., 1994). On the other hand, other reports have shown that cholesterol had a limiting (up to 30% mol/mol) enhancement of EE% (Abdelbary & El-gendy, 2008; Moazeni et al., 2010). Our results were found to agree with the final opinion. Figure 3.11 illustrates the effect of cholesterol levels on NTX EE%. The results demonstrated that the cholesterol level significantly (P < 0.05) affected the ability of niosomes to entrap NTX. Increasing the cholesterol concentration can improve the EE% of NTX, but only to a certain extent. Any further increase of cholesterol beyond a certain concentration (30% mol/mol for Span 40, and Span 60 and 40% mol/mol for Span 80) markedly decreased EE%. 115 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties The initial increase in NTX EE% is likely due to the initial intercalation of cholesterol molecules between the hydrocarbon chains of the surfactant bilayers, hence, the permeability to NTX (a water soluble molecule) decreases. The prepared niosomes established EE% peaked at 30% mol/mol. Span 80 niosomes exhibited EE% peaking at 40% mol/mol due to structural disorganisation and kinks in the oleyl chains of Span 80 niosomes. These vacancies in the bilayer chains can accommodate more cholesterol molecules to reach the peak of EE%. Further increases in cholesterol concentration could increase the rigidity of the bilayer membranes, by virtue of cholesterol’s rigid structure and characteristic inverted cone shape (Israelachvili et al., 1980). This results in disruption of the vesicles’ bilayer structure which compromises their ability to entrap NTX (Manosroi et al., 2003). This could lead to a decrease in the total amount of lipid available for encapsulating the drug. Consequently, there is a decrease in the ability of the niosomes to entrap such water-soluble molecules. These findings are displayed in Figure 3.7B; the glass vials show precipitates at a cholesterol level higher than 30 mol%. These results were in a good agreement with a previous study. Uncharged niosomes composed of Tween 60 and cholesterol at 1: 0.5 mol/mol had a higher EE% than those containing higher ratio of Tween 60 and cholesterol at molar ratio of 1:1 (Abdelbary & El-gendy, 2008). 116 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.11 Effect of cholesterol concentration on EE% for Spanbased niosomes (Results are expressed as mean values ± SD, n = 3) 3.4.4.3. Effect of total lipid content on NTX EE% and DL% The effect of total lipid content on the EE% and DL% of NTX was investigated (Figures 3.12 and 3.13). The results revealed a linear increase (R2 > 0.98) in the EE% of the drug with increasing the total lipid content of the same composition (Span 60:cholesterol in 7:3 mol/mol). This is due to the increasing the lipid/drug ratio available for hydration. 117 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.12 Effect of the total surfactant/lipid content on EE % for Span 60: cholesterol (7:3 mol/mol) niosomes (Results are expressed as mean values ± SD, n = 3) On the contrary, DL% non-linearly decreased with increasing surfactant/lipid content (Figure 3.13). Similar results have been reported with both liposomes (Kirby & Gregoriadis, 1984a) and proniosomes (Mokhtar et al., 2008). The DL% values were found to decrease with increasing the total surfactant/lipid amounts. This could be ascribed to the lower hydration efficiency associated with increasing the lipid content. The anhydrous lipid film becomes hydrated and rounds off into vesicles encapsulating a small portion of the aqueous phase. When the surfactant/lipid content increases, an inevitable portion of the surfactant/lipid never comes in contact with the aqueous phase. This unhydrated portion could increase with increasing the the surfactant/lipid content. 118 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.13 Effect of the total surfactant/lipid content on DL % for Span 60: cholesterol (7:3 mol/mol) niosomes (Results are expressed as mean values ± SD, n = 3) 3.4.4.4. Effect of initial NTX concentration on NTX EE% and DL% Figure 3.14 shows the effect of the initial amount of NTX on EE%. The results revealed that the EE% markedly decreased when the initial amount of NTX used increased. Since it is a weakly basic drug with a low octanol/buffer partition coefficient, NTX can show amphipatic properties, especially at the elevated temperature encountered during niosomes formation where ionisation of the nitrogenous basic drug (NTX) considerably decreased. The interaction of unionised species with bilayer membranes could not be ruled out especially at higher NTX concentrations and the elevated temperature during the self-assembly of the surfactant molecules. This interaction could disrupt the bilayer formation of niosomes and limit the ability of niosomes to entrap NTX. This was evidenced by precipitating out non-bilayer lipid aggregates when increasing the initial NTX concentration. 119 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.14 Effect of the initial amount of NTX on EE % for Span 60: cholesterol (7:3 mol/mol) niosomes (Results are expressed as mean values ± SD, n = 3) The effect of the initial amount of NTX on DL% is illustrated in Figure 3.15. The DL% slightly increased at an initial NTX concentration of 2 mg/ml and then steeply declined at a higher NTX concentration (3 mg/ml). The DL % results suggested that the initial NTX concentration of 2 mg/ml is optimal for NTX loading and it can be utilised for further NTX loading studies. 120 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.15 Effect of the initial amount of NTX on DL % for Span 60: cholesterol (7:3 mol/mol) niosomes (Results are expressed as mean values ± SD, n = 3) 3.4.5. DSC studies 3.4.5.1. Effect of cholesterol on gel/liquid phase transition of Span 60 niosomes DSC was employed to understand the interaction of cholesterol with the single-chain non-ionic surfactant Span 60, and to monitor its effect on the gel/liquid transition of the prepared niosomes. The bilayer membrane is an ordered structure which may exist in a gel or liquid crystalline state. Essentially, molecules are more mobile in a liquid crystalline state, enjoying the lateral diffusion within the bilayer that they are denied in the gel state (Israelachvili et al., 1980). The liquid crystalline state exists at a higher temperature than the gel state. An increase in temperature favours the transition from the gel to the liquid state because of the entropy gain (ΔS) associated with this transition. This ultimately leads to a lowering of the free energy (ΔG) of the system (Israelachvili et al., 1980). Figure 3.16 demonstrates the effect of cholesterol level on the gel-liquid transition of Span 60 niosomes. Table 3.4 presents two parameters: the gel/liquid transition temperature (Tm), which is the temperature peak point at which transition (melting) from gel to solid is half-complete; and 121 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties the transition enthalpy, which is the amount of heat in calories/mol required for the entire transition to take place. The thermogram of Span 60-based vesicles with 0% cholesterol shows a thermal event at 63oC (Figure 3.16), which was markedly higher than that of the constituting Span 60 surfactant (melting point 55oC) in the unhydrated form. This increase in the Tm of the Span 60 might be attributed to the hydration and self-assembly of the surfactant molecules forming closed bilayer structures in water. As such, greater energy was required to break the strong hydrophobic interaction between surfactant moieties in the hydrated state. However, unlike phospholipids, no pre-transition peak was seen with Span 60 (Figure 3.16). This might be due to the well defined structure and composition of the non-ionic surfactant used. Cholesterol exhibited a marked effect in decreasing both Tm and transition enthalpy. This was clearly evident when gradually decreasing both the area under the peaks of DSC thermograms and Tm values. The cholesterol contents in the membrane increased until the complete disappearance of the gel-liquid transition of Span 60 at 50% mol/mol. This is due to intercalation of cholesterol molecules between the lipid bilayer, which weakens Van der Waals forces between the hydrocarbon chains of surfactant’s bilayer membranes. These findings were in accordance with previous reports, where it was found that cholesterol could alter the fluidity of the bilayer membranes of liposomes (Taylor & Craig, 2003). Finally, it is worth noting that the residual gelliquid transition recorded for niosomes containing 30% and 40% mol/mol of cholesterol might be advantageous. Such transition can impart bilayer flexibility and thermo-responsiveness. As such this renders them promising carriers for ocular drug delivery. On the contrary, complete abolishment of the gel-liquid transition makes such niosomes rigid and as such less attractive for ocular applications. 122 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Table 3.4 DSC parameters of Span 60-based niosome dispersions containing different concentrations of cholesterol Tm (oC) ∆H (Cal/mol) 0 63.44 89.00 10 55 45.10 20 51 2.05 30 47.22 0.82 40 43.5 0.43 50 No transition - Cholesterol concentration (% mol/mol) 123 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Figure 3.16 DSC thermograms of Span 60-based niosomes containing different concentrations of cholesterol 124 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties 3.4.5.2. Effect of NTX on gel-liquid transition of Span 60 niosomes The effect of the encapsulated NTX on the Tm and enthalpy (heat content, ∆H) was also investigated (Figure 3.17). The results showed that the Tm and enthalpy values of NTX free and loaded niosomes (7 Span 60: cholesterol: 3 mol/mol) were 47.22, 46.8oC and 0.82, 0.73 cal/mol respectively. The results showed a slight difference in Tm and enthalpy of the NTX loaded niosomes and NTX free niosomes. However, the differences were insignificant (P > 0.05, Student t-test). Therefore, the entrapment of NTX molecules have a slight or even no effect on the fluidity of the bilayer membranes of the prepared niosomes, suggesting that the drug was encapsulated inside the aqueous core. These results are correlated well with the partition coefficient studies which indicate that NTX is a hydrophilic drug. Figure 3.17 Effect of NTX on gel/liquid transition temperature of (Span 60: cholesterol 7:3 mol/mol) niosomes 125 Chapter 3….Effect of Surfactants and Cholesterol on Niosomes Physical properties Conclusion • Span 60 is a promising surfactant, in terms of its ability to encapsulate NTX compared with other surfactants. • NTX solutions of 2 mg/ml and 1mmol of lipid will be used for further studies, for better NTX loading in niosomes. • Optimising cholesterol level can significantly alter the physical properties of the prepared niosomes, most notably their thermo-responsiveness. • Such behaviour is likely to render these vesicles more amenable to the ocular temperature and less likely to create a foreign body sensation (and subsequent reflex tear production and blinking) when instilled topically to the eye surface. Unfortunately, conventional niosomes (surfactant and cholesterol) have limited ability to encapsulate a considerable amount of NTX, in order to attain the required therapeutic dose (0.4 mg/ml). Therefore, the next chapter is concerned with studying the effect of some selected excipients and other non-conventional methods of niosome preparation, in an attempt to enhance NTX EE% and to obtain various niosomal formulations with versatile properties for the effective ocular delivery of NTX. 126 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 4. Effect of membrane additives on the physical properties of niosomes: size, morphology, spreading ability, rheological and stabilising properties 4.1. Introduction 4.1.1. Bilayer membrane additives Bilayer membrane additives can be defined as lipid excipients which are unable to form lipid vesicles (liposomes and niosomes) by their own but they can be accommodated by the bilayer domains. The membrane additives have been found in the lipid vesicles basically for two reasons; either to alter the biodistribution, or to improve the physical properties. Non-ionic surfactants can form niosomes in absence of other membrane additives under specified conditions. However, such niosomes have been found to be highly permeable and leaky to the solutes. Cholesterol is the most common membrane additive found in lipid vesicles. Cholesterol can significantly alter membrane permeability, membrane fluidity, vesicle stability, entrapment efficiency and niosomes size and it is widely known as a membrane stabiliser. Other membrane additives can be broadly classified into charge inducers and water-soluble surfactants. 4.1.1.1. Charge inducers Stearyl amine and dicetyl phosphate are positive and negative charge inducers respectively, and are frequently incorporated in the bilayer membranes. These charge inducers have been mainly utilised to improve the physical stability of the vesicular dispersions against aggregation (Uchegbu & Vyas, 1998); to prolong the half life of circulating lipid vesicles in plasma (Maurer et al., 2007); and to prolong the ocular residence time of topically instilled ophthalmic lipid vesicles (Kaur et al., 2004). For example, incorporation of positively charged lipids such as stearyl amine is essential for successful delivery of negatively charged polynucleotides and efficient transfection into the cells (Maurer et al., 2007). Also, incorporation of dicetyl phosphate in niosomes significantly improved gentamicin sulfate entrapment efficiency in the prepared niosomes for ocular delivery compared with the neutral niosomal formulations (Abdelbary & El-gendy, 2008). Negatively charged niosomes have been also found most effective in prolonging gentamicin release rate (Abdelbary & El-gendy, 2008). 127 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes On the other hand, introduction of water soluble surfactants into vesicles’ bilayer membranes produces liposomes and niosomes with numerous applications. 4.1.1.2. Water soluble surfactants Liposomes The addition of surfactant molecules to vesicular dispersions of phospholipids leads to breakdown or solubilisation of the liposomes and formation of a mixed micellar phase. Such vesicle to micelle transitions of phosphatidylcholine liposomes have been achieved by the addition of non-ionic surfactants of the polyoxyethylene cetyl ether class (Kim & Kim, 1991). In this system a region occurs where lamellar and mixed micelles coexist. In another study, sodium cholate was found to induce a vesicle to micelle transition in cholate-phosphatidylcholine systems (Walter et al., 1991b). On increasing the bile salt concentration, more multilamellar vesicles were detected. In addition, “open” vesicles, large (twenty to several hundred nanometres in diameter) bilayer sheets, and long (150-300 nm) flexible cylindrical vesicles were also detected (Walter et al., 1991b). Furthermore, new classes of liposomes, called Transferosomes®, have been developed upon mixing phospholipids with some water-soluble surfactants, such as sodium cholate and Tween 80. Transferosomes® have the same morphology as liposomes, but they have ultradeformable bilayer membranes. These non-conventional liposomes have been reported to penetrate intact skin when applied non-occlusively. To prepare these vesicles, surfactants such as sodium cholate and sodium deoxycholate have been incorporated in the vesicular membrane, in a concentration range from 20%-50% mol/mol (Cevc et al., 1996; Planas et al., 1992). Niosomes Niosomes derived from non-ionic surfactants of low aqueous solubility are analogues of liposomes (Florence & Baillie, 1989). Niosomes are similar in many respects to liposomes, being prepared in the same way forming unilamellar or multilamellar structures which entrap lipidsoluble or water-soluble solutes (Uchegbu et al., 1992). The incorporation of poly-24oxyethylene cholesteryl ether (Solulan C24) into niosomes has been found to generate atypical niosomal structures such as polyhedral niosomes and discomes (Uchegbu & Florence, 1995; Uchegbu & Vyas, 1998). 128 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Polyhedral niosomes Polyhedral niosomes first obtained from surfactant systems were non-uniform spherical vesicles and had between 4 to 12 straight edges of similar length (Uchegbu et al., 1997). These nonconventional structures are termed polyhedral niosomes. Typically, the reported polyhedral niosomes are formed from mixtures of hexadecyl diglycerol ether (C16G2) and Solulan C24, and could also be formed from C16G2, Solulan C24 and minimum amount of cholesterol. These polyhedral vesicles have been found to be stable for at least 36 days and able to entrap and release slowly water-soluble markers, such as carboxyfluorescein, and nucleotides (Uchegbu & Florence, 1995; Uchegbu et al., 1997; Uchegbu & Vyas, 1998). Luteinizing hormone releasing hormone (LHRH)-loaded polyhedral and spherical niosomes were prepared as a slow release depot system for intramuscular administration. Spherical conventional niosomes showed more stable membranes than polyhedral ones, suggesting that the conventional niosomes are more effective intramuscular depots (Arunothayanun et al., 1999). Giant niosomes (discomes) Large disc-like niosomes (discomes) of 11-60 µm in size were formed upon incubation of the preformed conventional (spherical) niosome dispersion with various proportions of Solulan C24 in a shaking water bath at a temperature of 74oC for 1 h (Uchegbu et al., 1992). Larger discomes were formed with higher Solulan C24 levels and vice versa. A C16G2: cholesterol molar ratio of 7:3 favours the formation of discomes. Increasing the cholesterol level in the vesicular dispersion above 30% mol/mol suppresses the production of discomes due to increasing the rigidity of niosomes (Uchegbu et al., 1992). Discomes are thermoresponsive vesicles, which become leakier as the temperature is increased from room temperature to 37oC (Uchegbu & Vyas, 1998). Discomes have been proposed a potential drug delivery system for ophthalmic administration (Uchegbu et al., 1992; Uchegbu & Vyas, 1998). They are capable of entrapping water-soluble drugs and their large faceted structure may prevent their drainage and absorption into systemic circulation via the nasolacrimal duct. They have been investigated to control the ocular delivery of timolol maleate (Vyas et al., 1998). The prepared discomes showed a sustained drug release pattern and systemic absorption was minimised to a negligible level (Vyas et al., 1998). However, the formation of this system requires the incubation of the niosomal dispersion with Solulan C24 at 75oC for at least 1 hr. 129 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes These preparation conditions might affect the chemical stability of some thermolabile therapeutic agents when incubated at such high temperature. 4.1.2. Surfactant and ocular delivery The use of surfactants in ophthalmic formulations and ocular drug delivery is occasionally recommended. This is due to their potential ability to increase bioavailability by increasing drug solubility, prolonging pre-corneal retention, and enhancing permeability(Jiao, 2008). Surfactants are used as preservatives in multiple dose ophthalmic formulations to maintain sterility and protect the formulations against microbial instability. For example, benzalkonium chloride and chlorhexidine (cationic surfactants) are used as preservatives in ophthalmic formulations. However, cationic surfactants are found to be highly irritant and cytotoxic (Furrer et al., 2000; Saettone et al., 1996a). The order of surfactant toxicity is found to be anionic > cationic >> nonionic (Lang et al., 2002). Many surfactants have been studied as drug ocular penetration enhancers. Non-ionic surfactants including Span 20, Span 40, Span 85, Tween 20, Tween 40, Tween 81, Brij 35, Brij 58, Myrj 52 and Myrj 53 were tested for their ability to increase the corneal permeability of fluorescein in human subjects (Marsh & Maurice, 1971). The study showed that an HLB range of 16-17 was found to be effective as a corneal penetration enhancer in 1% solution. Only Brij 58 was harmful at this concentration (Marsh & Maurice, 1971). In addition, some bile salts have been studied for drug corneal penetration enhancing effect and ocular toxicity and irritancy; these include sodium deoxycholate, sodium taurodeoxycholate, sodium ursodeoxycholate and sodium tauroursodeoxycholate. The study pointed out that the aforementioned surfactants are safe and effective as corneal penetration promoters (Saettone et al., 1996a). 130 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 4.2. Chapter aims The use of some membrane additives can alter the physical properties of niosomes, such as size, morphology and EE% of the encapsulated drug. The aim of this chapter is to study stearyl amine, dicetyl phosphate, sodium cholate and poly-24-oxyethylene cholesteryl ether (Solulan C24) as potential membrane additives. The excipients used can be broadly categorised into charged lipids (stearyl amine and dicetyl phosphate) and water-soluble surfactants (sodium cholate and Solulan C24). The following objectives are considered: • Studying the effect of different membrane additives and preparation methods on EE% of NTX. • Investigating the effect of the membrane additives on the physical properties of the prepared niosomes such as vesicle size, morphology and gel/liquid phase transition temperature. • Exploring the spreading ability and rheological properties of the prepared niosomes. • Studying the effect of NTX encapsulation by niosomes on NTX chemical stability under stress factors such as artificial daylight illumination and oxidising agents. 131 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 4.3. Materials and methods 4.3.1. Materials Stearyl amine (STA) and dicetyl phosphate (DCP) were purchased from Sigma-Aldrich, St. Louis, USA; poly-24-oxyethylene cholesteryl ether (C24) was a generous gift from Lubrizol Inc., Cleveland, USA; sodium cholate (CH) was a generous gift from New Zealand Pharmaceuticals, Palmerton North, New Zealand. All other chemicals were purchased or obtained from the same source mentioned in Chapter 2 and Chapter 3. 4.3.2. Methods 4.3.2.1. Niosome preparation TFH method Niosomes were prepared using the TFH method as described in Section 3.3.2.1. An amount equivalent to 1 mmol of total lipid was used. The following lipid ratios were studied: Span 60: cholesterol:additive (DCP, STA, C24 or CH) 7:3:0, 6.9:2.9:0.2, 6.75:2.75:0.5, 6.5:2.5:1 mol:mol:mol and the final percentages of the membrane additive used were 0%, 2%, 5% and 10% mol/mol respectively. Three formulations as well as control niosomes (F-S60) were selected based on the highest EE%. Their compositions are presented in Table 4.1. Table 4.1 Codes and composition of the prepared niosomal formulations Formulation code F-S60 F-DCP F-C24 F-CH Span 60 7 6.75 6.9 6.75 Molar ratio Cholesterol 3 2.75 2.9 2.75 DCP 0 0.5 0 0 C24 0 0 0.2 0 CH 0 0 0 0.5 132 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Freeze and thaw method Frozen and thawed multilamellar vesicles (FAT-MLV) were generated by subjecting 8 ml of MLV suspensions prepared using the TFH method to 5 alternate cycles of freezing in liquid nitrogen for 60 s and thawing in a water bath at 60°C for another 60 s (Mayer et al., 1985). Reverse phase evaporation (REV) method The REV method used was a modification of the method described by (Kirby & Gregoriadis, 1984b). A thin film of the lipid was formed, as previously described in the TFH method. The thin film was then dissolved in 12 ml of a mixture of ether:chloroform (1:1 v/v). Four ml of the aqueous NTX solution (4 mg/ml) of PBS was added to the organic phase, such that the organic: aqueous phase ratio was 3:1. This mixture was then sonicated for 3 min in a bath sonicator (Bandelin Snorex, Berlin, Germany), until an opalescent w/o emulsion was formed. The final dispersion was rotary evaporated at 60oC until a semi-solid gel-like mass or aqueous lipid dispersion (depending on the lipid contents) was obtained. The resulting system was purged with a stream of nitrogen gas for 3 min to remove any traces of the organic solvent. The final dispersion was diluted with 4 ml PBS solution and rotary evaporated at a speed of 200 rpm for at least 30 min. This method was previously used to generate large unilamellar vesicles (LUV) (Kirby & Gregoriadis, 1984b). Dehydration-rehydration method Dehydration-rehydration vesicles (DRV) were prepared by a modified version of the method described by (Hope et al., 1986). Eight ml of MLV dispersion prepared by the TFH method was frozen in liquid nitrogen, followed by lyophilisation in a freeze dryer (Labconco, Missouri, USA) overnight at a temperature of -20°C and vacuum≤ 0.133 mbar . The freeze-dried mass was hydrated with 4 ml (4 mg/ml) NTX solution of PBS pH 7.4 at 60°C. After complete lipid hydration, the final volume of the preformed niosomes was diluted to 8 ml using a PBS solution. 4.3.2.2. Physical properties Entrapment efficiency (EE %) of NTX EE% was determined as mentioned in method section Chapter 3. Niosome size and size distribution measurements The size of the prepared niosomes was determined by the laser diffraction technique (Mastersizer 2000, Malvern Instruments, UK). The samples were properly diluted with PBS and measured at 133 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 25 oC. The size was expressed in terms of volume diameter (D [4,3]) and the experiments were done in triplicate. Span was calculated using equation 4.1: equation 4.1 Where d(0.9) is the particle diameter at 90% cumulative size, d(0.1) is the particle diameter at 10% cumulative size and d(0.5) is the particle diameter at 50% cumulative size Morphology Cryo-SEM The ultrastructure of the prepared niosomes was studied using Cryo-SEM as mentioned in method section Chapter 3. Confocal laser scanning microscopy (CLSM) Niosomes preparation Niosomes were prepared using the REV method, as described in Section 4.3.2.1. The NTX solution was replaced with a 1.5 mM carboxyfluorescein (CF) solution of PBS pH 7.4 as a fluorescent probe. The free CF was separated from the niosomes by exhaustive dialysis. A cellulose bag containing 2 ml of the prepared niosomes was dialysed against a 500 ml PBS solution at 4oC. The PBS solution was replaced at least 3 times over 48 h. Niosomes imaging Three wells were drilled into a microscopic plastic slide by a laser cutter. Each well was 1 cm in diameter and 1.5 mm in depth. The wells were created to hold niosome samples, in order to avoid squashing of niosomes between the slide and the cover slip. Measurements were conducted on a Leica CLSM (Leica DMRXA-2 microscope fitted with a TCS-SP2 scan head, Leica Microsystems, Heidelberg, Germany) using a 40x water immersion lens; a zoom of 1; a pinhole with an Airy disk diameter of 2; and a combination of lasers and emission band pass filters (510521 nm) to visualise the prepared niosomes. The images were visualised and analysed with Leica Confocal Software® (Leica Microsystems, Heidelberg, Germany). DSC studies DSC studies were conducted to study the effect of different membrane additives on the membrane fluidity of the prepared niosomes (gel/liquid transition temperatures). Special batches 134 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes of the four niosomal formulations (10 mmol of total surfactant/lipid) were prepared by the TFH method. The DSC study was performed as mentioned in method section Chapter 3. Surface tension measurements The surface tension (γ) of the prepared niosomes was determined at the ambient conditions using an interfacial tensiometer (Torsion Balance, Malvern Wells, UK). All measurements were performed in triplicate. Contact angle and spreading coefficient measurements Contact angle measurements were performed using a drop shape analyser (goniometer) (KSVCAM 101, Helsinki, Finland). Goniometry is the analysis of the shape of a drop of test liquid placed on a solid surface. The basic elements of the goniometer include a light source, sample stage, lens and image capture. The contact angle can be assessed directly by measuring the angle formed between the solid surface (a glass slide) and the tangent to the drop surface (Figure 4.1). A Hamilton syringe was filled with each of the tested niosomal formulations. Approximately 20 µl of each formulation were dropped onto a glass slide at ambient conditions (Figure 4.2). The image of the drop was captured and measured by CAM 101 software. The measurements were performed in triplicate and compared with that of an NTX (0.4 mg/ml) aqueous solution in PBS. Figure 4.1 Equilibrium between forces acting on a drop of liquid on a solid surface 135 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Poor wetting (θ > 90o) Good wetting (90o > θ > 0o) Figure 4.2 KSV-CAM 101 goniometer setup for measuring contact angle, (A) poor wetting, and (B) good wetting of a liquid drop on a glass slide 136 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes The contact angle (θ) is defined as the angle formed by a liquid at the three phase boundary where a liquid (L), air (A) and solid (S) intersect (Figure 4.1). A representation of the several forces acting on a drop of liquid placed on a flat solid surface is outlined in Figure 4.1. The surface tension of the solid (γS/A) will favour spreading of the liquid, but this is opposed by the solid-liquid interfacial tension (γS/L) and the horizontal component of the surface tension of the liquid (γL/A) in the plane of the solid surface (γL/ACOS θ). Equating these forces gives equation 4.2, which is generally referred to as Young’s equation (equation 4.2) (Florence & Attwood, 1998c). γS/A = γS/L + γL/A COS θ equation 4.2 The condition required for complete wetting of a solid surface is that the contact angle should be zero. This condition is fulfilled when the forces of attraction between the liquid and the solid are equal to or greater than those between liquid and liquid (Florence & Attwood, 1998c). The type of wetting in which a liquid spreads over the surface of the solid is referred to as spreading wetting. The tendency for spreading can be quantified in terms of the spreading coefficient (S) using equation 4.3 (Florence & Attwood, 1998c). S= If θ γL/A (COS θ – 1) is larger than 0o, the term (COS equation 4.3 θ – 1) will be negative, and the value of S as well. The condition for complete or spontaneous wetting is thus a zero value for the contact angle. Rheological properties measurements Viscosity measurements were performed using a Brookfield DV-III programmable cone and plate rheometer (Brookfield Engineering Laboratories Inc., Stoughton, Massachusetts, USA). A CP-40 cone spindle was used and a rotation speed was 100 rpm. A water-jacketed sample cup was thermostatically controlled by circulating water pumped using a water pump (Biolab Scientific 137 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Limited, Victoria, Australia) immersed in a water bath maintained at the required temperature (25oC ± 0.5oC and 35oC ± 0.5oC). Brookfield Rheocalc operating software was employed for data acquisition and analysis. The system was calibrated and found to be accurate within ± 2% of the working range. Prior measurements, the tested niosomes were diluted with PBS to a concentration equivalent to the required therapeutic NTX concentration (0.4 mg/ml). A sample volume of 0.5 ml was used and the measurements were performed in triplicate. Viscosity was calculated using equation 4.4 (Martin et al., 1993). η= F G equation 4.4 Where G is shear rate (s-1) which is the difference of velocity between two planes of liquid separated by an infinitesimal distance, F is shear stress (dynes/cm2) which is the force per unit area required to bring about flow and η is the viscosity (Martin et al., 1993). 4.3.2.3. Effect of niosomal encapsulation on NTX oxidation An aqueous NTX solution of 0.4 mg/ml in PBS pH 7.4 or an equivalent concentration of NTX encapsulated in niosomes were exposed to 2% v/v hydrogen peroxide (H2O2) for 2 h at ambient conditions. Another group of samples were stored in a BINDER KBF 240 series stability chamber (BINDER, Tuttlingen, Germany) and subjected to artificial daylight illumination of 10,000 lux at 40oC. Samples were withdrawn at a suitable time interval and analysed for NTX content using the HPLC method. 4.3.3. Statistical analysis A one-way analysis of variance (ANOVA) followed by Tukey’s pairwise comparisons at the 5% significance level was used to test statistical significance differences between the prepared formulations and control for size measurements, contact angle, spreading ability, rheological and protective effects of niosomes. These were performed using GraphPad Software Version 3.05, San Diego California, USA. 138 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 4.4. Results and discussion The structures, molecular weights and critical micelle concentrations of the used additives are presented in Table 4.2. Table 4.2 Chemical structure, molecular weight, CMC and phase transition temperature of the investigated additives * The phase transition temperature (Tm) of the unhydrated samples of Solulan C24 and dicetyl phosphate were measured using DSC. Sodium cholate was obtained from a data sheet by the supplier. ** Critical micelle concentrations (CMC) of the two water-soluble surfactants (Solulan C24 and sodium cholate) were determined by surface tension measurements using a torsion balance. 139 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 4.4.1. Effect of selected membrane additives on NTX EE% The effect of different concentrations (0%, 2%, 5% and 10% mol/mol) of the selected membrane additives on NTX EE% was studied (Figure 4.3). The EE% values calculated for the prepared niosomes were highly dependent on the type and concentration of the additives used. F-S60 (0% additive) achieved EE% of 11.5% w/w whereas the EE% declined to 7% w/w in niosomes composed of Span 60:cholesterol:STA at molar ratio of 6.9:2.9:0.2. Incorporation of STA in a concentration above 2% mol/mol was unsuccessful. This was due to the formation of large lipid aggregates which precipitated out from the dispersion. In contrast, the addition of DCP significantly (P < 0.05) increased NTX EE%. For instance, there were 2.7-fold and 3.5-fold increases in the NTX EE% when 2% and 5% mol/mol of DCP were incorporated in the niosomes respectively (Figure 4.3). Figure 4.3 Effect of selected membrane additives concentration on NTX EE% for niosomes prepared using the TFH method (Results are expressed as mean values ± SD, n = 3) 140 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes STA and DCP are positively and negatively charged lipids respectively. Being a basic drug (pKa 8.380), NTX molecules carried positive charges at the selected buffer pH (7.40). Electrostatic interactions between charged drug molecules (NTX molecules) and STA or DCP were likely to be the main factor controlling the EE% for niosomes (Abdelbary & El-gendy, 2008; Junyaprasert et al., 2008). Further, as the concentrations of the ionic additives increased, niosomes’ charge densities increased. These observations could explain the marked reduction and significant increase of the EE% of NTX in STA- and DCP-containing niosomes respectively. Figure 4.3 also shows levelling off the DCP curve at a concentration of > 5% mol/mol. This could be ascribed to the saturation of the bilayer membranes of the niosomes with DCP molecules. The effect of different concentrations of C24 and CH on the EE% was also illustrated in Figure 4.3. The results revealed that the incorporation of C24 or CH in niosomes significantly (P < 0.05) increased NTX EE% compared with F-S60. The EE% values calculated for niosomes containing 5% mol/mol CH (F-CH) and 2% mol/mol C24 (F-C24) were 1.74-fold and 2.4-fold higher respectively than that for F-S60. These increases in the EE% were limited to critical concentrations (Table 4.2). For example, any further increase in C24 concentration beyond 2% mol/mol was accompanied by a significant (P < 0.05) decrease in EE% of NTX, whereas 5% mol/mol of CH was found to be optimal EE% of NTX. These critical concentrations were dependent on the type of the additive used. This could be due to the difference in the accommodating abilities of the bilayer membrane to the additives used. For instance, the structure of C24 is inherently bulky (mol wt 1443) compared with Span 60 (mol wt 431) and cholesterol (mol wt 386.7). The presence of 24 units of polyoxethylene chains, which are highly hydrophilic and hydrated with numerous water molecules, might account for such behaviour. Furthermore, different additives could show different affinities to the bilayer membrane and water (El Maghraby et al., 2004). The positive effect of C24 and CH on the EE% of NTX could be ascribed to the improved hydration of the thin lipid film and the enhanced quality of the formed lipid dispersion due to minimisation of vesicle aggregation. This is due to the steric effect of the bulky hydrophilic head group of C24 and the possible electrostatic repulsion between negatively charged CH moieties. On the other hand, the use of additives at concentrations higher than the critical concentrations markedly decreased NTX EE%. For example, NTX EE% declined to 14% and 20% in niosomes containing 10% mol/mol CH and 10% mol/mol C24 respectively. Such behaviour could be due to 141 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes the increase in the concentration of the water-soluble surfactant additives used (CH and C24), which in turn could lead to an increase in the permeability of the formed niosomes to the entrapped solutes. Additionally, increasing the concentration of the water-soluble surfactants could lead to self-assembling of the lipids into other aggregates, such as open lamellar structures or even mixed micelles. It has been reported that intermediate aggregates between closed bilayer vesicles and mixed micelles were identified with cholate-phosphatidylcholine systems using turbidity measurement (Ollivon et al., 1988; Walter et al., 1991a). Intermediate structures between niosomes and mixed micelles were also identified with hexadecyl diglycerol ether: cholesterol: C24: DCP 49:19.5:29:2 (Uchegbu et al., 1992). The additives used achieved a significant enhancement of EE% and Table 4.2 shows formulation codes and compositions of the selected niosomal formulations. Thanks to these additives, F-DCP, F-C24 and F-CH formulations were of clinical significance, as they were able to carry the required therapeutic concentration (0.4 mg/ml), whereas F-S60 (0% additive) niosomes carried a sub-therapeutic dose of NTX. Studying the effect of different preparation methods could be other solution to enhance EE% for F-S60 control niosomes. 4.4.2. Effect of preparation methods on NTX EE% The four selected niosomal formulations (F-S60, F-DCP, F-C24 and F-CH) were prepared using three different methods (FAT-MLV, REV and DRV) and were compared with the conventional TFH method. Figure 4.4 outlines the effect of the method of preparation on NTX EE%. The results showed that, for the same niosomal formulation, NTX EE% values were significantly (P < 0.05) different and dependent on the method of preparation. For instance, the NTX EE% values for F-DCP prepared by TFH, FAT-MLV, DRV and REV were 40%, 29%, 43% and 61.5% respectively. Niosomes prepared using the REV method showed the highest ability to encapsulate NTX. On the contrary, those prepared using the FAT-MLV method showed the lowest EE%. There was a 1.4-fold decrease in the EE% obtained from FAT-MLV F-DCP compared with that obtained from TFH F-DCP. These findings could be ascribed to the different preparation methods. Subjecting the bilayer vesicles to drastic conditions such as 5 alternate cycles of freezing and thawing was thought to rupture the bilayer membrane and consequently increase their permeability to the entrapped solutes. These results were in a good agreement with those reported on ethosomes (liposomes containing ethanol) (Maestrelli et al., 2009). According to this report, 142 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes the effect of preparation methods was studied for ethosomal formulations for topical delivery of the local anaesthetic agent, benzocaine. The prepared ethosomes were prepared using TFH, FATMLV, REV and small unilamellar vesicles (SUV) by sonication or extrusion through 200 nm polycarbonate filters. Both SUV and FAT-MLV ethosomes showed the lowest EE%. Figure 4.4 Effect of the preparation method on NTX EE% (Results are expressed as mean values ± SD, n =3) 143 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes On the other hand, it was found that a 1.5-fold increase in the EE% for REV F-DCP compared with TFH F-DCP. This could be attributed to the ability of niosomes to encapsulate large aqueous cores during the slow evaporation of the organic phase and conversion from inverted micelles to bilayer vesicles. Niosomes prepared by the DRV method showed higher EE% than TFH niosomes. This could be due to the increase in the lipid/drug ratio which was achieved when rehydrating the freeze-dried niosomes in half of the original solution volume (Hope et al., 1986; Kirby & Gregoriadis, 1984a). This result was in a good agreement with that mentioned for the effect of total lipid content on EE% (Figure 3.12) in Chapter 3. 4.4.3. Niosome size and distribution measurements The effects of the membrane additives used on the volume diameter (D [4,3]) and span values for REV niosomes were studied (Table 4.3 and Figure 4.5). The results showed that the average D [4,3] value for F-DCP was 7.5 ± 0.59 µm, whereas that for F-S60 was 8.30 ± 1.14 µm. Also, the span index for F-DCP (1.68) was less than that for F-S60 (2) indicating a narrow distribution of size and polydispersity (unimodal). Similar span values (1.3 to 2.75) were reported for DCP liposomes and prepared by the REV method (Mehanna et al., 2009). Incorporation of DCP imparted negative charges to niosomes; and the resultant electrostatic repulsion is likely to account for the reduction in the tendency for niosomes to aggregate. 144 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Table 4.3 Effect of selected membrane additives on D [4,3] and span values for niosomes prepared by the REV method (Results are expressed as mean values ± SD, n = 3) REV Formulation F-S60 F-DCP F-C24 F-CH D [4,3] µm 8.30 ± 1.14 7.50 ± 0.59 22.41 ± 1.40 5.37 ± 0.81 Span 2.0 ± 0.15 1.68 ± 0.13 5.55 ± 1.40 1.72 ± 0.05 The D [4,3] calculated for F-CH was 5.37± 0.81 µm, compared with 8.3 ± 1.14 µm for F-S60. The results showed that incorporation of CH in the prepared niosomes significantly decreased (P < 0.01) D [4,3] for F-CH. CH is a highly hydrophilic molecule (HLB 16.70) (El Maghraby et al., 2004). The introduction of such a highly polar molecule should allow niosomes to self-assemble into smaller closed bilayer structures. Similar results have been observed with liposomes where the introduction of CH to the phospholipid bilayer markedly decreased the size of the liposomes formed. Hence, CH is well-known as an edge activator (Cevc et al., 1996; El Maghraby et al., 2004). More interestingly, the D [4,3] for F-C24 was found to be 22.41 ± 1.40 µm and span index was found to be 5.55. This reflects a 2.7-fold increase in F-C24 niosome size compared with FS60 niosomes. C24 has a bulky structure (mol wt 1443) and highly hydrophilic long poly-24oxyethylene chains. The bulky structure and the water molecules associated with the polyoxyehtylene chains could lead to an increase in the cross-sectional area of the head groups and produce giant niosomes and a relatively broad size distribution (bimodal). Similar findings were reported upon incorporation of C24 in hexadecyl diglycerol ether (C16 G2) and Span 40based niosomes (Uchegbu et al., 1992; Vyas et al., 1998). Giant vesicles or discomes, with sizes ranging from 11 to 60 µm, were detected upon incubating the preformed C16 G2- or Span 40based niosomes with C24 solution at 75oC (Uchegbu et al., 1992; Vyas et al., 1998). These giant niosomes are called discomes and coexist with smaller spherical niosomes (2-10 µm) and suggested as a preferable vehicle over conventional spherical niosomes (Uchegbu et al., 1996b). Discomes are more likely to resist nasolacrimal drainage by virtue of larger sizes. Hence, they are thought to offer improved ocular bioavailability than conventional niosomes (Uchegbu & Florence, 1995; Uchegbu et al., 1997; Vyas et al., 1998). 145 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes F-S60 F-DCP F-C24 F-CH Figure 4.5 Size-frequency distribution curves of the prepared niosomes 146 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 4.4.4. Morphology and lamellarity Microstructure and lamellarity were studied using Cryo-SEM, and the results are shown in Figure 4.6 and Figure 4.7. The micrographs revealed spherical and concentric bilayers, arranged in an onion-like organisation, indicating the multilamellarity of F-S60 and F-CH niosomes. On the other hand, F-DCP vesicles were large, spherical and unilamellar (Figure 4.6). These structures are characteristic of vesicles generated by the REV method. Large unilamellar vesicles (LUVs) are likely to self-assemble during the transition from inverted micelles to bilayer vesicles upon the slow evaporation of the organic phase (Kirby & Gregoriadis, 1984b). Incorporating CH and DCP in the bilayer membrane produced spherical niosomes similar to those of F-S60, suggesting that CH and DCP had no obvious effect on the morphology. Incorporation of C24 in the bilayer membrane yielded large or giant vesicles (approx. 22 µm) with oval shapes. These giant vesicles coexisted with smaller spherical vesicles (Figure 4.7). Large elliptical vesicular structures have been produced when incubating the performed Span 40-based niosomes in a C24 solution for 1 h at 75oC. The size of these giant structures ranged from 11 to 60 µm and they had a disc-like appearance; hence the term discomes (Uchegbu et al., 1992). These discomes were reported to be an enhanced and controlled ocular delivery system (Mainardes et al., 2005; Uchegbu & Vyas, 1998). Their inherent large size was thought to prevent discomes from being rapidly washed out by tear dynamics. Also, their non-uniform spherical structure could provide a better fit on the ocular surface (Mainardes et al., 2005; Vyas et al., 1998). For instance, discomes have been found to achieve higher entrapment efficiency and improved the ocular bioavailability of timolol maleate (Vyas et al., 1998). 147 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes F-S60 F-DCP F-S60 F-DCP Figure 4.6 Cryo-SEM micrographs of F-S60 and F-DCP prepared using the REV method 148 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes F-C24 F-CH F-C24 F-CH Figure 4.7 Cryo-SEM micrographs of F-C24 and F-CH prepared using the REVmethod 149 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figures 4.8-4.11 show CLS micrographs of the prepared niosomes loaded with a fluorescent probe (CF). Typical uniform spherical-shaped niosomes were imaged from conventional niosomes (F-S60, Figure 4.8). F-DCP and F-CH were found to be spherical in shape, as shown in Figures 4.9 and 4.11 respectively. These results indicate that the incorporation of DCP and CH into bilayer membranes has no observable effect on niosomes morphology. On the other hand, oval giant vesicles (approx. 20 µm) were imaged for F-C24 niosomes (Figure 4.10). These results were in a good agreement with those obtained from Cryo-SEM micrographs and confirmed discomes formation. As previously mentioned, incorporating such a bulky surfactant (mol wt 1443) with a long ployoxyethylene chain of C24 could influence niosomes geometry, especially at a relatively low level of cholesterol typically ≤ 30% mol/mol (Uchegbu et al., 1992; Uchegbu et al., 1996b). In this study, the discomes were prepared in a single step and under relatively mild conditions. These were formed at 60oC, whereas all the literature on discomes necessitates incubation of the preformed spherical niosomes in a solution of C24 for 1 h at 75oC (Uchegbu et al., 1992; Vyas et al., 1998). This method of preparation could be of value for encapsulating thermolabile watersoluble drugs. 150 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.8 CLS micrographs of F-S60 niosomal formulation loaded with CF produced by the REV method 151 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.9 CLS micrographs of F-DCP niosomal formulation loaded with CF produced by the REVmethod 152 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.10 CLS micrographs of F-C24 discomes loaded with CF produced by the REV method 153 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.11 CLS micrographs of F-CH niosomal formulation loaded with CF produced by the REV method 154 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 4.4.5. DSC studies The DSC studies were exploited in the previous chapter to study the interaction of cholesterol with the bilayer membranes, and to find out the location of entrapped NTX molecules in Span 60 niosomes. Cholesterol exhibited a marked effect in decreasing both Tm and transition enthalpy. This was clearly evident when gradually decreasing both the area under the peaks of DSC thermograms and Tm values. The cholesterol contents in the membrane increased until the complete disappearance of the gel/liquid transition of Span 60 at 50% mol/mol. Additionally, the DSC studies pointed out that NTX molecules are located in the aqueous domains due to the fact that NTX is a hydrophilic. Figure 4.12 Effects of different membrane additives on gel/liquid transition temperatures of the prepared niosomes In this study, DSC was employed to understand further the interaction of the selected membrane additives with the bilayer domains of Span 60 niosomes, and to monitor their effects on the gel/liquid transition temperatures of the prepared niosomes. The residual phase transition of FS60 niosomes (7:3 mol/mol Span 60: cholesterol) was completely abolished upon incorporating the selected membrane additives (Figure 4.12). Such thermal behaviours suggest the inclusion of the additives used into the bilayer membranes of the prepared niosomes. 155 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figures 4.13-4.15 show hypothetical positions of the membrane additives based on DSC data. It is believed that cholesterol is a common membrane stabilizer and can be accommodated by the bilayer membranes. Its steroidal nucleus aligns itself parallel to the hydrocarbon chains of the surfactants and the hydroxyl groups project to the adjacent ester linkages of surfactants’ polar head groups. Cholesterol has a rigid structure, poor water solubility and extreme hydrophobicity. Cholesterol can offer membrane stabilising properties to vesicles; alter vesicle sizes; abolish their phase transition; decrease their membrane permeability; and alter membrane fluidity (Israelachvili et al., 1980; Israelachvili & Mitchell, 1975; Uchegbu & Florence, 1995). The fully extended (all trans) double hydrocarbon chains of DCP align themselves parallel to the hydrocarbon chains of Span 60, and the polar phosphate groups of DCP are parallel to the polar heads the of the sorbitan ester groups. This type of interaction could impart more packing by filling any irregularities running through the bilayer membranes of niosomes. Such improvement in the packing properties could reduce the permeability to the entrapped water-soluble solutes (Figure 4.13). The extremely hydrophobic steroidal nucleus of C24 is more likely to align itself parallel to the accommodating hydrocarbon chains of Span 60 in a manner similar to that of cholesterol molecules in niosomes. The long poly-24-oxyethylene chains align themselves parallel to the sorbitan ester groups (Figure 4.14). Although sodium cholate (CH) has a steroidal nucleus like cholesterol, it is energetically favoured to flip vertically, compared to the orientation of cholesterol in the bilayer membranes. This is due to presence of the carboxylate functional group at the hydrocarbon terminal, and hydroxyl groups attached to the steroidal nucleus. This means the carboxylate group can be accommodated by the sorbitan ester groups and the steroidal nucleus aligns itself parallel to the hydrocarbon chains of the surfactant. Further, due to the high polarity of carboxylate groups and the presence of hydroxyl groups attached to the steroidal nucleus, CH molecules more preferentially protrude into the aqueous environment and consequently, some vacancies in the bilayer membrane could be created (Figure 4.15). 156 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.13 Hypothetical position occupied by DCP in the bilayer membrane of F-DCP based on DSC 157 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.14 Hypothetical position occupied by C24 in the bilayer membrane of F-C24 based on DSC 158 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.15 Hypothetical position occupied by CH in the bilayer membrane of F-CH based on DSC 159 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes 4.4.6. Surface tension measurements Figure 4.16 Surface tension measurements for the prepared niosomes compared with NTX aqueous solution (0.4 mg/ml) (Results are expressed as mean values ± SD, n =3) * Significant difference (P < 0.05), ** insignificant difference (P > 0.05) Figure 4.16 shows the surface tension (γ) measurements for the prepared niosomes compared with NTX aqueous vehicle (0.4 mg/ml in PBS). The γ values measured for the prepared niosomes ranged from 38 to 41 dynes/cm. These results revealed that the γ values for the prepared niosomes were significantly (P < 0.001) lower than that for the aqueous solution (72 dynes/cm). However, there was no significant (P > 0.05) difference amongst the γ values measured for the prepared niosomes. It is well accepted that the lower the surface tension is, the easier it is for the formulations to wet the hydrophobic surface of the corneal epithelium (Pawar & Majumdar, 2006; Rathore & Majumdar, 2006). For example, two gatifloxacin eye drops were prepared with two distinctly 160 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes different surface tension properties and their transcorneal permeation was evaluated. Control eye drops (3 mg/ml gatifloxacin isotonic solution) and optimised formulations composed of 3 mg/ml gatifloxacin isotonic solution containing benzalkonium chloride (BAC) (0.01%) and disodium edetate (EDTA) (0.01%). The amounts of gatifloxacin permeated through excised goat cornea from the control and optimised formulations were 29 and 83 µg respectively. The optimised formulation showed 2.9-fold increase in the amount of the drug permeated. This was attributed to the lower surface tension of the optimised formulations (41.2 dynes/cm) than that (69.5 dynes/cm) of the control. Hence, a better spreading ability and a permeation-enhancing effect were observed with the optimised eye drops (Rathore & Majumdar, 2006). Both BAC (cationic surfactant) and EDTA (chelating agent) are penetration enhancers. Nevertheless, BAC was found to have a more pronounced penetration enhancing effect than EDTA (Rathore & Majumdar, 2006). This in part due to its ability to lower the surface tension compared with EDTA and also due to disruption of corneal epithelium (Rathore & Majumdar, 2006). This report can highlight the role of surface tension of the vehicle in enhancing the corneal penetration via the exclusive corneal epithelium tight junctions. 4.4.7. Contact angle and spreading coefficient measurements The contact angle is a measure of the spreading or wetting of a solid surface by a liquid, adhesion and biocompatibility (Florence & Attwood, 1998c; Lerk et al., 1977). Low values indicate that the liquid spreads or wets well while high values indicate poor wetting. If the angle is less than 90°, the liquid is said to wet the solid. If it is greater than 90° it is said to be non-wetting. A zero contact angle represents complete wetting (Florence & Attwood, 1998c; Lerk et al., 1977). 161 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Table 4.4 Contact angle and spreading coefficient measurements for the prepared niosomes (Results are expressed as mean values ± SD, n=3) Formulation PBS solution F-S60 F-DCP F-C24 F-CH Contact angle (θo) 57.00 ± 1.80* 42.00 ± 1.00 40.00 ± 2.20 43.00 ± 1.81 44.00 ± 2.60 Spreading coefficient ( dynes/cm) -32.79 ± 1.90* -9.79 ± 0.81 -9.00 ± 0.94 -11.20 ± 0.90 -11.65 ±1.26 * The PBS solution was significantly (P < 0.01) different to the prepared formulation, while all the niosomes were insignificantly different from each other (P > 0.05) Table 4.4 shows the contact angle (θ) and the spreading coefficient (S) measurements of the prepared formulations compared with the plain vehicle (PBS). The θ and S values of the prepared formulations ranged from 40o ± 2.20o to 44o ± 1.81o, and -9.00 ± 0.94 to -11.40 ± 0.90 dynes/cm respectively, compared with those for PBS which were 57o ± 1.80o and -32.79 ± 1.90 dynes/cm. The θ values of all tested niosomes were significantly less (P < 0.05) than that of the aqueous vehicle. Additionally, the S values of the prepared formulations were up to 5 times higher than that of the aqueous vehicle. These results suggest that spreading of niosomes on the solid surface is more energetically favoured than rounding itself. These findings suggest that the prepared formulations have better wetting properties than the aqueous vehicle. Hence, the prepared formulations could more promptly and easily wet, spread and adhere to the hydrophobic surface of the cornea than the aqueous vehicle. 4.4.8. Rheological properties measurements Many reports have shown that increasing the viscosity of ophthalmic solutions increases the ocular residence time and the pharmacological effect (Burgalassi et al., 2000; Lang et al., 2002; Ludwig, 2005). However, applying an extensive gel into the eye can blur the vision and make the adjustment of the dose very difficult (Winfield et al., 1990). Increasing the ocular bioavailability due to increasing the viscosity is limited to a plateau, where further increases in the viscosity produces a slight or no increase in the ocular bioavailability. 162 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes This plateau effect has been found to be dependent on drug and formulation type (Lang et al., 2002). It has been reported that an increase in the viscosity range over 1 to 12.5 mPa.s using methylcellulose as a viscosity imparting agent showed a 3-fold decrease in the drainage rate. This decrease in the drainage rate increased the concentration of drug in the precorneal tear film (Chrai & Robinson, 1974). In addition, Isopto® and Liquifilm® are widely used commercial viscous vehicles and consist of 0.5% hydroxypropyl methylcellulose and 1.4% polyvinyl alcohol. The viscosity ranges for these products are 10-30 cps and 4-6 cps respectively, and both have the convenience of being an eye drop (Lang et al., 2002). In this study, the rheological properties of the prepared niosomes were studied at two different temperatures 25oC (ambient temperature) and 35oC (ocular temperature). Data was collected by Reholac software. A direct relationship between F and G at a constant viscosity (η) was obtained for all tested niosomes and PBS which indicates Newtonian flow properties (Figure 4.17). Figure 4.18 shows the η values of PBS solution and the prepared niosomes at 25oC and 35oC. Generally, the η values were higher at 25oC than at 35oC. The effect of temperature on the η values was significant (P < 0.05) in some instances (Table 4.5). For example, the η values for FS60, F-C24 and F-CH were significantly lower (P < 0.05) at 35oC. These results reveal the dependency of viscosity on temperature. Hydrogen bonds, responsible for solvent-solvent or solute-solvent interactions, can be broken by thermal movement at higher temperatures (Martin et al., 1993). 163 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.17 Representative rheograms for the aqueous vehicle (PBS) and the prepared niosomes 35oC 164 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.18 Viscosity values for the aqueous solution (PBS) compared with the prepared niosomes at two different temperatures 25oC and 35oC (Results are expressed as mean values ± SD, n=3) 165 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Table 4.5 Tukey’s pair wise comparison of viscosity values for the prepared niosomes at 25oC and 35oC PBS solution F-S60 F-DCP F-C24 F-CH 25; 35 25; 35 Temperature(oC) 25; 35 25; 35 25; 35 PBS solution NS - - - - F-S60 S;S S S;S - - F-DCP S;S S;S NS - - F-C24 S;S S;S S;S S - F-CH S;S S;S NS ; NS S;S S S = significant difference (P < 0.05) NS = non-significant (P > 0.05) One S or NS indicates the statistical significance difference for one formulation at 25oC and 35oC S;S or NS;NS indicates the statistical significance differences between two formulations at 25oC and 35oC Irrespective of the temperature, all η values for the prepared niosomes were significantly higher (P < 0.01) than the aqueous vehicle. The viscosity range for the prepared niosomes was 1.7 to 8.2 times higher than that for the aqueous vehicle at 25oC. Similarly, the viscosity for the prepared niosomes was found to be 1.7 to 6.3 higher than that for the aqueous vehicle at 35oC. The tested niosomes consisted of different sizes, compositions and lipid contents. The order of the total surfactant/lipid contents of the tested niosomes was as follows F-S60 > F-CH > F-C24 > F-DCP. The highest η values (9.7 and 5.5 cps) were recorded for F-S60 niosomes at both 25oC and 35oC, whereas the lowest η values recorded for the prepared niosomes were 2 and 1.5 cps for F-DCP and F-CH niosomes respectively. F-C24 discomes came in the middle and demonstrated a 2.8fold increase in the η value compared with that for the aqueous vehicle at 25oC. The highest surfactant/lipid content estimated for F-S60 showed the highest η value. Conversely, F-DCP showed the lowest η value, which is mainly due to having the lowest surfactant/lipid 166 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes concentration. These results suggest that the higher the surfactant/lipid concentration, the higher the vesicle-solvent interactions are more likely to occur and consequently the greater the viscosity of the final dispersion. Other factors such as size and composition of niosomes could affect the viscosity of niosomal dispersions. For example, the total surfactant/lipid content for F-CH was approximately double that for F-DCP and the η values for F-DCP and F-CH showed nonsignificant difference (P > 0.05) (Figure 4.18). F-DCP niosomes had a lower concentration but a larger size compared with F-CH niosomes which proves the surfactant/lipid content is not the sole factor determining the viscosity of the niosomal dispersion. Similarly, the η value for F-C24 discomes was significantly higher (P < 0.05) than that for F-CH, although F-CH had a higher content of surfactant/lipid than F-C24. The average diameter for F-C24 discomes was approximately 4 times larger than that for F-CH niosomes. Also, inclusion of C24, with a head group consists of 24 units of highly hydrophilic polyoxyethylene units, in F-C24 discomes increases vesicles-solvent interactions. These findings suggest that many factors such as the lipid content, size and composition of niosomes control the viscosity of the niosomal dispersions. Conclusively, the prepared niosomes showed η values significantly higher than the aqueous vehicle at the same time they have the convenience of being eye drops. It is well accepted that more viscous vehicles have longer precorneal residence times than simple eye drop solutions. This is another potential advantage for the prepared niosomes as ocular delivery vehicles for NTX rather than their spreading and wetting ability. 4.4.9. Effect of niosomal encapsulation on NTX oxidation From the preformulation studies performed in Chapter 2, autoxidation was found to be the main degradation pathway of NTX as indicated from the S-shaped degradation profile, discolouration of NTX solutions stored for forced degradation studies and MS studies. It is well accepted that most liquid-phase autoxidations are free-radical reactions. Free radicals are atoms and molecules that have an unpaired valence electron. Metal ions, light, heat and peroxides are involved as catalysts (Johnson & Gu, 1988; Lachman et al., 1986). Light energy, similar to heat, provides the activation energy necessary for oxidation reactions. Daylight is a primary source for creating free radicals and an initiator of the autoxidation propagation (Florence & Attwood, 1998a). One of the genuine properties a certain drug delivery 167 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes system should have is the ability to offer a chemical protective effect to the drug. Niosomes have been proposed as systems capable of improving the chemical stability of photosensitive drugs such as doxorubicin (Uchegbu & Florence, 1995) and tretinoin (Manconi et al., 2003). The prepared formulations were tested for their ability to protect the encapsulated NTX molecules from the light-induced degradation and the oxidizing agent such as H2O2. Figure 4.19 shows the percentages remaining of NTX for the free PBS solution of NTX and NTX encapsulated niosomes after exposure to daylight and H2O2 (2% v/v). The results showed that 84% and 20% of NTX solution potency was lost after exposure of the free NTX solution to H2O2 (2% v/v) and daylight illumination (10,000 lux) respectively; and the potency of NTX loaded niosomes was also adversely affected by exposure to H2O2 but was significantly (P < 0.05) lower than that of the free drug solution. In terms of protective effects against light-induced degradation (photo-degradation) and apart from F-S60, NTX molecules encapsulated in the prepared niosomes exhibited 24 h protection against the photo-degradation. Figure 4.19 Effect of niosomes encapsulation on the chemical stability of NTX against oxidation and daylight illumination (Results are expressed as mean values ± SD, n=3) 168 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Table 4.6 Tukey’s pair wise comparison of oxidation and photolysis for the prepared niosomes NTX solution F-S60 F-DCP F-C24 Oxid; Phot Oxid; Phot Oxid; Phot Oxid; Phot F-S60 S; NS - - - F-DCP S;S NS; S - - F-C24 S;S NS ; S NS ; NS - F-CH S;S S;S NS ; NS NS ; NS S = significant difference (P < 0.05) NS = non-significant (P > 0.05) Oxid = Oxidation due to H2O2 Phot = Photo-oxidation Longer term (10 days) photo-oxidation profiles of NTX from the PBS solution of NTX and NTX encapsulated niosomes were studied. Figure 4.20 shows semi-logarithmic plots of % NTX remaining versus time. Linear correlations were obtained (R2 > 0.99) for both the PBS solution of NTX and NTX encapsulated niosomes indicating first-order photo-degradation kinetics. Exposure of NTX aqueous solution to the artificial daylight obviously accelerated degradation of NTX. The observed degradation rate constants (kobs) for the aqueous NTX solution stored in the dark and exposed to the daylight illumination were found to be 0.02 and 0.22 day-1 respectively. The calculated kobs value for the aqueous NTX solution exposed to the daylight illumination was 11-fold faster than that of the dark stored sample (Figure 4.21). This indicates the significant impact of the daylight illumination as a potential initiator for autoxidation reactions. The calculated kobs values for the photo-oxidised NTX niosomes were significantly smaller (P < 0.001) than the photo-degraded the PBS solution of NTX. A 1.4-3-fold decrease in kobs was estimated for NTX encapsulated niosomes compared with the free NTX solution. These results could be attributed in part to the ability of niosomes to protect niosomes-encapsulated NTX from peroxide radicals formed by the daylight illumination by virtue of lipid bilayer membranes. Also, 169 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes the bilayer membranes could scavenge the free radicals and prevent them from propagation and slowing autoxidation kinetics (Roda et al., 1998). These findings show the protective effect of the prepared niosomes against photolytic and oxidative degradation of NTX. However, the calculated kobs value for NTX F-S60 was significantly higher (P < 0.001) than all other NTX encapsulated niosomes. The lower protective effect for F-S60 could be ascribed to leakage of NTX molecules from F-S60 due to the presence of a residual gel/liquid transition and its thermo-responsiveness. Abolishment of the gel/liquid transition of the bilayer membranes due to incorporation of the membrane additives would improve the stability of niosomes especially at higher temperatures than the gel/liquid transition temperature (Manconi et al., 2003). These results also suggest that the composition of the bilayer membranes has an influence on the photostability. Figure 4.20 First-order degradation kinetics for PBS solution of NTX and NTX encapsulated niosomes under artificial daylight illumination at 40oC in PBS pH 7.4 (Results are expressed as mean values ± SD, n=3) 170 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Figure 4.21 First-order degradation rate constants for PBS solution of NTX and NTX encapsulated niosomes under artificial daylight illumination at 40oC in PBS pH 7.4 (Results are expressed a mean values ± SD, n = 3) * Significant difference (P<0.05) and no significant difference ** P > 0.05 * Significant difference (P < 0.05), ** non-significant difference (P > 0.05) 171 Chapter 4….Effect of Membrane Additives on the Physical properties of Niosomes Conclusion • Four different niosomal formulations were successfully prepared, encapsulating a considerable amount of NTX. • Both the composition and method of preparation were found to have a significant effect on NTX EE%. • Inclusion of membrane additives (DCP, C24 and CH) to the bilayer membrane generated niosomes with versatile morphology and sizes as confirmed by Cryo-SEM and CLSM. As such discomes were detected with F-C24. • The contact angle and spreading coefficient measurements of niosomes exhibited better wetting and spreading abilities than the aqueous vehicle. • The prepared niosomes were significantly more viscous than the PBS solution. The lipid content, size and composition of niosomes are the main factors affecting the viscosity. • Exposure of NTX solution to artificial daylight illumination can produce extensive degradation of NTX. The prepared formulations were able to significantly protect encapsulated NTX from the photo-degradation compared with free NTX solutions. From the abovementioned key findings, the next and final experimental chapter in the thesis is concerned with evaluating the prepared niosomes for ocular delivery. 172 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5. Evaluation of niosomal formulations for ocular delivery of naltrexone hydrochloride 5.1. Introduction In recent decades, researchers testing ocular dosage forms have recorded toxicological signs of ocular tissues exposed to topically applied drugs. Ocular tissues, such as the cornea and conjunctiva, are susceptible to injuries and adverse ocular effects, either from the administered drug or from excipients used in formulating pharmaceutical products (Basu, 1984; Li et al., 2008). For instance, amphotericin B and ketoconazole (antifungal agents) can cause corneal oedema and corneal abnormalities when administered topically (Foster et al., 1981). Excessive use of topical anaesthetics can produce corneal lesions and ulcers (Basu, 1984). Further, anti-inflammatory corticosteroids are found to retard epithelial corneal wound healing and induce glaucoma (Basu, 1984; Basu et al., 1981; Li et al., 2008). Ocular side effects due to excipients in the pharmaceutical products have also been reported. For instance, benzalkonium chloride (BAC), a quaternary ammonium cationic surfactant, is a commonly used preservative in ophthalmic products. BAC has been reported to cause corneal opacity, a decrease in corneal epithelial microvili, conjunctival hyperaemia (red eye) and delayed wound healing (Li et al., 2008; Pfister & Burstein, 1976). Assessment of the toxicity of ophthalmic formulations and the potential for ocular irritation represents an essential step in the development of new ocular delivery systems (Basu, 1984; Lang et al., 2002). On the regulatory viewpoints, there is relatively little guidance from ICH regulatory expectations for non-clinical toxicity needed for ocular drugs, including those with novel delivery approaches (Avalos et al., 1997; Short, 2008). European regulatory authorities recommend an ocular tolerance study (CPMMP/SWP/21/00), they propose that these studies consist of a single-dose study in a small number of rabbits (1-3), with a drop size of 20 to 30 µl in a single dose administration, along with observation and scoring for any ocular abnormalities (Short, 2008). There are many tests and assays which can be utilised to evaluate the ocular potential of ocular pharmaceuticals. 173 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery A brief review of the different tests used for assessing the ocular toxicity of ophthalmic dosage forms is included below. 5.1.1. Eye irritation tests 5.1.1.1. In vivo ocular irritation test (Draize test) Standard procedures for assessing the eye irritancy potential of a chemical substance are based on the Draize eye irritation test or rabbit eye test (Draize et al., 1944). According to the Draize test, 100 µl of liquid or 100 mg of a solid test material are placed into the lower conjunctival sac of the albino rabbit eye and the eyelids are held shut for a fixed period. Any ocular response is assessed by an observer at various time intervals for a period of up to three weeks after treatment. The contralateral eye is used as an untreated control. The scoring procedures evaluate the irritation response of three ocular tissues, namely the cornea, iris, and conjunctiva. A numerical score was given to different aspects of the irritation observed (Table 5.1). The maximum score is 110, with 80 points possible from the cornea, 10 from the iris and 20 from the conjunctiva. Since damage to the cornea is so critical, its contribution to the scoring is the greatest. Table 5.1 Ocular tissue and score for assessing ocular irritation using the Draize test, adopted from (Draize et al., 1944) Ocular tissue Maximum Cornea score = opacity (0-4) X area (0-4) X 5 80 Iris score = grading value (0-2) X 5 10 Conjunctiva score = [redness (0-3) + oedema (0-4) + discharge (0-3)] X 2 20 Total score 110 Over the years, the Draize test has received both technical and ethical criticism (Anderson & Russell, 1995; Bruner, 1992). Technically, the reproducibility of the test is poor, as scoring and interpretation are subjective, depending on the visual observation of the observer. Also, it tends to over-predict the human response, because it uses a high dose of the test materials and the site of application is at the deep conjunctival sac of the rabbit eye. Ethically, the use of large numbers of 174 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery live animals, the application of large doses of painful and stressful test materials and the length of recovery time are criticised by animal welfare groups (Anderson & Russell, 1995; Bruner, 1992). 5.1.1.2. Modification of Draize test (Low volume eye test) One of drawbacks of the Draize test is the over-prediction of human risk resulting from using a large amount of the test material (100 µl for liquid or 100 mg for solid). A modification of the test is carried out; this uses only 10 µl of the test materials, which is applied directly to the cornea instead of the lower conjunctival sac, and the eyelids are not held shut. This procedure gives rise to a procedure called the low volume eye test (LVET) (Griffith et al., 1980). The applied volume and the site of application closely mimic accidental human exposures (Avalos et al., 1997; Bruner, 1992). 5.1.1.3. In vitro ocular toxicity models Over the past three decades, there has been growing need in developing an in vitro alternative to in vivo animal studies for ocular toxicity by ethics committees and animal welfare organisations, to provide more objective and reproducible results, avoid the use of animals and end animal suffering (Bruner, 1992). Potential in vitro alternatives include (Anderson & Russell, 1995; Bruton et al., 1981; Curren & Harbell, 1998): • Cytotoxicity tests • Chorioallantoic membrane of the chicken eggs • Isolated enucleation eyes • Combination of chorioallantoic membrane and isolated enucleation eyes These in vitro tests can offer many advantages over the conventional in vivo tests. These include reducing the number of animals involved, and using more quantifiable and objective end-point measurements. These tests are also more convenient and less time-consuming (Avalos et al., 1997). Cytotoxicity tests The 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) test is a colorimetric reaction that relies on the ability of viable cells to reduce yellow tetrazoliums to purple formazans, a reaction catalysed by the mitochondrial succinate dehydrogenase of the viable cells. 175 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery The MTT assay is also subject to some limitations, including the chemical reduction of tetrazolium by the test agents. Also, chemical interference occasionally occurs due to the cellular reduction of tetrazolium. There is also a variety of metabolic conditions (e.g. depletion of glucose and variations in pH) which might affect the production of formazan from tetrazolium (Rubinstein et al., 1990; Skehan et al., 1990). Another rapid and sensitive cytotoxicity test relies on the interaction of sulforhodamine B (SRB) with the viable cellular protein content. The SRB assay shows good linearity with the number of cells (Rubinstein et al., 1990). The SRB assay offers major practical advantages over the MTT assay. Notably, the MTT assay involves a timed step, in which cells are incubated with tetrazolium and the number of viable cells is estimated on the basis of the amount of formazan produced. When many hundreds of 96-well microtiter plates are handled daily in the laboratory, scheduling this step and subsequently reading the plates becomes difficult. In the SRB assay, cells in the culture wells are chemically fixed at the end of the assay, allowing simultaneous staining and further processing of extremely large batches of plates without time-critical steps (Rubinstein et al., 1990). There are some limitations to the use of cytotoxicity testing of water-insoluble solids or oilybased formulations (e.g. w/o emulsions), since such test materials do not come into direct contact with the aqueous cell culture media. Also, cell culture media lack the protective shield of the precorneal tear film, and as a result the test materials are directly exposed to the test tissue. Other drawbacks include a possible interaction of the test materials (drug or formulation additives) with the components of the growth medium for the cell line, producing false results. The chorioallantoic membrane (CAM) test The CAM is the vascularised respiratory membrane that surrounds an embryonic bird developing in the egg. It is a complete tissue including veins, arteries and capillaries. This test was initially developed as a possible model for predicting the irritant effect of the chemical on the conjunctiva (Leighton et al., 1985). The test material is applied directly to the membrane surface, and then incubated for a certain period, depending on the protocol used. The membrane is inspected visually and changes in morphology (size of the lesion and presence of necrosis or capillary haemorrhage) are scored (Bruner, 1992). The CAM test has been used to test the toxicity and irritancy of some surfactants commonly used in pharmaceutical industries. The test can successfully discriminate between the test materials according to their irritancy potential (Bagley 176 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery et al., 1991a, 1991b). It has also been found to correlate well with rankings of the same preparation from the in vivo rabbit eye test (Draize test) (Leighton et al., 1985). A modification of the CAM test, called the Hen’s egg test-CAM (HET-CAM), was developed by (Luepke, 1985). The HET-CAM is a rapid, sensitive and inexpensive test, whereby fresh fertilised viable eggs are incubated at 37oC ± 0.5 and relative humidity of 62.5 ± 7.5%. On day 10 of incubation, the egg’s shell is scratched around the air cell and the inner egg membrane is removed. The test substance is instilled onto the membrane, and after 20 seconds the membrane is irrigated with warm water. The membrane is then scored for irritant effects including hyperaemia, haemorrhage, and coagulation at 0.5, 2 and 5 min. A variety of household materials were tested including pyrithiones, phenols and isothiozolinones. The HET-CAM showed a good correlation with the ranking of the same preparation tested using the Draize test (Luepke, 1985). In addition, the HET-CAM has achieved good discrimination and sensitivity to the irritation potential for pharmaceutical excipients and agricultural pesticides (Alany et al., 2006; Budai et al., 2010). For instance, several formulation additives have been tested for ocular irritancy using the HET-CAM test, including ethyl oleate, sorbitan mono-laurate, polyoxyethylene 20 sorbitan mono-oleate, 1-propanol, 1-butanol, 1-hexanol, 1,2 propanediol, and 1,2 hexandiol. The test successfully discriminated between the individual test materials in a homologous series. For example, ocular irritation of n-alcohols showed that there was a stronger irritation potential above a critical hydrocarbon chain length (2–3 carbon atoms). Moreover, the introduction of a second hydroxyl group caused a reduction of irritation only for 1,2-propanediol compared with 1propanol (Alany et al., 2006). The HET-CAM offers greater advantages than the Draize test. These advantages include reduced animal suffering, less subjective scoring and more reproducible scoring systems. The HET-CAM is also more rapid, less expensive (no animals or animal housing) and uses simpler methods (Alany et al., 2006; Bagley et al., 1991a, 1991b; Budai et al., 2010; Leighton et al., 1985). Isolated enucleated eyes Isolated enucleated eyes (IEEs) from bovines (Muir, 1985), pigs (Bruton et al., 1981), chickens (Bruton et al., 1981) and rabbits (Scaife, 1985) were fortunately used as potential alternatives tests. The results obtained from the IEE test correlated well with in vivo data. IEEs do not require the use of live animals but they use animal by-products and ready to use animal tissues. More advantageously, IEE tests do not necessitate animal and human ethics’ approval and as such they 177 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery save the time required for that. The main argument against the acceptance of the use of isolated rabbit eyes is that there are no rabbit slaughterhouses to supply rabbit meat for human consumption. Therefore, laboratory rabbits are still needed as eye donors (Prinsen & Koeter, 1993). The bovine corneal opacity and permeability (BCOP) assay is a good example for IEE assay. The BCOP assay involves an assessment of corneal opacity and disruption of the corneal barrier (as assessed by the passage of a fluorescent dye) after exposure to the test material. This test has been widely utilised as an in vitro alternative to the in vivo test (Draize test). The BCOP assay is currently applied across cosmetic and pharmaceutical industries in order to evaluate the ocular irritation potential of surfactants, pharmaceutical intermediates and final products (Cooper et al., 2001; Gautheron et al., 1992; Muir, 1985; Vanparys et al., 1993). Combined in vitro assay There is general agreement that the HET-CAM test serves as a possible model for conjunctival irritation testing, as it responds to irritant substances with an inflammatory reaction similar to that produced by conjunctival tissue (Alany et al., 2006; Anderson & Russell, 1995; Bruner, 1992). However, good eyesight relies on the cornea as a refractive component. The cornea serves as the gateway to the eye for external images. The transparency and smoothness of the cornea is essential to maintaining its refractive and protective functions (Nishida, 2005). Consideration must be given to the safety of the corneal tissue when using the developed ocular formulation. It is not surprising that both corneal and conjunctival damage together constitute 100 out of 110 possible points when scoring the Draize test. Therefore, it has been found more advantageous to develop an in vitro alternative model to investigate the safety of the test material on both the cornea and the conjunctiva (Weterings & Vanerp, 1987). Combined isolated enucleated eyes and HET-CAM models for the cornea and conjunctiva respectively were developed. These combined tests were evaluated against the effects observed in the rabbit test. A good correlation was obtained with a broad range of chemical substances (Weterings & Vanerp, 1987). Moreover, this method is easy to perform, inexpensive, reproducible and uses less subjective scores than the in vivo rabbit test (Avalos et al., 1997; Budai et al., 2010; Weterings & Vanerp, 1987). However, predicting ocular irritation using opacity and permeability endpoints is challenging when the test substances produce a delayed reaction by interacting with nucleic acids and mitochondrial proteins, rather than causing an immediate loss of epithelial integrity. Therefore, 178 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery histological examination of the cornea after treatment with the test substances can provide comprehensive assessment of the depth of injury and cellular damage of the three principle layers of the cornea (Curren et al., 2000; Curren & Harbell, 1998). In this thesis, combined HET-CAM and bovine eye tests are conducted to study the irritation potential of the test substances used and the prepared niosomes, followed by histological evaluation of the excised bovine cornea. 5.1.2. Toxicity of niosomes There are limited research reports on the toxicity of niosomes. Different niosomal formulations were evaluated for inhibiting the cell proliferation of SV-40 human keratinocytes (Hofland et al., 1991). The effects of molecular structure of the surfactant and the incorporation of cholesterol on the cell proliferation were investigated. The surfactants were composed of different hydrocarbon chain lengths (12, 14, 17 and 18) and polyoxethylene chain lengths (3 and 7). Also, ether- and ester-type surfactants were examined. It was found that both the hydrocarbon chain length and the polyoxyethylene chain length had minor effects on cell proliferation. However, the bond (an ether or ester bond) by which the alkyl chain was linked to the polyoxyethylene head group had a strong effect on the cell proliferation. The concentration of the ether type-surfactants that inhibited cell proliferation by 50% was 10 times lower than for the ester-type surfactants. The cholesterol content of the bilayers has not been found to have an effect on cell proliferation. The low toxicity of ester-type surfactants compared with ether-type surfactants could be attributed to enzymatic degradation of ester bonds (Hofland et al., 1992). In some instances, encapsulation of some drugs in niosomes reduces the toxicity. For example, niosomes encapsulating vincristine decreased the neurological symptoms, diarrhea and alopecia following intravenous administration of vincristine (Parthasarathi et al., 1994). On the other hand, evaluating the physical stability of niosomes with respect to changes in niosome size and drug retention ability should be carried out at an early stage of formulation development. Therefore, the tendency of niosomes to aggregate and leakage of entrapped therapeutic molecules over time have been investigated (Guinedi et al., 2005; Manosroi et al., 2003; Uchegbu & Vyas, 1998). 179 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5.1.3. Physical stability of niosomes One of the most important features of niosomes compared with liposomes is their chemical stability. Niosomes are less vulnerable to chemical degradation or oxidation than phospholipids (Baillie et al., 1985; Zuidam et al., 2003). However, both liposomes and niosomes can equally show physical stability problems. Vesicle aggregation, bilayer fusion and drug leakage are the most common physical changes exhibited. Physically stable niosomal dispersions must exhibit a constant particle size and a constant level of entrapped drug over time. It has been shown that niosomes are also more physically stable than liposomes (Mukherjee et al., 2007; Nasr et al., 2008). For example, aceclofenac Span 60 niosomes showed higher physical stability (in terms of particle aggregation and drug leakage) than liposomes when stored for three months at 4oC (Nasr et al., 2008). Also, both reconstituted and lyophilised Span 20-based niosomes were more physically stable than liposomes when stored at 4 and 20oC for 3 months (Mukherjee et al., 2007). Additionally, Tween 61 niosomes with 25% ethanol and Span 60 niosomes with 20% ethanol (elastic niosomes) exhibited no sedimentation, no layer separation and an unchanged particle size at 4, 27 and 45oC within three months (Manosroi et al., 2008b). It is worth mentioning that the main factors determining the stability of niosomes include surfactant/lipid type, original size and temperature of storage. First, the choice of surfactant determines the fluidity of the bilayer membrane. It has been found the higher the gel-liquid transition temperature, the higher the drug-retentive properties of niosomes. This is because the bilayer membranes are always in a gel state at the storage conditions. The likelihood of carboxyfluorescein-loaded Span niosomes leaking was found to be in the following order: Span 60 > Span 40 > Span 20 > Span 80 (Yoshioka et al., 1994). Also, the incorporation of cholesterol (membrane stabiliser) in niosomes improves their physical stability (Uchegbu & Vyas, 1998). Secondly, niosomes sized between 1 and 10 µm have been found to be more stable than those in the sub-micron range. Thermodynamically, smaller niosomes have higher surface free energy and tend to aggregate more than larger ones in order to lower the excess free energy (Moazeni et al., 2010). For example, niosomes composed of hexadecyl diglycerol: cholesterol: Solulan C24 at a 40:40:10 molar ratio with an original size of 70 nm were monitored for changes in size. The size of the niosomes increased 250-fold. This result demonstrated that the original size has a significant effect on the stability of niosomes (Uchegbu & Vyas, 1998). 180 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Thirdly, the storage temperature for niosomes must be controlled, as a change in temperature often leads to a change in the nature of the surfactant bilayer membranes (Tanaka, 1990; Uchegbu & Vyas, 1998). In this context, Span 40 niosomes have been found to be more stable than liposomes at higher storage temperatures such as 25 and 37oC since Span 40 has a higher transition temperature than phospholipids (Ning et al., 2006). 5.2. Chapter aims This chapter aims to evaluate the prepared niosomes for ocular delivery of NTX. The following specific objectives are considered: • Investigating the ocular irritation potential of the surfactants and additives used, and the prepared niosomes at different concentrations (1-10% w/v of total lipid). This involves the use of the HET-CAM and bovine eye tests to study conjunctival and corneal irritation potential respectively. • Examining the extent of corneal tissue damage from the prepared niosomes via histological sectioning and examination of the bovine corneas after exposure to the test substances at different time points. • Exploring the in vitro release profiles and release kinetics of NTX from the prepared niosomes. • Studying the ex-vivo permeation of NTX from a simple NTX solution and from the prepared niosomal formulations using excised cow corneas. • Investigating the physical stability of the prepared niosomes through monitoring changes in particle size and NTX EE% over 90 days at different temperatures. 181 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5.3. Materials and methods 5.3.1. Materials NTX, solvents and buffer salts were purchased from the same sources mentioned in Chapter 3 and Chapter 4. Fertilised Shaver Brown hen’s eggs were purchased from Bromley Park Hatcheries Ltd, Tuakau, New Zealand; whole cow eyes were collected from Auckland Meat Processors, Auckland, New Zealand and were used on the same day; cellulose membrane, molecular weight cut-off 12,000-14,000, was purchased from Sigma-Aldrich, St. Louis, USA. 5.3.2. Methods 5.3.2.1. Ocular irritation studies Conjunctival (modified HET-CAM) test Preparing and growing the CAMs Freshly collected fertilised hen’s eggs (Shaver brown) were incubated at 37.5oC ± 0.5oC and 66 ± 5% relative humidity (RH) for 3 days according to the HET-CAM method (Luepke, 1985). The eggs were incubated horizontally in growing trays and gently rotated every day to ensure that the embryo was properly positioned. On days 3 and 4, the eggshells were sterilised by spraying them with a rectified spirit and opened by cracking the underside of the egg against the edge of a plastic Petri dish. The egg was then poured into a growing chamber, according to the modified HET-CAM method (Alany et al., 2006). The growing chamber was made of a cylindrical plastic holder, to which a piece of cellophane membrane (commercial Glad Wrap™) was attached to make a pouch. The cellophane pouch was fixed using a plastic circular sleeve. Once in the growing chamber, the egg was examined to test the viability of the embryo (intact CAM and yolk) (Figure 5.1). Only viable embryos with an intact CAM and yolk sac were further incubated at 37.5oC ± 0.5oC and 66% ± 5% of relative humidity (RH). These were the optimum conditions for CAM growing (Luepke, 1985). The temperature and RH were recorded every hour using a RH/Temp data logger (EL-USB-2, UK) operated by Easylog software. Figure 5.2 shows a 10-day recording of the temperature and RH% inside the in-house modified egg incubator. 182 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Three day old dead embryo with intact yolk sac Three day old embryo with broken yolk and CAM CAM Three day old embryo with intact yolk and CAM Ten day old embryo with viable CAM covering the entire surface Figure 5.1 Development stages of the growing embryos 183 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Figure 5.2 Temperature, percentage relative humidity (% RH) and dew point inside the egg incubator over the 10-day incubation period 184 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Test substances and noisomal formulations The ocular irritation potential of the following test substances was investigated: • NTX solution (40 mg/ml) in PBS pH 7.4. • Span 60 and cholesterol, DCP, C24 and CH in their pure (powder) forms. • Surfactant solutions of 0.3% w/v CH and 0.4% C24 solutions in PBS representing the actual concentrations of the surfactant additives in the proposed niosomal formulations. • The plain niosomal formulations (F-S60, F-DCP, F-C24 and F-CH) at three different total surfactant/lipid concentrations (1, 5 and 10 % w/v). • NTX loaded F-S60, F-DCP, F-C24 and F-CH niosomal formulations. Irritation testing, scoring and classification On day 10, 0.2 ml (for liquid test materials) or 0.1 g (for solid test materials) of the test substances were dropped onto the membrane. For each test substance three eggs were used. NaOH (0.5 M) was used as a positive control for a strong irritant, acetone as a moderate irritant, propylene glycol as a slight irritant, and normal saline as a negative control (Alany et al., 2006). After the application of the test substance, the blood vessels and capillaries were examined for irritant effects. The irritant effects were hyperaemia, haemorrhage and clotting at different time points post-application for 5 min (Figure 5.3). A time-dependent numerical score was allocated to each test substance or formulation (Table 5.2). The sum of the time-dependent numerical scores for all three irritant responses gave a single numerical value. This value interpreted the irritation potential of the test substance (Table 5.3). The mean score value of the test allowed the assessment by a classification scheme analogous to the well-known Draize test (Luepke, 1985). 185 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Table 5.2 HET-CAM scoring system Score Effect time (min) 0.5 2 5 Hyperemia 5 3 1 Haemorrhage 7 5 3 Clotting/coagulation 9 7 5 Table 5.3 Classification of cumulative scores in HET-CAM Cumulative score Interpretation 0-0.9 None 1.0-4.9 Slight 5.0-8.9 Moderate 9.0-21.0 Severe 186 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery No response Hyperaemia Haemorrhage Clotting Figure 5.3 Vascular responses used to score the test substances 187 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Corneal (Bovine eye) test Biological materials Cow eyes were obtained from a local slaughterhouse (Auckland Meat Processors, Auckland, New Zealand). The eyes were excised by standard procedures within 10 min of the animal being killed. The collected eyes were transported to the laboratory in cold saline (8-10oC). The eyes were examined for epithelium detachment, corneal opacity and corneal vascularisation. Eyes with corneal damage or abnormalities were discarded. A summary of the average dimensions of the cow corneas, compared with human corneas is presented in Table 5.4. Table 5.4 Average dimensions of human and cow corneas Species Cornea Cornea Cornea diameter surface area thickness mm mm2 mm Human 11 95 0.52 Cow 20 314 1 Data collected from (Worakul & Robinson, 1997) and own measurements Test substances and niosomal formulations Four different controls were used for validation purposes; 0.5 M NaOH was used as a strong irritant control, acetone as a moderate irritant, propylene glycol as a slight irritant, and normal saline as a negative control. The same components and formulations described in the HET-CAM test were also investigated for their corneal irritation potential. Irritation testing, scoring and classification Small plastic cups were used to hold the eyes (cornea upwards), and the whole kept in the humid atmosphere of a closed water bath at 37oC ± 0.5oC for 10 min (Weterings & Vanerp, 1987). A silicon O-ring (thickness 1.78 mm, an internal diameter 7.6 mm) was carefully placed on the central part of the cornea, to localise the application site and for easy and reproducible test material application (Figure 5.4). One drop of saline was applied inside the ring and the eyes were equilibrated in a closed water bath for 5 min. The test substance was applied to the cornea inside the ring at a volume of 0.1 ml or 0.1 g of the solid test substance. After 30 s, the eyes were rinsed with saline (approximately 10 ml), followed by further incubation in the closed water bath 188 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery for 10 min. Then, the extent of corneal injury was assessed by evaluating the opacity, followed by application of sodium fluorescein solution (2% w/v pH 7.4) to examine the integrity of the corneal epithelium, using an examination lamp and cobalt blue filter (Leica, GmbH, Germany). The observations were graded according to individual numerical scores for opacity, epithelial integrity (degree of staining) and epithelial detachment (Weterings & Vanerp, 1987). Corneal injuries and scores are demonstrated in Figure 5.5 and Figure 5.6 and Table 5.5. The sum score was calculated and the mean scores for each of the 3 exposed eyes was used to interpret the corneal irritation potential (Table 5.6). Figure 5.4 Excised cow eyes immersed in normal saline with silicone O-rings centered on top of the cornea and incubated a water bath thermostatically equilibrated at 37oC ± 0.5oC 189 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Table 5.5 Bovine eye scoring system Opacity Score Epithelial integrity Score Epithelial detachment Score None 0 None 0 No gross abnormalities 0 Slight 1 Diffuse and weak 0.5 Wrinkling of corneal surface 2 Marked 2 Confluent and weak 1 Loosening of epithelium 3 Severe 3 Confluent and intense 1.5 Epithelium absent 4 Opaque 4 - - - - Table 5.6 Classification of cumulative scores in bovine eye test Cumulative score Interpretation ≤0.5 None 0.6-1.9 Slight 2.0-4.0 Moderate >4 Severe 190 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Transparent cornea (none) Opaque cornea No staining (none) Complete staining and loss of lustre (no epithelium) Figure 5.5 Degree of corneal opacity (left) and fluorescein permeability (right) used to score the test substances [non-irritant (upper) and strong irritant (lower) models] 191 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Marked opacity Diffuse and weak opacity Confluent and intense staining Confluent and weak staining Figure 5.6 Degree of corneal opacity (left) and fluorescein permeability (right) used to score the test substances [moderately irritant (upper) and slightly irritant (lower) models] 192 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Histopathological evaluation of the bovine corneas The treated areas of the cornea in the previous bovine eye test and bovine eyes treated with the prepared niosomes for longer periods of time (1, 3 and 8 h) were histologically studied. The samples were dissected and fixed in 10% v/v neutral buffered formalin (NBF) for at least 24 h, and then transferred to 70% v/v ethanol for another 24 h. The fixed samples were transferred into labelled cassettes and placed in 70% v/v ethanol (Zagon et al., 2006a). They were then sent to the histology laboratory for tissue processing, embedding, sectioning and staining. Each section was paraffin-embedded, bisected into two equal halves and finally mounted in a paraffin block so that a section of each half cut and placed on a single microscopic slide. The slides were stained with haematoxylin and eosin (H&E). The stained corneal sections were imaged using a light microscope (Leica DMRE, GmbH, Germany) and the thickness of stroma was measured using Leica DMRE software. 5.3.2.2. In vitro release studies In vitro release studies were performed using standard Franz diffusion cells. The Franz–diffusion cells had a 15 mm diameter orifice (providing a diffusion area of 1.7 cm2) and were thermostated by means of a water jacket connected to a VTC-220 heat circulator (Logan Instrument Corporation, Somerset, NJ, USA). Receptor chambers (12 ml volume) were filled with PBS pH 7.4 (osmolality 297 mOSm/kg) and stirred constantly using small magnetic bars. Donor and receptor chambers were separated by means of a 12,000-14,000 mol wt cut-off dialysis membrane, presoaked in the receptor medium overnight prior to the experiment. The temperature was set at 35oC ± 0.5ºC. Two ml of each formulation were loaded into the donor compartment before occluding the chamber with Parafilm. Care was taken to avoid air bubbles at the membrane/liquid interface. Samples (0.4 ml each) were withdrawn at predetermined time points for up to 12 h, and replaced with an equal volume of the receptor medium. The experiments were done in triplicate and the samples were determined using the HPLC method as described in Section 2.3.2. 5.3.2.3. In vitro release data analysis Three in vitro release parameters were determined for comparing NTX release profiles from the prepared niosomes. • Q2h: the percentage drug release per unit surface area at 2 h. • Q6h: the percentage drug release per unit surface area at 6 h. • DE %-12: the percentage dissolution efficiency at 12 h (equation 5.1). 193 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 12 DE % − 12 = ∫ y.dt 0 1200 .100% equation 5.1 12 The DE%-12 was defined as the area under the % release-time curve ( ∫ y.dt ) at 12 h, expressed 0 as a percentage of the area of the rectangle described by 100% release at the same time which is 12 1200 (Khan, 1975). The ∫ y.dt was estimated using the trapezoidal method. 0 5.3.2.4. In vitro release kinetic studies To elucidate the mechanism of drug release from the prepared niosomes, the release data was put through equation 5.2 which is widely known as the Korsmeyer-Peppas equation (Korsmeyer et al., 1983). Mt = kt n M∞ equation 5.2 Mt/M∞, t, k and n are the fraction of drug release at time t, the kinetic constant, release exponent respectively (Peppas, 1985). The release exponent (n) indicates the general operating release mechanism; an n value of 0.43 indicates Fickian diffusion-mediated release; 0.43 < n < 0.85 indicates non-Fickian release or coupled diffusion and erosion; and n = 0.85 indicates erosionmediated release (zero-order kinetics) (Peppas, 1985; Ritger & Peppas, 1987). The release exponent n is the slope of the log fraction of the drug release (Mt/M∞) versus log time plot. 5.3.2.5. Ex-vivo permeation studies Enucleated bovine eyes were carefully examined for any corneal damage such as epithelium detachment, corneal opacity and corneal vascularisation, before the corneas were dissected. The dissection was performed with extreme care to avoid touching the surface of the cornea. Therefore, a small ring of sclera tissue was left around the cornea (Figure 5.7). The scleral tissue served as a gasket and permitted the cornea to be mounted where the endothelium was facing the receptor compartment and the epithelium was facing the donor compartment. Franz-diffusion cells were used for ex-vivo study. A modification was made in the donor and receptor compartments to hold cow corneas and maintain normal corneal curvature without 194 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery wrinkling (Figure 5.7). The receptor compartment was concave and the donor component was convex providing a diffusion area of 1.7 cm2. The temperature was kept at 35oC ± 0.5oC. The receptor chambers were filled with a modified ringer’s solution to preserve the integrity of the excised cornea over 8 h (Schoenwald & Huang, 1983). The permeation medium comprised sodium chloride (122 mM), sodium bicarbonate (25 mM), magnesium sulfate (1.2 mM), dibasic sodium phosphate (0.5 mM), calcium chloride (1.4 mM), glucose (10 mM) and 4-(2hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES, 10 mM). The medium’s pH and osmolality were 7.4 and 299 mOSm/kg respectively. The medium was constantly stirred using small magnetic bars. A two ml-sample of each formulation was introduced into the donor compartment. Samples of 0.4 ml were withdrawn at predetermined time points for up to 12 h and replaced with fresh receptor medium. The experiments were performed under non-occlusive condition, to allow air permeation through the corneal tissues (n = 5). The samples were analysed as previously mentioned in Section 2.3.2. Figure 5.7 Excised bovine cornea, modified parts and top view of the final assembly of the Franz- diffusion cell (from left to right) 5.3.2.6. Ex-vivo data analysis The apparent permeability coefficient (Papp, cm/s) was calculated using equation 5.3 (Schoenwald & Huang, 1983). 195 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Papp = ∆Q ∆t (3600 )ACo equation 5.3 ∆Q/∆t is the permeability rate of NTX across the excised cow corneas. It was calculated from the slope of the straight line obtained from the plot of the amount of NTX permeated versus time curve. Co is the initial drug concentration (µg/cm3), A is the corneal surface area (cm2) and 3600 allows for the conversion of hours to seconds. The lag time (tL) was determined by extrapolating the steady-state line to the time axis. 5.3.2.7. Physical stability of the prepared niosomes Four different batches of F-S60, F-DCP, F-C24 and F-CH were prepared, purged with nitrogen gas and stored in screw-capped 20-ml glass vials at three different temperatures (4oC, 25oC and 37oC) for three months. Samples were withdrawn at specified time intervals and evaluated for: • Changes in size by measuring the average D [3,4] as previously mentioned in size measurements at Section 4.3.2.2. • The amount of NTX retained in the niosomes, by measuring the residual EE% as previously described in EE% measurements at Section 3.3.2.2. 5.3.3. Statistical analysis A one-way analysis of variance (ANOVA) followed by Tukey’s pairwise comparisons at the 5% significance level was used to test statistical significance differences between the prepared formulations for in vitro release studies, kinetic studies, ex vivo studies and physical stability studies. These were performed using GraphPad Software Version 3.05, San Diego California, USA. 196 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5.4. Results and discussion 5.4.1. Conjunctival (HET-CAM) test During the cracking stage of the CAM preparation process, about 30% were excluded for several reasons, some of which were beyond experimental control like eggs had dead embryos (Figure 5.1) and infertile eggs. In addition, the technique of cracking the shell against the side of the growing chamber did not allow adequate control of breaking and in some occasions resulted in damaging of the young CAM and/or the yolk membrane (Figure 5.1). In some instances, the damage of yolk membrane was in the form of small leaks that were not detected initially and showed up after 24-72 h. These findings suggest the need for an alternative technique for the eggshell breaking with better control of the breaking process. This would reduce the chances of damaging the embryo and extra-embryonic membranes and thus increasing the number of embryos incubated. There are many methods of opening the eggs reported in the literature (Alany et al., 2006; Auerbach et al., 1974; Luepke, 1985). According to the original method, the CAMs were fully grown in the eggshells until test day. On day 10 of incubation, a window was created in the eggshell using a dentist rotary saw-blade and the inner egg membrane was removed to expose the CAM for applying test substances (Luepke, 1985). However, exposing the CAM by this method was found to damage the CAM occasionally when eggshell particles fell onto the CAM surface during the drilling procedure. Moreover, the window created in the eggshell is relatively small and hence gives limited access to and visibility of the CAM. This impedes the evaluation of vascular responses and minimises the total area of CAM surface available for testing (Lawrence et al., 1986). A modified method was reported in the literature, where the chick embryo was grown in a Petri dish from day 3 onwards to allow ready access to the entire CAM surface for better visibility and convenience (Auerbach et al., 1974). This method was slightly modified for the studies performed in this thesis using in-house made growing chambers to replace the Petri dishes. Accordingly, the growing chambers can offer the embryo a more natural environment to grow. The curved, elastic nature of the cellophane membrane mimics the inner membrane of the eggshell compared with the flat, rigid surface of the Petri dish. This could explain the higher percentage of embryos surviving (60%) compared with that obtained when using a Petri dish to grow the embryos (50%). These findings suggest that the cracking procedure and growing 197 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery conditions are critical to the survival of the embryos and therefore the number of CAMs available for test substances. Figure 5.8 shows the cumulative HET-CAM scores for the controls, NTX solution (40 mg/ml) in PBS, the test substances used in fabricating the niosomal formulations and the prepared niosomes. Strong irritant Moderate irritant Slight irritant Non irritant Figure 5.8 Cumulative HET-CAM scores for the controls and representative examples of the test substances (Results expressed as mean values ± SD, n = 3) 198 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery The average cumulative scores calculated for NTX solution, Span 60 powder, cholesterol powder and DCP powder were found to be < 0.9. These results reveal that these test substances are practically non-irritant when applied to the surface of the CAM. Span 60 and cholesterol are regarded as non-irritant and are widely used in the cosmetic and food industry (Buehler, 2003; Lawrence, 2003). In contrast, CH powder was found to be strongly irritant with a cumulative score of 14 ± 1.4. This result is in accordance with the in vivo rectal irritation in rats. A solution of 25 mM (equivalent to 1.1 % w/v) CH was found to be strongly irritant to rectal mucosa of rats. Immediate strong contractions were observed after application into the rectum. Lumen congestion, oedema and haemorrhage were also detected after 20 min (Kinouchi et al., 1996). All the prepared niosomal formulations, total lipid concentrations up to 10 % w/v, were assessed for the irritation potential and the results did not show any signs of inflammatory reactions at the nominated time points. Similar results have been reported on niosomes assessed using the in vivo rabbit eye test (Draize test) (Abdelbary & El-gendy, 2008). Niosome formulations composed of Tween 60:cholesterol:DCP at 1:1:0.1 molar ratio, Tween 80: cholesterol at 1:1 molar ratio and Brij 35:cholesterol: DCP at 1:1:0.1 molar ratio did not show any signs of redness, inflammation or increased tear production when applied onto the eyes of albino rabbits (Abdelbary & El-gendy, 2008). C24 powder developed minimal irritation potential such as hyperaemia after 3 min. This indicates that C24 is slightly irritant when applied on the surface of the CAM. Further, no signs of vascular toxicity were detected after application of niosomes containing sodium cholate (F-CH). The concentrations of CH in the tested F-CH niosomes were 0.13% to 0.53% w/v. It is likely that the amount of CH in these formulations was insufficient to produce an irritant effect. Rabbit corneas demonstrated good ocular tolerability and no signs of ocular damage after the application a 0.5 % w/v CH solution onto rabbit eyes (Furrer et al., 2002b). Another possible explanation is that CH molecules were included in the bilayer membrane of the prepared niosomes, which is likely to reduce the direct contact of free CH molecules with the ocular surface. 199 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5.4.2. Cornea (bovine eye) test Figure 5.9 shows the cumulative bovine eye test scores for the controls, NTX solution (40 mg/ml) in PBS, the test substances used in manufacturing the prepared niosomal formulations and the prepared niosomes. The cornea has both refractive and protective functions. The normal cornea is transparent, lustrous and completely impermeable to fluorescein dye, due to the exclusive tight junctions of the corneal epithelium. Figure 5.5 shows photographs for normal cornea before and after staining. Apart from CH powder, all the test substances and the prepared formulations did not show any signs of corneal injuries (corneal opacity, corneal permeability (stained green) or epithelial damage), and the cumulative corneal scores were < 0.5. Additionally, all the prepared niosomal formulations in the three concentrations tested showed no signs of corneal damage when applied to the corneas. Moreover, the lustre of the niosome-tested corneas was as normal as the negative control corneas. Similar results were reported on the in vivo rabbit eye test for niosomes (Guinedi et al., 2005). Niosomes composed of Span 60: cholesterol 7:4 mol/mol were applied topically on rabbit eye test daily for 40 days. Rabbits’ corneas were excised and assessed histologically for corneal irritation. The tested corneas showed slight stromal oedema and no major harmful corneal signs, and as such the tested niosomes were regarded as safe for short and long term treatment (Guinedi et al., 2005). 200 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Strong irritant Moderate irritant Slight irritant Non irritant Figure 5.9 Cumulative bovine eye scores for the controls and representative examples of the test substances (Results expressed as mean values ± SD, n = 3) Table 5.7 shows the interpretations of the cumulative test scores for the HET-CAM and bovine eye tests. The irritation potential for all the prepared niosomes and the excipients used was equally interpreted by the HET-CAM (the conjunctival model) and bovine eye assay (the corneal model). These findings indicate that a good correlation was achieved between the two toxicity models. The only exception to these results was recorded for acetone. Based on the HET-CAM test, acetone was regarded as a strong irritant, whereas it was found a moderate irritant when applied on the bovine cornea. The scoring system of bovine cornea is based on assessing the disruption of the test substances to the epithelium barriers and consequently, corneal opacity and permeability induced by the test substances are estimated. Acetone, an organic solvent, could induce damage to the corneal tissues other than the immediate effects recorded for the surfactants used. Therefore, histological examination to the corneal layers could be helpful as discriminatory tool and confirmatory to the HET-CAM and bovine corneal opacity and permeability (BCOP) tests. 201 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Table 5.7 Summary of HET-CAM and bovine eye test interpretations 202 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5.4.3. Histopathological evaluation of the bovine corneas Histological sectioning, fixing and H&E staining of the corneal samples were carried out to help understand and evaluate the degree of corneal injuries for the test substances. Negative (Figure 5.10A) and positive (Figure 5.11 and Figure 5.12) controls were included to facilitate the interpretation of the results. The negative controls provide the baseline against which histological changes are compared. Furthermore, the histology of the negative control corneas was used in this study to evaluate the quality/acceptability of the sections from batch to batch, to minimise or rule out batch processing and cattle age effects. Older cattle have thicker corneal sections where Descemet’s membrane increases in thickness with the age of the donor. Also, poorly-prepared sections (over-trimmed or poorly-embedded) show thicker corneal layers (Cuellar et al., 2002). In this study, a negative control was used as a baseline for each processed batch to evaluate the irritation potential for the test substances (Table 5.8). 5.4.3.1. Evaluating the negative control- treated corneas Figure 5.10 A and Figure 5.10 B show photomicrographs for the whole corneal layers at 10x magnification; and the epithelial layers at 40x magnification for the saline-treated cornea (the negative control). The cornea is an avascular tissue (no circulatory system), hence no leukocyte infiltration is evident. Since the test materials were applied topically onto a silicon O-ring on the surface of the cornea, the evaluation was conducted top-down, starting with the upper epithelium and proceeding down through the epithelial layers, via the stroma and down to the endothelium. The depth of corneal injury basically depends on the penetration of the test substance through the three principles corneal layers: the epithelium, stroma and endothelium. Figure 5.10 B shows magnified corneal epithelium which is composed of three distinct layers (the basal cell layer, wing cell layer and squamous cell layer). The basal cell layer is a columnar-cell layer directly attached to the basement membrane above Bowman’s layer. Several cell-layers of wing cells are on top of the basal layer. The superficial layer or squamous layer is composed of flattened with limited cytoplasm and highly condensed nuclei. The second principal layer is the stroma which accounts for 90% of the total thickness of the cornea. It is composed of wellorganised collagen matrix fibres with dispersed keratocytes. The third layer is endothelium, which is a single layer of flattened cells attached to the basal surface of Descemet’s membrane. 203 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery (A) Corneal layers (10x magnification) (B) Epithelial layer (40x magnification) Figure 5.10 Photomicrographs (A 10x; B 40x) of H&E stained corneal sections of negative control treated with saline for 30 s 204 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5.4.3.2. Evaluating the positive control-treated corneas Epithelial defects and stromal oedema are two common causes of ocular irritation after topical application of test substances (Harbell et al., 2006; Harbell et al., 1999). Cell loss, vacuolisation (presence of vacuoles or holes), pyknosis (nuclei coagulation), saponification (due to the effect of alkali) and separation of cells from the Bowman’s membrane are characteristic lesions observed in the epithelium (Figure 5.11 and Figure 5.12). The lesions of stroma are predominantly presented in form of stromal oedema, keratocytes pyknosis and vacuolisation. The stromal swelling or oedema is associated with the loss of the normal ordered array of extracellular collagen matrix fibres. Stromal swelling is directly related to corneal opacity. The goal of scoring the corneal lesions is not to report separately but rather to correlate histological scores with the opacity and permeability values. It is accepted that the greater the stromal oedema, the greater the corneal opacity found (Nishida, 2005). A potential utilisation of histological scoring of the corneal lesions was to quantify the degree of opacity by measuring stromal thickness using employed Leica software. Stromal swelling could be manifested in the form of vacuoles in the organised collagen matrix. As the degree of vacuolisation increases, the overall thickness of the stroma is expected to increase. Lastly, the damage to endothelium would be expected to result from mechanical damage rather than the topical application of test substances. 205 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery (A) 5x magnification (B) 40x magnification Figure 5.11 Photomicrographs (A at 10x magnification; B at 40x magnification) of H&E stained corneal sections of acetone-treated cornea for 30 s 206 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Figure 5.11 A & B shows the corneal layers treated with acetone and a magnified photomicrograph for the epithelium layer. Squamous layer vacuolisation, squamous layer sloughing and wing layer coagulation were observed (Figure 5.11 B). The stroma showed a significant increase (P < 0.001) in thickness compared with the stroma of the saline-treated cornea. The stromal thickness was 2 times thicker than that for the negative control-treated. Further, marked stromal collagen matrix condensation and keratocytes condensation were observed. The lower opacity and permeability scores calculated for acetone-treated eyes from the bovine eye test are attributed to coagulation of the epithelium and consequently this could decrease the permeability of the epithelium to the fluorescent dye. Hence, additional scores should be counted for the marked coagulation and condensation of acetone-treated cornea as shown by histological examination. Figure 5.12 A & B shows the corneal layers treated with NaOH (0.5M) and a magnified photomicrograph for the remnants of the epithelium and stroma. Extensive corneal damage was observed for NaOH treated control. For example, epithelial rupture, saponification and epithelial loss were registered. Furthermore, the stroma was exhibited extensive vacuolisation and swelling. These results were correlated well with opacity and permeability scores from the bovine eyes test. The marked opacity and loss of lustre induced by NaOH can be attributed to complete loss of epithelial barriers. 207 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery (A) 5x magnification (B) 40x magnification Figure 5.12 Photomicrographs (A at 5x magnifications; B at 40x magnification) of H&E stained corneal sections of NaOH (0.5 M)-treated cornea for 30 s 208 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5.4.3.3. Evaluating the test substances Test substances applied for 30 s Figure 5.13 A-F shows photomicrographs of H&E stained histological sections of corneas treated with CH powder, C24 powder, F-S60, F-DCP, F-C24 and F-CH for 30 s according to the bovine eye assay. Figure 5.13 A demonstrates an extensive epithelial loss and swelling or oedema induced by CH powder. Further, epithelial detachment, vacuolisation and stromal collagen matrix vacuolisation were observed (Table 5.8). This indicates that CH is a strongly irritant substance. In contrast, intact epithelial layer and mild stromal oedema were recorded for C24 powder indicating that C24 is a slightly irritant. These results are in accordance with those obtained from the opacity and permeability scores. Additionally, there were no obvious signs of epithelial loss or stromal oedema recorded for the prepared niosomes. The slight superficial sloughing of squamous layer exhibited by F-DCP and F-CH (Figure 5.13 D&F) indicates good corneal tolerability. These results are also well correlated with opacity and permeability scores. 209 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery (A) CH powder (5x magnification) (B) C24 powder (10x magnification) (C) F-S60 (10x magnification) (D) F-DCP (10x magnification) (E) F-C24 (10x magnification) (F) F-CH (10x magnification) Figure 5.13 Photomicrographs (A at 5x magnification; B, C, D, E and F at 10x magnification) of H&E stained corneal sections treated with CH Powder, C24 powder and the prepared niosomes for 30 s 210 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Test substances applied for 1, 3 and 8 h The duration of exposure was also studied by applying the prepared niosomes for a longer time 1, 3 and 8 h. The treated corneas were dissected, fixed, sectioned, stained and photographed. Figures 14-16 show photomicrographs of H&E stained corneas treated with the prepared niosomes for 1, 3 and 8 h respectively. Table 5.8 summarises the histological toxicity and stromal thickness for all test substances at the different time points. After 1 h, all the prepared niosomes showed minimal to no histopathological signs. Intact epithelium and non-significant differences between corneal stromal thicknesses were observed after 1 h compared with 30 s-treated samples. Similarly, the prepared niosomes did not demonstrate notable differences after 3 h. However, epithelial vacoulisation and slight epithelial swelling induced only by F-CH (Figure 5.15 D). After 8 h, moderate stromal collagen matrix vacoulisation exhibited by the prepared niosomes (Figure 5.16 A-D) and marked epithelial cell loss and erosion induced only by F-CH (Figure 5.16 D). The stromal oedema induced by the prepared niosomes is likely to be reversible and the cornea regains its normal configuration and normal structure after stopping treatment (Guinedi et al., 2005; Hughes et al., 2004). 211 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery (A) F-S60 (10x magnification) (C) F-C24 (10x magnification) (B) F-DCP (10x magnification) (D) F-CH (10x magnification) Figure 5.14 Photomicrographs (A, B, C and D at 10x magnification) of H&E stained corneal sections treated with the prepared niosomes for 1 h 212 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery (A) F-S60 (10x magnification) (C) F-C24 (10x magnification) (B) F-DCP (10x magnification) (D) F-CH (10x magnification) Figure 5.15 Photomicrographs (A, B, C and D at 10x magnification) of H&E stained corneal sections treated with the prepared niosomes for 3 h 213 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery (A) F-S60 (10x magnification) (C) F-C24 (10x magnification) (B) F-DCP (10x magnification) (D) F-CH (10x magnification) Figure 5.16 Photomicrographs (A, B, C and D at 10x magnification) of H&E stained corneal sections treated with the prepared niosomes for 8 h 214 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Table 5.8 Histological corneal lesions scores for controls and the prepared niosomes at different time points (Thickness values are expressed as mean values ± SD, n = 3) Test substance Epithelium Cell loss Stroma Vacuolisation/coagulation Thickness Collagen matrix (µm) Vacuolisation Control (30 s) - -/- 657 ± 60 - NaOH (1M) (30 s) +++ +++/+++ 1180 ± 267 +++ Acetone (30 s) + +/+++ 1278 ± 110 + CH powder (30 s) +++ +++/+++ 1280 ± 258 +++ C24 Powder (30 s) - +/- 756 ±211 + F-S60 (30 s) - -/- 653 ± 64 - F-DCP (30 s) - -/- 649 ± 21 - F-C24 (30 s) - -/- 663 ± 15 - F-CH (30 s) - -/- 745 ± 22 - Control (1 h) - -/- 740 ± 39 - F-S60 (1 h) - -/- 621 ± 70 - F-DCP (1 h) - -/- 672 ± 58 - F-C24 (1 h) - -/- 691 ± 12 - F-CH (1 h) - -/- 664 ± 17 - Control (3 h) - -/- 674 ± 91 - F-S60 (3 h) - -/- 720 ± 15 + F-DCP (3 h) - -/- 733 ± 49 + F-C24 (3 h) - -/- 632 ± 14 + F-CH (3 h) - -/- 677 ± 25 + Control (8 h) - +/- 687 ± 73 + F-S60 (8 h) - +/- 702 ± 16 + F-DCP (8 h) - +/- 635 ± 31 + F-C24 (8 h) - +/- 679 ± 28 + F-CH (8 h) ++ ++/- 781 ± 28 + 215 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery 5.4.4. In vitro release studies In vitro dissolution or release testing is a regulatory requirement in drug development and quality control to warrant that a dosage form will release the active pharmaceutical ingredient at appropriate rate and extent (Burgess, 2010). Development of an in vitro release test should consider physiological conditions such as the temperature, the pH of the fluid, ionic composition, buffer capacity, surface tension and hydrodynamic stirring rate, in order to closely represent the in vivo performance. In this study, many physiological factors were considered, such as the ocular temperature (35oC), pH (7.4) and osmolality (isotonic with the tear fluid) in order to simulate ocular surface conditions. The in vitro release performances of NTX from the prepared niosomes were studied on Franzdiffusion cells. An aqueous solution of PBS was used as a release medium which is composed of electrolytes (NaCl 137 mM, Na2HPO4 10 mM, KH2PO4 1.47 mM, KCl 2.68 mM), a pH value of 7.4, osmolality of 278 ± 5 mOSm/kg, stirred and equilibrated at 35oC ± 0.5. The tested niosomes in the donor compartment were separated from the receptor compartment, which was filled with 12 ml of the release medium, by a cellulose membrane. Figure 5.17 shows various release profiles of NTX from the aqueous solution and the prepared niosomes. Levelling off the NTX release from NTX solution formulation observed at the 6 h point is shown in Figure 5.18. This can be attributed to attaining equilibrium between the donor and receptor compartments. The sample size used (0.4 ml) was insufficient to maintain a concentration gradient between the donor and receptor compartments. Unperturbed NTX release profiles were obtained when relatively larger sample sizes (1 ml and 2 ml) were used (Figure 5.18). Such flattening was not observed with the prepared niosomal formulations over the 12 hrelease period. The partitioning of NTX molecules through the bilayer membranes of the niosomes was the rate-controlling step in the release process. Hence, an appreciable concentration gradient was maintained. 216 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Figure 5.17 In vitro release profiles of NTX from the NTX solution and the prepared niosomes (Results are expressed as mean values ± SD, n = 3) 217 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Figure 5.18 Effect of sample volume on the diffusion of NTX from the aqueous solution (Results represent mean values, n = 3) Table 5.9 In vitro release parameters for NTX from the prepared niosomes compared with NTX aqueous solution (Results are expressed as mean values ± SD, n = 3) Formulation Q2h (%.cm-2) Q6h (%.cm-2) DE% -12.cm-2 NTX- solution 17.06 ± 1.70 35.29 ± 3.05 26.34 ± 0.34 F-S60 4.53 ± 0.26 11.00 ± 0.29 8.86 ± 0.47 F-DCP 2.29 ± 0.11 6.00 ± 0.41 5.03 ± 0.38 F-C24 5.38 ± 0.94 13.88 ± 1.77 11.60 ± 1.4 F-CH 8.15 ± 0.21 20.11 ± 0.16 17.01 ± 0.29 218 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Table 5.10 Tukey’s pair wise comparison of Q2h, Q6h and DE%-12 for NTX from the aqueous PBS solution and the prepared niosomal formulations NTX - solution F-S60 F-DCP F-C24 Parameters Q2h; Q6h;DE% -12 Q2h; Q6h;DE%-12 Q2h; Q6h;DE%-12 Q2h; Q6h;DE%-12 F-S60 S;S;S - - - F-DCP S;S;S S;S;S - - F-C24 S;S;S NS ; S ; S S;S;S - F-CH S;S;S S;S;S S;S;S S;S;S S = significant difference (P < 0.05) NS = non-significant (P > 0.05) The release parameters (Q2h, Q6h and DE %-12) calculated for the NTX solution were 17.06, 35.29 and 26.34 respectively (Table 5.9). These parameters were significantly higher (P < 0.01) than those calculated for the prepared niosomal formulations. For example, the Q6h value calculated for NTX solution was 5.88 and 2.54 times greater than that for F-DCP and F-C24 respectively. These results demonstrate the NTX release retarding efficiency for the prepared niosomes. Similar results were obtained from acyclovir encapsulated liposomes. The in vitro release rate for acyclovir from liposomes was markedly lower than that for an aqueous solution of acyclovir indicating that the prepared liposomes established controlled acyclovir release (Law et al., 2000). The drug release rate and extent were significantly different (P < 0.5) amongst the prepared niosomal formulations (Table 5.9 and Table 5.10). F-DCP showed the highest NTX retarding efficiency, whereas NTX release rate was highest for F-CH. The drug release rates for F-S60 and F-C24 came in the middle. For instance, the Q6h values for F-S60, F-DCP and F-CH were 11, 6 and 20 %.cm2 respectively. These results revealed good discrimination amongst the drug release profiles for the prepared niosomes. This is partly due the ability of the developed in vitro release test to show perfect sink conditions. In addition, this can be attributed to the effect of lipid composition and membrane additives used to fabricate the tested niosomes. 219 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery The highest NTX release retarding efficiency ascribed for F-DCP niosomes could be due to the incorporation of DCP in the bilayer membranes. This could lead to two possible effects: firstly, the electrostatic interaction between the positively charged NTX molecules and the negatively charged DCP was likely to slow the NTX release from the vesicles. Secondly, the double hydrocarbon chains of DCP could impart more order and improve the packing of the bilayer membranes, as previously outlined in Figure 4.13. Similar results have been reported with charged liposomes. The lowest release rate was obtained from positively charged liposomes entrapping acetazolamide (weak acid). This has been ascribed to the electrostatic interaction between positively charged liposomes and the negatively charged drug (Hathout et al., 2007). Incorporating bulky structures with long and highly hydrophilic poly-24-oxyethylene chains (C24 molecules) in the bilayer layer of F-C24 could increase the interstitial spaces between the bilayer membranes. This could increase the permeability of the membrane to water-soluble solutes and hence the drug release rate. F-CH showed 1.18- and 3.33-fold increases in Q6h compared with those calculated for F-S60 and F-DCP respectively. CH molecules could interact with the hydrocarbon region such that its carboxylate group interacts with the head group of Span 60 and the steroidal nucleus aligns itself parallel to the acyl chains. Such orientation is likely to create some vacancies in the hydrophobic domains of the bilayer membranes (Figure 4.15). This could provide increase NTX permeability through the bilayer membranes and increase release rate for NTX compared with F-DCP and FS60. 5.4.5. In vitro release kinetic studies The Higuchi model has been used to explain the drug release kinetics for many niosomal formulations. The Higuchi model was also used to interpret the mechanism of water-soluble and insoluble drugs from uni- and multi-lamellar niosomes (Bayindir & Yuksel, 2010; Guinedi et al., 2005). In this study, the Korsmeyer-Peppas equation was utilised to elucidate the release kinetics of NTX from the prepared niosomes. This equation does not force the time data points to a particular exponent (t1/2) like Higuchi equation does. In Addition, it can give more insights on other drug release mechanisms, such as Fickian (diffusion), non-Fickian and erosion-mediated (zero-order) release. Further, this equation has been successfully used to explain release 220 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery mechanisms from thin film, cylindrical, disc and spherical controlled release devices (Ritger & Peppas, 1987). Table 5.11 summarises the release exponent (n) values and regression coefficients (R2) calculated for the tested formulations. The n values for F-S60, F-DCP and F-C24 were found to be close to 0.85. These results suggested that the general operating release mechanism was zero-order (erosion-mediated release). For F-CH, the n value was found to be 0.77 suggesting that the NTX released by coupled diffusion and erosion mechanism, however, the dominant release mechanism was by erosion. Zero-order kinetics prevails if n > 0.66 (Mockel & Lippold, 1993). Table 5.11 In vitro release kinetic parameters for NTX from the prepared niosomes (Results are expressed as mean values ± SD, n=3) Formulations Release exponent (n) Regression coefficient (R2) F-S60 0.80 ± 0.015 0.996 F-DCP 0.83 ± 0.071 0.998 F-C24 0.80 ± 0.061 0.999 F-CH 0.77 ± 0.032 0.999 5.4.6. Ex-vivo corneal permeation studies Figure 5.19 shows transcorneal permeation profiles of NTX from the investigated formulations using excised bovine corneas. The following transcorneal permeation parameters were calculated: steady-state flux, apparent permeability coefficient (Papp) and lag time (tL) (Table 5.12). Table 5.13 shows statistical significance differences among the transcorneal permeation parameters for NTX from the aqueous solution and the prepared niosomes. The prepared niosomes achieved controlled NTX permeation through the excised bovine cornea compared with NTX solution. The steady-state flux and Papp calculated for F-DCP were 2.4 times lower than those for the NTX solution. In addition, the permeation parameters obtained from the NTX solution were 1.6 times higher than those from F-CH. These results correlated well with the in-vitro release data suggesting that the corneal permeation process of NTX is dependent on its release characteristics. 221 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Similar results were reported for lipid-based delivery systems, such as liposomes and solid lipid nanoparticles (SLNs), using excised rabbit corneas and bioengineered human corneas (Attama et al., 2009; Law et al., 2000). The permeation rate and apparent permeability coefficient for acyclovir-encapsulated liposomes and timolol maleate loaded SLNs were significantly controlled compared with that of the corresponding drug solutions. When the prepared acyclovir liposomes, which demonstrated lower permeability, were applied on rabbit eyes, higher ocular bioavailability for acyclovir was obtained, compared with that for the aqueous drug solution (Law et al., 2000). This is attributed to the ability of liposomes to provide longer precorneal residence time and less precorneal drug loss due to tear dilution and spillage than the acyclovir aqueous solution. 222 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Figure 5.19 Transcorneal permeation profiles of NTX from the NTX solution and the prepared niosomes using excised cow corneas (Results are expressed as mean values ± SD, n = 5) 223 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Table 5.12 Steady state flux, apparent permeability coefficient (Papp) and tL for NTX from the aqueous PBS solution and the prepared niosomes though excised cow corneas (Results are expressed as mean values ± SD, n = 5) Steady state flux Papp x 106 tL (µg/h) (cm/s) (h) NTX solution 18.74 ± 2.00 7.65 ± 0.83 1.27± 0.26 F-S60 8.00 ± 0.37 3.26 ±0.36 0.68 ± 0.35 F-DCP 7. 89 ± 1.23 3.22 ± 0.73 0.5 ± 0.30 F-C24 9.30 ± 2.25 3.79 ± 0.76 0.90 ± 0.44 F-CH 11.88 ± 1.45 4.85 ± 1.00 1.00 ± 0.23 Formulation Table 5.13 Tukey’s pair wise comparison of steady state flux, Papp and tL for NTX from the aqueous PBS solution and the prepared niosomal formulations NTX - solution F-S60 F-DCP F-C24 Parameters Flux; Papp; tL Flux; Papp; tL Flux; Papp; tL Flux; Papp; tL F-S60 S;S;S - - - F-DCP S;S;S NS ; NS ; NS - - F-C24 S ; S ; NS NS ; NS ; NS NS ; NS ; NS - F-CH S ; S ; NS S ; S ; NS S;S;S NS ; NS ; NS S = significant difference (P < 0.05) NS = non-significant (P > 0.05) 224 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery To demonstrate the degree of correlation between in vitro and ex vivo studies, a plot of Papp values versus in vitro release parameters (Q2h, Q6h and DE%-12h) was constructed (Figure 5.20). A direct relationship was established between Papp and in vitro release parameters for NTX solution and the prepared formulations. These findings show the good in vitro/ex vivo correlation and confirm the ability of the prepared niosomes to retard NTX release on the excised corneal surface. Figure 5.20 Relationship between NTX release rate (Q2h, Q6h and DE%-12h values) and apparent permeability coefficient (Papp) for NTX from the NTX solution and the prepared niosomes 225 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery With regard to the lag time (tL), the NTX solution produced the longest tL value, yet produced the highest steady-state flux, whereas F-DCP achieved the shortest. The tL values for F-DCP and FS60 demonstrated 2.54 and 1.86 times shorter than that for the aqueous NTX solution. The tL values for F-C24 and F-CH came in the middle (Table 5.12). These results could be attributed to the fact that the formulated niosomes possess better wetting and spreading abilities on the lipophilic corneal surface than the aqueous solution. The calculated tL values were in accordance with the contact angle and spreading coefficient measurements for the prepared niosomes. Superior wetting and spreading abilities in niosomes can promote intimate contact with the lipophilic corneal epithelium, increase the available corneal surface area in contact with the applied formulations and promote better corneal absorption. Further, by virtue of their amphiphilic nature, niosomes are reported to impart a corneal permeation-enhancing effect (Saettone et al., 1996b). These were thought to enhance the penetration of NTX, and consequently shorten the lag time. To support the previous assumption, a θ versus tL plot was drawn (Figure 5.21). A direct relationship between the θ and tL values was obtained. Figure 5.21 Relationship between contact angle and lag time from the NTX solution and four niosomal formulations 226 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery It is worth noting that negatively charged niosomes (F-DCP) produced a shorter tL than neutral ones (F-S60). This suggests that negatively charged niosomes could be more effective than neutral ones for improving the permeation of NTX through the cornea, despite the negatively charged nature of the epithelial corneal surface. Similar results were observed elsewhere. Negatively charged multilamellar liposomes showed enhanced corneal permeability of penicillin G (water soluble drug), compared with both neutral and negatively charged unilamellar liposomes (Schaeffer & Krohn, 1982). Further, these findings support the hypothesis that flux rate is independent on the lag time. Similar results have been reported with other colloidal drug delivery systems for transdermal delivery of indomethacin (El Maghraby, 2010). Longer tL values were reported for Self–microemulsifying drug delivery systems (SMEDDS), although these formulations exhibited the greatest flux values. In the same report, application of indomethacin in microemulsion forms produced a transdermal flux value relatively smaller than that obtained from SMEDDS but with a shorter tL value. 5.4.7. Physical stability of the prepared niosomes The physical stability of the prepared niosomes against aggregation and leakage of the payload was studied over three months at three different temperatures; 4, 25 and 35oC. More specifically, the effects of ageing (storage time) and storage temperature on the physical stability were evaluated. 5.4.7.1. Effect of ageing on volume diameter D[4,3] The prepared niosomes were stored for three months at three different temperatures (4, 25 and 35oC). The three different temperatures were used to represent fridge, room and physiological ocular temperatures. Over the 90-day storage period, samples were withdrawn and evaluated for size by measuring D [4,3]. Table 5.14 shows the D [4,3] values recorded for the prepared niosomes over 3 months’ storage at the three different temperatures. Apart from F-C24, the prepared niosomal formulations showed slight (< 2%) to significant (40%) increases in the D [4, 3] values compared with the original D [4, 3] values over the storage period at the three different temperatures. After 90 days, the average D [4, 3] values measured for F-S60 were 8.48, 9.57 and 9.13 µm at 4, 25 and 35oC respectively. There was no significant increase (P > 0.1) in the D [4,3] at 4oC, whereas the D [4,3] values significantly increased (P < 0.05) after 3 months at 25 and 35oC. Similar trends were recorded for F-DCP and F-CH. 227 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery These results showed that the ambient storage temperature is more likely to have a noticeable effect on the stability of the prepared niosomes against aggregation. At higher temperatures, the niosomes were relatively unstable compared with the cold storage temperature. The thermal energy imparted to vesicles increases when the temperature increases. Hence, the rate and force of collision between vesicles increases. Niosomes tend to aggregate and coalesce to become more energetically stable and, consequently, niosome size increases. These findings suggest that the storage temperature of choice is 4oC where the prepared niosomes showed minimal changes in sizes over three months. Conversely, the prepared discomes (F-C24) exhibited a notable decrease in the D [3,4] during storage. This decrease was significant (P< 0.05) when the temperature increased from 4 to 25oC. For example, the D [4,3] values for F-C24 were 17.2 and 12.54 µm at 4 and 25oC respectively after 3 months. The respective decreases in the D [4,3] value compared to that for the freshly prepared discomes were 23% and 44%. F-C24 is considered to be a giant discoid-shaped niosome, but the results showed that the prepared discomes were not able to retain their characteristic size for a period of 2 months at higher storage temperatures. These results were in accordance with the fact that discomes lose their unique discoidal morphology at higher temperatures (> 35oC) and turn into small spherical vesicles due to partitioning C24 molecules throughtout of the bilayer membranes (Uchegbu et al., 1996b; Uchegbu & Vyas, 1998). These results suggest that cold storage is also the storage temperature of choice for discomes in order to maintain their morphological stability. 228 Chapter 5….Evaluation of Niosomal Formulations for Ocular Delivery Table 5.14 Effect of ageing on the volume diameter (D [4,3]) for the prepared niosomes over three months at three different temperatures (Results are expressed as mean values ± SD, n = 3) Control (0 time) Formulation 4oC 25oC 35oC 1 month 4oC D [4,3] µm 25oC 2 month 35oC 4oC D [4,3] µm 25oC 3 month 35oC 4oC D [4,3] µm 25oC 35oC D [4,3] µm F-S60 8.30 ± 1.14 8.30 ± 1.14 8.30 ± 1.14 8.12 ± 0.42 8.39 ± 0.33 9.05 ± 0.75 8.49 ± 0.21 8.67 ± 0.43 9.75 ± 1.12 8.48 ± 0.21 9.57 ± 1.11 9.13 ± 1.33 F-DCP 7.50 ± 0.59 7.50 ± 0.59 7.50 ± 0.59 7.56 ± 0.66 7.72 ± 0.33 9.39 ± 1.59 7.89 ± 1.12 8.34 ± 0.76 11.18 ± 1.10 8.10 ± 1.34 9.17 ± 1.85 10.71 ± 0.98 F-C24 22.41 ± 1.40 22.41 ± 1.40 22.41 ± 1.40 22.3 ± 4.30 13.00 ± 0.77 16.79 ± 1.60 20.16 ± 2.74 13.60 ± 1.53 13.00 ± 1.22 17.20 ± 4.2 12.54 ± 1.73 10.50 ± 2.5 F-CH 5.37 ± 0.81 5.37 ± 0.81 5.37 ± 0.81 4.99 ± 0.08 5.67 ± 1.33 6.60 ± 0.96 5.28 ± 0.39 6.84 ± 1.51 9.96 ± 2.12 5.09 ± 1.02 8.77 ± 1.65 14.22 ± 3.80 229 Chapter 5….Evaluation of the Prepared Niosomes for Ocular Delivery 5.4.7.2. Effect of ageing on NTX retention Another physical stability parameter to determine is the drug leakage from the prepared niosomes over the 90-day storage period. The percentages of NTX retained in niosomes were measured for three months at the three different temperatures; 4, 25 and 35oC. A blanket of nitrogen gas was used to replace air in the head space of the glass vials in order to minimise or slow down autoxidation of NTX. This was carried out to determine the amount of NTX retained in the niosomes without interference from the drug loss due to oxidative degradation over the period of study. It has been reported that the degradation rate of an oxidisable drug can be retarded by replacing air in ampoules with inert nitrogen gas (Johnson & Gu, 1988). Firstly, the effects of storage temperatures on percentages of NTX retained in the prepared niosomes over the 90-day period are presented in Figure 5.22. The results showed that the prepared niosomes were relatively more stable to NTX leakage at 4oC than at higher temperatures. After 90 days, the percentages of NTX retained in F-S60 were 85, 64.5 and 53% at 4, 25 and 35oC respectively. Similarly, the respective percentages of NTX retained in F-DCP at the three temperatures were 87, 73 and 66%. The effect of temperature on NTX leakage from all the prepared niosomes was significant (P < 0.001). This effect is more pronounced with ageing (storage time). For example, the drug leakage after a month was not significant (P > 0.05) at 4oC but a statistically significant (P < 0.05) difference was obtained after three months. These results could be attributed to defects in the niosomes’ bilayer integrity, as a consequence of ageing and the effect of temperature on the fluidity of the bilayer membranes leading to partitioning of NTX molecules throughout the bilayer membranes. At 4oC, the bilayer membranes are extensively in a gel state, hence NTX leakage is relatively minimal compared with the state of bilayer membranes at higher temperatures. Secondly, the effect of bilayer membrane composition on the NTX leakage over three months is outlined in Figure 5.23. The percentages of NTX retained in F-S60, F-DCP, F-C24 and F-CH after three months at 4oC were 85%, 87%, 83% and 80.5% respectively. The NTX leakage from F-CH was significantly higher (P < 0.05) than that calculated for both F-S60 and F-DCP. Further, the percentages of NTX retained in F-S60, F-DCP, F-C24 and F-CH after three months at 25oC were 64.5%, 73%, 58 and 52% respectively. In addition to the temperature, incorporation of the 230 Chapter 5….Evaluation of the Prepared Niosomes for Ocular Delivery membrane additives was found to have an additional effect on niosomes stability. The percentages of NTX retained for the prepared niosomes were significantly different (P < 0.05) due to the incorporation of the membrane additives used. F-DCP was found to be the most stable formulation, whereas F-CH showed the highest NTX leakage. The electrostatic interaction between the positively charged NTX and negatively charged DCP partly decreases the partitioning of NTX throughout the bilayer membrane. The packing effect of DCP on the bilayer membrane can also decrease its permeability to NTX compared with the conventional niosomes (F-S60). In contrast, the incorporation of CH can alter the nature of the bilayer membrane by disturbing its packing properties in a manner opposite to cholesterol which is attributed to its highly hydrophilic nature and the introduction of two hydroxyl groups in the steroidal nucleus. CH can create vacancies in the bilayer membrane as shown in Figure 4.15. Hence, the partitioning of NTX throughout F-CH was faster than F-S60. However, the higher NTX leakage calculated for F-C24 rather than F-S60 is mainly ascribed to the discomes breaking down into relatively smaller sizes. It is worth mentioning that over the three-month study period, no degradation peaks were detected for NTX by the HPLC chromatograms. These could be attributed to two possible reasons: firstly, replacing the oxygen gas in the heads of the glass vials with inert nitrogen gas can minimise the oxygen concentration available to initiate autoxidation (Johnson & Gu, 1988); secondly, the influence of vesicular encapsulation and its stabilising effect against NTX oxidative degradation, as shown and discussed earlier in Chapter 4, Section 4.4.9. 231 Chapter 5….Evaluation of the Prepared Niosomes for Ocular Delivery Figure 5.22 Effect of storage temperature on NTX retention in the prepared niosomes over three months (Results are expressed as mean values ± SD, n = 3) 232 Chapter 5….Evaluation of the Prepared Niosomes for Ocular Delivery Figure 5.23 Effect of niosome composition on NTX retention over three months (Results are expressed as mean values ± SD, n = 3) 233 Chapter 5….Evaluation of the Prepared Niosomes for Ocular Delivery Conclusion • The prepared niosomes were found to be non-irritant, when applied onto the CAM surfaces (Conjunctival model) and bovine corneas. • The HET-CAM test, BCOP assay and histological examination were well correlated and successfully used to study the ocular irritation potential of niosomes and can be considered as an alternative in vitro toxicity tests for the in vivo rabbit eye test. • The in vitro release rate (Q2h and Q6h) and extent (DE-12%) were significantly dependent on surfactant/lipid composition. • The prepared niosomes demonstrated significant controlled release and enhanced penetration of NTX on the excised bovine corneas, compared with the aqueous NTX solution. • The prepared niosomes were found to be more physically stable at 4oC than ambient conditions over 90 days. The four niosomal formulations were demonstrated versatile sizes, morphological features, in vitro NTX release and transcorneal permeation profiles. These key findings point out the utility of the prepared niosomes as a safe, effective and new ophthalmic pharmaceutical for the opioid growth factor antagonist-NTX- for treatment of diabetic complications on the surface of the eye. 234 General Discussion, Conclusion and Future Directions 6. General discussion, conclusion and direction for future research 6.1. General discussion and conclusion Diabetic keratopathy is the term given to diabetic complications in the cornea and affects up to 60% of diabetics (Schultz et al., 1981 ). Diabetic keratopathy or diabetic corneas demonstrate corneal nerve damage, reduced corneal sensation and delayed wound repair. The current treatment consists of topical lubricants and antibiotics. However, these treatments, even in combination, are not effective in most cases (McLaughlin et al., 2010; Schultz et al., 1981 ). Recently, a group of macromolecules (insulin and nerve growth factor) and small molecules (naltrexone and nicergoline) have shown effects which accelerate corneal wound healing, but only NTX has both desirable pharmacological properties and a high safety profile with ocular tissues. However, there has been no data yet on formulating NTX for ocular use. Eye drops are the most popular ophthalmic dosage form. This is because of simplicity and patient acceptability. However, these advantages are offset by low bioavailability, frequent instillations, and the typical pulse-type entry of the drug, which might promote both patient incompliance and systemic side effects. Hence, topical ocular delivery is still a challenging endeavour; the two available strategies to improve both patient compliance and ocular bioavailability are to increase corneal permeability and prolong precorneal residence time. Niosomes (non-ionic surfactant vesicles) were proposed in this study as a potential ocular delivery system for NTX. Niosomes are economic and more chemically stable alternatives to liposomes; they have the convenience of being eye drops; they can be fabricated from biodegradable and biocompatible materials; can provide controlled drug-release and penetrationenhancing effects; and they can protect the encapsulated drug molecules from enzymatic degradation. For the effective ocular drug delivery, the following properties should be considered during niosomes fabrication: • The size of the niosomes should be > 10 µm to resist drainage by reflex tearing. 235 General Discussion, Conclusion and Future Directions • The shape of the prepared niosomes should show some irregularities to fit properly into the cul-de-sac of the eye and lodge on the eye surface. • Ocular niosomes should ideally be thermo-responsive in order to release the drug content in a controlled yet timely manner, before being flushed by blinking and nasolacrimal drainage. This thesis was concerned with designing and characterising the major attributes of niosomes as ocular delivery systems for NTX. Prior to the formulation stage, the fundamental physical properties of NTX, such as aqueous solubility, lipid solubility and ionisation constant as well as developing and validating an HPLC method, were determined (Chapter 2). The results showed that NTX is freely water-soluble and is a hydrophilic moiety. The HPLC method was rapid and sensitive. The short run time of approximately 4 min, with isocratic flow, makes it ideal for routine analysis. LOD and LOQ were 0.55 µg/ml and 1.85µg/ml respectively. Additionally, the method developed was stability-indicating. There was no interference with the excipient used and the assay can be applied to quantify NTX in niosomal formulations. The forced degradation studies showed that autoxidation is the major pathway of the degradation of NTX. The results of NTX preformulation studies recommended using a surfactant-based delivery system to enhance corneal penetration through the lipophilic epithelium of the cornea, and to encapsulate NTX in bilayer vesicles in an attempt to protect NTX against oxidative stresses. In Chapter 3, two different classes of non-ionic surfactants (Span® and Brij®) were used for forming niosomes. The surfactants used for niosomes preparation were Span 20, Span 40, Span 60, Span 80, Brij 52 and Brij 72. These surfactants are widely used in food, cosmetic and pharmaceutical industries. They are also regarded as non-toxic and non-irritant materials (Lawrence, 2003; Yu, 2003). The surfactants used are composed of different hydrocarbon chain lengths, degree of saturation and various HLB values. Cholesterol was used as a membrane stabiliser at different concentrations. The effects of both surfactants and cholesterol on the niosomes’ physical properties, such as EE %, D [4,3] and gel/liquid phase transition temperature, were studied. Surfactant/lipid mixtures composed of different surfactant: cholesterol molar ratios (1:0, 9:1, 8:2, 7:3, 6:4 and 1:1 mol/mol) were studied. The surfactant:cholesterol ratio 7:3 was found to show the highest NTX EE%. Generally, Spans-based niosomes were found to be more 236 General Discussion, Conclusion and Future Directions physically stable than Brij-based niosomes. The Brij-based niosomes showed phase separation and creaming upon storage, irrespective of the cholesterol level. Span 60: cholesterol 7:3 mol/mol niosomes showed a residual phase transition, with the onset of the gel/liquid transition at the ocular temperature (approximately 35oC) indicating its thermoresponsiveness. In addition, The Span 60: cholesterol (7:3) niosomes exhibited the highest NTX EE% compared with other niosomes but the amount of NTX entrapped by the niosomes was still sub-therapeutic (< 0.4 mg/ml). In fact, these findings support the hypothesis that hydrophilic drugs are less preferentially entrapped by niosomes than lipophilic drugs (Essa, 2010). Hence, other attempts were sought elsewhere to improve EE% of NTX. The effect of membrane additives and various methods for niosome preparation were studied. In Chapter 4, both charged lipids (dicetyl phosphate [DCP] and stearyl amine [STA]) and watersoluble surfactants (poly-24-oxyethylene cholesteryl ether [C24] and sodium cholate [CH]) were investigated as bilayer membrane additives. The selected membrane additives were studied for their effects on the physical properties of niosomes. Different concentrations (0-10% mol/mol) of the selected additives were used. Niosomal formulations containing 2% to 5% mol/mol of C24, DCP and CH encapsulated a considerable amount of NTX. For example, F-DCP showed a 3-fold increase in NTX EE%, compared with the conventional niosomes (F-S60), using the TFH method. Additionally, the REV method was superior to other methods (TFH, MLV-FAT and DRV methods). REV niosomes showed roughly a 1.5-fold increase in EE% compared with TFH niosomes. The REV method has more lipid hydration efficiency than the TFH method. Also, the REV method generates niosomes with larger aqueous core to accommodate a considerable amount of water-soluble solutes compared with the TFH method. Incorporation of CH and DCP produced spherical niosomes similar to the conventional (F-S60) niosomes in shape, as shown by Cryo-SEM and CLSM. CH containing niosomes (F-CH) achieved smaller D [4,3] (5 µm) than that (8 µm) for F-S60, whereas introduction of C24 into the bilayer membrane produced niosomes significantly larger sizes (22 µm) with an oval-to-disc shape, indicating the formation of atypical giant niosomes, better known as discomes. It is worth mentioning that ocular drug bioavailability can be affected by the size of instilled ophthalmic suspensions. It is likely that as the particle size increases, eye discomfort could be created. As a consequence, tear production and blinking rate are likely to increase, leading to a 237 General Discussion, Conclusion and Future Directions decrease in residence time. It has been reported that an ophthalmic suspension made up of solid particles should have a diameter of 10 µm or less. This is to ensure uniform ocular absorption and prevent eye irritation (Lang et al., 2002). However, the upper limit of particles mentioned in the literature is based on a study on ophthalmic suspensions of water-insoluble anti-inflammatory steroids (Schoenwald & Stewart, 1980). According to this study, dexamethasone suspensions of different particle sizes were prepared. Suspensions A, B and C were formulated with particle sizes of 5.57, 11.5 and 22 µm respectively. The area under aqueous humour drug concentration versus time curve (AUC) was determined for each suspension formulation. The results showed that as the particles decrease in size, the time taken to reach peak concentration (Cmax) decreased and AUC increased. The ocular bioavailability obtained from a dexamethasone suspension, measured in terms of the AUC of suspensions A-C, can be expressed in the following order: A > B > C. The AUC was markedly dependent on the particle size of the suspension. These results were attributed to the fact that as the particles increase in size, the in vivo dissolution rate decreases, such that the particles are removed from the conjunctival sac before dissolution is completed. Therefore, both the rate and extent of penetration into aqueous humour decreased. Based on observation of the animals during the study, there was no reason to believe that dexamethasone induced abnormal tearing or blinking at an average size of 5.75-22 µm. There is no firm conclusion to support the need for particle size being less than 10 µm to minimise irritation in the eye (Schoenwald & Stewart, 1980). Nevertheless, shape, concentration and density are additional factors that make it difficult to select a specific particle size above which irritation or discomfort might result. The authors came to the conclusion as particle sizes increase, the surface area available for drug dissolution decreases. Hence, the drug dissolution rate decreases providing dissolution rate-limited absorption. For niosomes ophthalmic dispersion, the entrapped drug molecules are in the soluble molecular forms and they are either entrapped in the bilayer domains or encapsulated in the aqueous domains, depending on the drug physical properties. Partitioning of the entrapped/encapsulated drug molecules throughout the bilayer membranes at the eye temperature could be the ratecontrolling step for ocular drug absorption from niosomes. Controlled and constant drug partitioning is likely to be important for consistent ocular bioavailability. To design niosomes capable of responding to ocular temperature and release the entrapped drug molecules at 238 General Discussion, Conclusion and Future Directions controlled and constant rate (thermo-responsiveness) is more favourable than rigid niosomes (without gel/liquid transition temperature close to ocular temperature). The vesicles could spread promptly and in an even layer on the precorneal surface. Hence, niosomes should ideally be thermo-responsiveness for optimum ocular bioavailability. Other factors could affect the ocular bioavailability of the drug molecules entrapped by niosomes are; firstly, the ability of niosomes to carry the therapeutic molecules (EE %); secondly, the shape of niosomes could be an additional factor. Oval shaped niosomes (discomes) are thought to be of value for ocular delivery, due to better fitting on the conjunctival sac (retained by inner canthus) than spherical niosmes (Mainardes et al., 2005; Uchegbu & Vyas, 1998). Niosomes > 10 µm are suitable for ophthalmic drug delivery (Sahin, 2006). For instance, large niosomes (discomes) sized between 12 and 60 µm were found to entrap a relatively larger amount of timolol maleate (Vyas et al., 1998). The prepared discomes produced a controlled drug release and high bioavailability compared with the aqueous drug solution. The systemic drug absorption measured from discomes was at a negligible level (Vyas et al., 1998). Returning to Chapter 4, DSC thermograms revealed that the additives used completely abolished the residual gel/liquid transition of F-S60 niosomes indicating that the tested additives were accommodated by the bilayer membranes of the prepared niosomes. Niosomes achieved marked improvement in spreading ability compared with the aqueous vehicle. Viscosity measurements also demonstrated a 1.7- to 8.2-fold increase in the viscosity compared with that for the aqueous solution. These results demonstrate the ability of the tested niosomes to spread and wet easily the lipophilic corneal surface as well as prolong precorneal residence time. More importantly, the prepared niosomes were able to protect chemically the encapsulated NTX from photolytic degradation compared with dark storage. In Chapter 5, the HET-CAM test, bovine eye assay and histological examination of the bovine corneas were utilised for evaluating the ocular irritation and toxicity of NTX, the raw materials used to fabricate niosomes and the prepared niosomes. Further, in vitro release studies, release kinetics and the transcorneal permeation of NTX from the prepared niosomes were studied. Finally, investigations of the physical stability of the prepared niosomes were performed at three different temperatures 4, 25 and 35oC to evaluate their ability to maintain their size and retain their NTX payload. 239 General Discussion, Conclusion and Future Directions The results obtained from the HET-CAM (conjunctival) model were correlated well with those obtained from the corneal model. The HET-CAM and bovine eye tests were thought to be simple, less expensive, rapid and discriminating in vitro ocular toxicity models for niosomes as a potential ocular drug delivery. Histological examination of the corneas treated with the prepared niosomes at different time points 30 s, 1 h, 3 h and 8 h showed slight stromal oedema and no harmful signs. Such effects are reversible upon discontinuation of treatment or can be easily reversed by ocular administration of a simple hyperosmotic salt solution (Dawson & Edelhauser, 2010). In vitro drug release parameters (Q2h, Q6h and DE-12h) calculated for the tested niosomes indicate significantly prolonged (P < 0.01) NTX release rate. The calculated Q6h values for the prepared niosomes were approximately 1.6 to 6 times lower than those for the NTX solution. The four different niosomes generated four well-discriminated controlled-release profiles for NTX and the general operating release mechanism was zero-order release. Parallel to the in vitro release data, ex vivo transcorneal permeation profiles showed that niosomes potentially controlled NTX permeation though the excised bovine corneas. The steady flux rate values calculated for the prepared niosomes were 1.6 to 2.4 times slower than those for the NTX solution. This controlled flux rate is likely to be desirable for prolonging precorneal residence time and reducing the typical pulse type entry of eye drop solutions. Furthermore, controlling NTX release and permeation is specifically advantageous for taste-masking NTX, since the NTX solution is extremely bitter taste. The lag time (tL) values calculated for the prepared niosomes were shorter than those for the aqueous vehicle, suggesting that the prepared niosomes additionally have permeation-enhancing effects. A linear correlation between tL values and contact angles (θ) was discovered. In addition, Chapter 5 focused on studying the physical stability of the prepared niosomes stored at three different temperatures; 4, 25 and 35oC for 90 days. The effect of storage time (ageing) on size and the amount of NTX retained in niosomes was evaluated. Ageing provoked various changes in niosomes’ D [4,3] and in the percentages of NTX retained. These changes were dependent on the storage temperature and the composition of niosomes. Minimum changes in D [4,3] and the highest percentage of NTX retained were recorded at 4oC. 240 General Discussion, Conclusion and Future Directions In conclusion, this thesis has presented NTX as a new and potential treatment for diabetic keratopathy. It has also studied the physicochemical properties of NTX necessary for developing a new liquid dosage form. The main attributes of niosomes as an ocular delivery system is to prolong and enhance corneal uptake of NTX. These delivery devices can provide the pharmaceutical formulator with insights into using a surfactant-based delivery system as a potential ophthalmic pharmaceutical dosage form for NTX. 6.2. Limitations and future direction This project is an attempt to develop ocular niosomes for an ambitious therapeutic entity (NTX). Four different niosomal formulations have been designed and tailored to fit the anatomical and physiological barrier of the eye surface in terms of suitable size, morphology and thermoresponsiveness. Although the formulation approach was adopted in this thesis in an attempt to protect NTX molecules from oxidative stresses, it has apparently seemed to be insufficient. Other strategies, to overcome NTX oxidative degradation, should be considered for any future work, most notably, the use of stabilisers. Stabilisers are pharmacologically inactive ingredients added to the formulation to decrease the rate of decomposition of the active ingredient (Lang et al., 2002). A typical example for the stabilisers is antioxidants. Antioxidants can reduce or prolong the autoxidation process by different mechanisms. • Chain terminators are molecules with weak bonds to hydrogen atoms, such as thiol or phenol groups, which attack and scavenge free radicals and hence do not propagate oxidation chains reactions. • Sacrificial reductants (e.g. ascorbic acid) are compounds that can be oxidised more readily than the drug. They effectively scavenge oxygen while they are themselves consumed (Waterman et al., 2002). Other stabilisers to protect oxidisable drugs are chelating agents, such as editic acid and citrate, which can inhibit oxidative degradation by chelating any traces of heavy metals (Waterman et al., 2002). The pH also has an important effect on NTX chemical stability in aqueous solution, as it has been demonstrated that oxidation reaction is pH-dependent. The tear film pH 7.4 was selected as vehicle’s pH, although it is not the optimum pH to offer maximum chemical stability to NTX. This is because: 241 General Discussion, Conclusion and Future Directions • The physiological pH can offer greater eye comfort and better drug permeability. • Being a basic drug, NTX molecules exist predominantly in their ionised (the least absorbable) form below physiological pH. • Drug loss from the cul-de-sac of the eye on cheek and via nasolacrimal drainage can be exacerbated following instillation of drug solutions buffered below physiological pH due to increased drainage and induced lacrimation (Ahmed & Patton, 1984; Lang et al., 2002). • Prolonged tear film pH depression can reduce corneal permeability (Francoeur et al., 1983; Rathore & Majumdar, 2006). Therefore, the propective research will study the influence of various stabilisers on the chemical stability of NTX. With respect to the physical stability of the prepared niosomes, there are two approaches to improve the physical stability of lipid vesicles at ambient conditions. Dispersing liposomes in a viscous gel has been used either to reduce the rapid leakage of the encapsulated drug from liposomes (Meisnera & Mezeib, 1995) or to minimise the burst release effect observed with liposomes (Mehanna et al., 2009, 2010). This approach may be extrapolated to improve the physical stability of niosomes. However, the drug release from such a system is likely to be complex, as the drug molecules have to release from the bilayer membranes and diffuse through the viscous gel within the short ocular residence time. Hence, the improvement in the physical stability of the niosomes can be offset by reducing the ocular bioavailability of the administered drug in such vehicles. Additionally, a topical application of a viscolised gel is less convenient than eye drops in terms of ocular administration and the adjustment of the dose. In addition to the previous approach, converting the final liposomal/niosomal liquid dispersion to a powder form by lyophilisation (freeze-drying) or spray-drying not only enhances the physical stability of the vesicles (Ingvarsson et al., 2011), but can also dramatically reduce the oxidative instability of oxidisable drug molecules, by minimising the formation of hydroxyl free radicals (Uchegbu & Vyas, 1998; Waterman et al., 2002). These advantages can be offset by high production costs; long production time; and ending up with more complicated formulations containing cryoprotectants. The ready-to-use eye drops are the easiest and most convenient for producing and administering a dosage form. Accordingly, the 242 General Discussion, Conclusion and Future Directions direction of our future research on niosomes will be concerned with studying different strategies for improving the physical stability of niosomes. An ophthalmic dosage form must be sterile; however, the niosomes used in this thesis were prepared under non-sterile conditions because the study is still in the early stages of formulation development. Sterilisation of the final lipid-based (niosomes/liposomes) products is a challenging endeavour. Heat sterilisation is not recommended for lipid-based delivery systems (Zuidam et al., 1993). Dry heat and steam sterilisation are unsuitable and destructive for lipid-based formulations with a gel/liquid transition temperature which is relatively lower than temperature involved in heat sterilisation, since this may cause extensive drug leakage from the bilayer vesicles. Membrane filtration is also unsuitable for particulate delivery systems larger than the pore size (0.22 µm) of the bacterial filters. On the other hand, preparation under aseptic conditions, gas (ethylene oxide) and gamma irradiation could be useful, since minimal heat is generated during the sterilisation process. Liposomes and niosomes are prepared in a sterile form by passing all organic lipid, buffer and drug solutions through bacterial filters; and then the final product is prepared under aseptic conditions (Abdelbary & El-gendy, 2008; Hathout et al., 2007). Gamma irradiation has high penetration power, so it can be applied to a packaged product and is applicable to heat-sensitive drugs. Many marketed pharmaceuticals are gamma radiationsterilised, such as ophthalmic ointments, drops and injectables (Waterman et al., 2002). These sterilisation methods could be of a potential value for producing sterile niosomes, and hence the effect of gamma irradiation on the physical stability of niosomes could be a potential research line for future work. Future work will also look at studying precorneal retention time in vivo (in rabbit models) by evaluation of the nasolacrimal drainage of encapsulated radioisotopes in niosomes and discomes, and compared with an aqueous solution of the radioactive material using a gamma camera. This is in order to evaluate the ability of niosomes and discomes to prolong the precorneal residence time, compared with aqueous solution. In conclusion, the future of NTX in diabetic keratopathy has been documented (Abdelkader et al., 2011; McLaughlin et al., 2010). NTX competitively blocks the interaction between OGF and OGFr (ζ-opioid receptors) on the basal corneal epithelial cells. This leads to a marked increase in DNA synthesis, cell division and cell migration as well as regeneration of corneal epithelium and 243 General Discussion, Conclusion and Future Directions improvement of diabetic keratopathy (McLaughlin et al., 2010; Zagon et al., 1997; Zagon et al., 2000). Other diabetic complications are often experienced by diabetic patients are keratoconjunctivitis sicca or commonly known as dry eye symptoms and delayed corneal sensation. Recently, NTX has demonstrated an ability to reverse dry eye symptoms and improve delayed diabetic corneal sensation in diabetic rat models (Zagon et al., 2009). However, the exact mechanism for these effects is still unknown (McLaughlin et al., 2010). It is worth mentioning that there is an ongoing collaboration between the Department of Ophthalmology and our research group in the School of Pharmacy to test the hypothesis that topical ocular application of insulin or naltrexone promotes corneal nerve regeneration. Diabetic rats will be used to model neurotrophic keratopathy and non-invasive in vivo confocal microscopy will be utilised to image and quantitatively analyse the corneal nerves non-invasively over time. 244 References References Abdelbary, G., & El-gendy, N. (2008). Niosome-encapsulated gentamicin for ophthalmic controlled delivery. AAPS PharmSciTech, 9, 740-747. Abdelkader, H., Patel, D., McGhee, C., & Alany, R. G. (2011). New therapeutic approaches in treatment of diabetic keratopathy. Clinical & Experimental Opthalmology, 39, 259-270. Adibkia, K., Omidi, Y., Siahi, M., Javadzadeh, A., Barzegar-Jalali, M., Barar, J., Maleki, N., Mohammadi, G., & Nokhodchi, A. (2007). Inhibition of endotoxin-induced uveitis by methylprednisolone acetate nanosuspension in rabbits. Journal of Ocular Pharmacology and Therapeutics, 23, 421 -432. Aggarwal, D., Garg, A., & Kaur, I. P. (2004). Development of a topical niosomal preparation of acetazolamide: preparation and evaluation. Journal of Pharmacy and Pharmacology, 56, 1509–1517. Ahmed, I., & Patton, T. F. (1984). Effect of pH and buffers on ocular penetration of pilocarbine in rabbits. International Journal of Pharmaceutics, 19, 215-227. Ahmed, I., & Patton, T. F. (1987). Disposition of timolol and inulin in the rabbit eye following corneal versus non-corneal absorption. International Journal of Pharmaceutics, 38, 9-21. Ahmed, I., & Patton, T. F. (1989). Importance of the noncorneal absorption route in topical ophthalmic drug delivery. Investigative Ophthalmology & Visual Science, 26, 584-587. Akil, H. S., Watson, S. J., Young, E., Lewis, M. E., Khatchaturian, H., & Walker, J. M. (1984). Endogenous opioid: biology and function. Annual Review of Neuroscience, 7, 223-255. Akinci, A., Cetinkaya, E., & Aycan, Z. (2007). Dry eye syndrome in diabetic children. European Journal of Ophthalmology, 17, 873-878. Alany, R. G., Rades, T., Nicoll, J., Tucker, I. G., & Davies, N. M. (2006). W/O microemulsions for ocular delivery: Evaluation of ocular irritation and precorneal retention. Journal of Controlled Release, 111, 145–152. Albert, A., & Serjeant, E. P. (1971). Introduction. In A. Albert & E. P. Serjeant (Eds.), The Determination of Ionization Constants (pp. 1-8). London: Chapman and Hall Ltd. Ali, Y., & Lehmussaari, K. (2006). Industrial perspective in ocular drug delivery. Advanced Drug Delivery Reviews, 58, 1258–1268. Allen, S. J., Robertson, A. G., Tyler, S. J., Wilcock, G. K., & Dawbarn, D. (2001). Recombinant human nerve growth factor for clinical trials: protein expression, purification, stability and 245 References characterisation of binding to infusion pumps. Journal of Biochemical and Biophysical Methods, 47, 239-255. Alsante, K. M., Martin, L., & Baertschi, S. W. (2003). A stress testing benchmarking study. Pharmaceutical Technology February, 60-72. Anderson, D., & Russell, T. (1995). In D. Anderson & T. Russell (Eds.), The Status of Alternative methods in Toxicology (pp. 21-33). Cambridge, UK: The Royal Scociety of Chemsitry. Anderson, J. A. (1980). Systemic absorption of topical ocularly applied epinephrine and dipivefrin. Archives of Ophthalmology, 98, 350-353. Apfel, S. C. (2002). Nerve growth factor for the treatment of diabetic neuropathy: What went wrong, what went right, and what does the future hold? International Review of Neurobiology, 50, 393–413. Arunothayanun, P., Turton, J. A., Uchegbu, I. F., & Florence, A. T. (1999). Preparation and in vitro/in vivo evaluation of luteinizing hormone releasing hormone (LHRH)-loaded polyhedral and spherical/tubular niosomes. Journal of Pharmaceutical Sciences, 88, 3438. Attama, A. A., Reichl, S., & Muller-Goymann, C. C. (2009). Sustained release and permeation of timolol from surface-modified solid lipid nanoparticles through bioengineered human cornea. Current Eye Research, 34, 698-705. Auerbach, R., Kubai, L., Knighton, D., & Folkman, J. (1974). A simple procedure for the longterm cultivation of chicken embryos. Developmental biology, 41, 391-394. Avalos, J., Jacobs, A., & Wilkin, J. K. (1997). Toxicty testing for ocular drug products. In K. Green (Ed.), Advances in Ocular Toxicology (pp. 261-268). New York: Plenum. Azeem, A., Anwer, K., & Talegaonkar, S. (2009). Niosomes in sustained and targeted drug delivery: some recent advances. Journal of Drug Targeting, 17, 671-689. Azmin, M. N., Florence, A. T., Handjani-vila, R. M., Stuart, J. F. B., Vanlerrberghe, G., & Whittaker, J. S. (1985). The effect of non-ionic surfactant vesicle (niosome) entrapment on the absorption and distribution of methotrexate in mice. Journal of Pharmacy and Pharmacology, 37, 237-242. Azmin, M. N., Florence, A. T., Handjani-vila, R. M., Stuart, J. F. B., Vanlerrberghe, G., & Whittaker, J. S. (1986). The effect of niosome and polysorbate 80 on the metabolism and excretion of methotrexate in mouse. Journal of Microencapsulation, 3, 95-100. 246 References Badawi, A. A., El-Laithy, H. M., El-Qidra, R. K., El mofty, H., & El dally, M. (2008). Chitosan based nanocarriers for indomethacin ocular delivery. Archives of pharmacal research 31, 1040-1049. Baeyens, V., Felt-Baeyens, O., Rougier, S., Pheulpin, S., Boisramé, B., & Gurny, R. (2002). Clinical evaluation of bioadhesive ophthalmic drug inserts (BODI®) for the treatment of external ocular infections in dogs. Journal of Controlled Release, 85, 163-168. Baeyens, V., Kaltsatos, V., Boisramé, B., Varesio, E., Veuthey, J. L., Fathi, M., Balant, L. P., Gex-Fabry, M., & Gurny, R. (1998). Optimized release of dexamethasone and gentamicin from a soluble ocular insert for the treatment of external ophthalmic infections. Journal of Controlled Release, 52, 215-220. Bagley, D. M., Rizvi, P. Y., Kong, B. M., & Salva, S. J. (1991a). Evaluation of the vascular components of the chorioallantoic membrane assay as a model for eye irritation potential: II. Cutaneous and Ocular Toxicology, 10, 105 - 113. Bagley, D. M., Rizvi, P. Y., Kong, B. M., & Salva, S. J. (1991b). Factors affecting use of the hen's egg chorioallantoic membrane as a model for predicting eye irritation potential: I. Cutaneous and Ocular Toxicology, 10, 95 -104. Baillie, A. J., Coombs, G. H., Dolan, T. F., & Laurie, J. (1986). Non-ionic surfactant vesicles, niosomes, as a delivery system for the anti-leshmanial drug, sodium stibogluconate. Journal of Pharmacy and Pharmacology, 38, 502-505. Baillie, A. J., Florence, A. T., Hume, L. R., Muirhead, G. T., & Rogerson, A. (1985). The preparation and properties of niosomes-non-ionic surfactant vesicles. Journal of Pharmacy and Pharmacology, 37, 863-868. Balakrishnana, P., Shanmugama, S., Lee, W. S., Lee, W. M., Kim, J., Oh, D. H., Kim, D. D., Kim, J. S., Yoo, B. K., Choia, H. G., & Yong, C. S. (2009). Formulation and in vitro assessment of minoxidil niosomes for enhanced skin delivery. International Journal of Pharmaceutics, 377, 1-8. Bangham, A. D., Standish, M. M., & Watkins, J. C. (1965). Diffusion of univalent ions across the lamellae of swollen phospholipids. Journal of Molecular Biology, 13, 238–252. Barnett, J., Chow, J., Nguyen, B., Eggers, D., Osen, E., Jarnagin, K., Saldou, N., Straub, K., Gu, L., & Erdos, L. (1991). Physicochemical characterization of recombinant human nerve growth factor produced in insect cells with a baculovirus vector. Journal of Neurochemistry, 57, 1052–1061. 247 References Basu, P. K. (1984). Toxic effects of drugs on the corneal epithelium: a review. Cutaneous and Ocular Toxicology 2, 205-227. Basu, P. K., Avaria, M., & Jankie, R. (1981). Effect of hydrocortisone on the mobilization of leucocytes on the corneal wounds. British Journal of Ophthalmology, 65, 694-698. Bayindir, Z. S., & Yuksel, N. (2010). Characterization of niosomes prepared with various nonionic surfactants for paclitaxel oral delivery. Journal of Pharmaceutical Sciences, 99, 2049 - 2060. Benoliel, A. M., Kahn-Perles, B., Imbert, J., & Verrando, P. (1987). Insulin stimulates haptotactic migration of human epidermal keratinocytes through activation of NF-kappa B transcription factor. Journal of Cell Science, 110, 2089–2097. Blumberg, H., & Dayton, H. B. (1973). Naloxone, naltrexone and related noroxymorphones. In M. C. Braude, L. S. Harris, E. L. May, J. P. Smith & J. E. Villareal (Eds.), Narcotic Antagonists. Advances in Biochemical Psychopharmacology (Vol. 8, pp. 33-43). New York: Raven Press. Bochot, A., Fattal, E., Boutet, V., Deverre, J. R., Jeanny, J. C., Chacun, H., & Couvreur, P. (2002). Intravitreal delivery of oligonucleotides by sterically stabilized liposomes. Investigative Ophthalmology & Visual Science, 43, 253–259. Bochot, A., Fattal, E., Grossiord, J. L., Puisieux, F., & Couvreur, P. (1998). Characterization of a new ocular delivery system based on a dispersion of liposomes in a thermosensitive gel. International Journal of Pharmaceutics, 162, 119-127. Bonini, S., Lambiase, A., Rama, P., Capriogli, G., & Aloe, L. (2000). Topical Treatment with Nerve Growth Factor for Neurotrophic Keratitis. Ophthalmology, 107, 1347–1352. Bourges, J. L., Gautier, S. E., Delie, F., Bejjani, R. A., Jeanny, J. C., Gurny, R., BenEzra, D., & Behar-Cohen, F. (2003). Ocular drug delivery targeting the retina and retinal pigment epithelium using polylactide nanoparticles. Investigative Ophthalmology & Visual Science, 44, 3562–3569. Boyd, B. J. (2003). Characterisation of drug release from cubosomes using the pressure ultrafiltration method. International Journal of Pharmaceutics, 260, 239–247. Brown, S., Lamberts, D., Reid, T., Nishida, T., & Murphy, C. (1997). Neurotrophic and anhidrotic keratopathy treated with substance P and insulin-like growth factor 1. Archives of Ophthalmology, 115, 926-927. 248 References Bruner, L. H. (1992). Ocular irritation. In J. M. Frazier (Ed.), In vitro toxicity testing: Applications to safety evaluation (pp. 149-190). New York: CRC Press. Bruton, A. B. D., York, M., & Lawrence, R. (1981). The in vitro assessment of severe eye irritants Food and Cosmetics Toxicology, 31, 471-480. Budai, P., Lehel, J., Tavaszi, J., & Kormos, E. (2010). HET-CAM test for determining the possible eye irritancy of pesticides. Acta Veterinaria Hungarica, 58, 369 - 377. Buehler, J. D. (2003). Cholesterol. In R. C. Rowe, P. L. Sheskey & P. L. Weller (Eds.), Handbook of Pharmaceutical Excipients (4th ed., pp. 155 - 157). London: Pharmaceutical Press. Burgalassi, S., Chetoni, P., Panichi, L., Boldrini, E., & Saettone, M. F. (2000). Xyloglucan as a novel vehicle for timolol: pharmacokinetics and pressure lowering activity in rabbits. Journal of Ocular Pharmacology and Therapeutics, 16, 497–509. Burgess, D. L. (2010). Drug dissolution: method development. Paper presented at the The Australian & New Zealand Chapters of the Controlled Release Society: In vitro drug release and dissolution testing workshop, Auckland, New Zealand. Cavalli, R., Gasco, M., Chetoni, P., Burgalassi, S., & Saettone, M. (2002). Solid lipid nanoparticles (SLN) as ocular delivery system for tobramycin. International Journal of Pharmaceutics, 238, 241-245. Cavalli, R., Morel, S., Gasco, M. R., Saettone, M. F., & Chetoni, P. (1995). Preparation and evaluation in vitro of colloidal lipospheres containing pilocarpine as ion-pair. International Journal of Pharmaceutics, 117, 243–246. Cavanagh, H. D., Pihlaja, D., Thoft, R. A., & Dohlman, C. H. (1976). The pathogenesis and treatment of persistent epithelial defect. Transactions of the American Academy of Ophthalmology and Otolaryngology, 81, 754–769. Cevc, G., & Blume, G. (1992). Lipid vesicles penetrate into intact skin owing to the transdermal osmotic gradients and hydration force. Biochimica et Biophysica Acta, 1104, 226–232. Cevc, G., Blume, G., Schatzlein, A., Gebauer, D., & Paul, A. (1996). The skin: a pathway for systemic treatment with patches and lipid-based agent carriers. Advanced Drug Delivery Reviews, 18, 349-378. Chan, J., El Maghraby, G. M., Craig, J. P., & Alany, R. G. (2007). Phase transition water-in-oil microemulsions as ocular drug delivery systems: In vitro and in vivo evaluation. International Journal of Pharmaceutics, 328, 65-71. 249 References Chan, J., El Maghraby, G. M., Craig, J. P., & Alany, R. G. (2008). Effect of water-in-oil microemulsions and lamellar liquid crystalline systems on the precorneal tear film of albino New Zealand rabbits. Journal of Clinical Ophthalmology, 2, 129-137. Chang, C. Y., Green, C. R., McGhee, C. N. J., & Sherwin, T. (2008). Acute wound healing in the human central corneal epithelium appears to be independent of limbal stem cell influence. Investigative Ophthalmology & Visual Science, 49, 5279-5286. Chang, S. C., & Lee, V. H. (1987). Nasal and conjunctival contributions to the systemic absorption of topical timolol in the pigmented rabbit: implications in the design of strategies to maximize the ratio of ocular to systemic absorption. Journal of Ocular Pharmacology, 3, 159-169. Cheng, F., McLaughlin, P. J., Banks, W. A., & Zagon, I. S. (2009). Passive diffusion of naltrexone into human and animal cells and upregulation of cell proliferation. American Journal of Physiology, 297, R844 - R852. Cheng, Y., Xu, Z., Ma, M., & Xu, T. (2008). Dendrimers as drug carriers: applications in different routes of drug administration. Journal of Pharmaceutical Sciences 97, 123-143. Chetoni, P., Burgalassi, S., Monti, D., Najarro, M., & Boldrini, E. (2007). Liposomeencapsulated mitomycin C for the reduction of corneal healing rate and ocular toxicity. Journal of Drug Delivery Science and Technology, 17, 43-48. Chrai, S. S., Makoid, M. C., Eriksen, S. P., & Robinson, J. R. (1974). Drop size and initial dosing frequency problems of topically applied ophthalmic drugs. Journal of Pharmaceutical Sciences, 63, 333-338. Chrai, S. S., Patton, T. F., Mehta, A., & Robinson, J. R. (1973). Lacrimal and instilled fluid dynamics in rabbit eyes. Journal of Pharmaceutical Sciences, 62, 1112-1121. Chrai, S. S., & Robinson, J. R. (1974). Ocular evaluation of methylcellulose vehicle in albino rabbits. Journal of Pharmaceutical Sciences, 63, 1218–1223. Cisarik-Fredenburg, P. (2001). Discoveries in research on diabetic keratopathy. Optometry, 72, 691-704. Colangelo, A. M., Finotti, N., Ceriani, M., Alberghina, L., Martegani, E., Aloe, L., Lenzi, L., & Levi-Montalcini, R. (2005). Recombinant human nerve growth factor with a marked activity in vitro and in vivo. Proceedings of the National Academy of Sciences, 102, 18658-18663. 250 References Cooper, K. J., Earl, L. K., Harbell, J. W., & Raabe, H. A. (2001). Prediction of ocular irritancy of prototype shampoo formulations by the isolated rabbit eye (IRE) test and bovine corneal opacity and permeability (BCOP) assay. Toxicology in Vitro, 15, 95-103. Crabtree, B. (1984). Review of naltrexone, a long-acting opioate antagonist. Clinical Pharmacy, 3, 273-279. Craig, J. P., Purslow, C., Murphy, P. J., & Wolffsohn, J. S. W. (2010). Effect of a liposomal spray on the pre-ocular tear film. Contact Lens & Anterior Eye, 33, 83–87. Cuellar, N., Merrill, J., Clear, M., Mun, G., & Harbell, J. W. (2002). The application of Benchmarks for the evaluation of the potential ocular irritancy of aerosol fragrances. The Toxicologist, 66, 243-244. Curren, D. R., Evans, M. G., Raabe, H. A., Ruppalt, R. R., & Harbell, J. W. (2000). Correlation of histopathology, opacity and permeability of bovine corneas exposed in vitro to known ocular irritants. Veterinary Pathology, 37, 557. Curren, R. D., & Harbell, J. W. (1998). In vitro alternatives for ocular irritation. Environmental Health Perspectives, 106, 485-492. Daniele, S., Gilbard, J., & Schepens, C. (1992). Treatment of persistent epithelial defects in neurotrophic keratitis with epidermal growth factor: a preliminary open study. Graefe's Archive for Clinical and Experimental Ophthalmology, 230, 314–317. David, T., Rieck, P., Renard, G., Hartmann, C., Courtois, Y., & Pouliquen, Y. (1995). Corneal woun healing modulation using basic fibroblast growth factor after excimer lasephotorefractive keratectomy. Cornea, 14, 227–234. Dawson, D. G., & Edelhauser, H. F. (2010). Corneal edema. In A. L. Levin & D. M. Albert (Eds.), Ocular Disease: Mechanisms and Management (pp. 64-73). New York: Elsevier. Dogru, M., Katakami, C., & Inoue, M. (2001). Tear function and ocular surface changes in noninsulin-dependent diabetes mellitus. Ophthalmology, 108, 586-592. Draize, J. H., Woodward, G., & Calvery, H. O. (1944). Methods for the study of irritation and toxicity of substances applied topically to the skin and mucus membranes. Journal of Pharmacology and Experimental Therapeutics, 82, 377-390. du Toit, L. C., Pillay, V., Choonara, Y. E., Govender, T., & Carmichael, T. (2011). Ocular drug delivery – a look towards nanobioadhesives. Expert Opinion on Drug Delivery, 8, 71-94. 251 References Duzman, E., Anderson, J. A., Vita, J. B., Lue, J. C., Chen, C. C., & Leopold, I. (1983). Topically applied oxymetazoline. Ocular vasoconstrictive activity, pharmacokinetics, and metabolism. Archives of Ophthalmology, 101, 1122-1126. Edwards, R. H., Selby, M. J., Garcia, P. D., & Rutter, W. J. (1988). Processing of the native growth factor precursor to form biologically active nerve growth factor. The Journal of Biological Chemistry, 263, 6810–6815. Egelhaaf, R. M., Epand, R. F., & Maekawa, S. (2003). The arrangement of cholesterol in membranes and binding of NAP-22. Chemistry and Physics of Lipids, 122, 33-39. Ehlers, N., & Hjortdal, J. (2006). The cornea: epithelium and stroma. In E. Bittar (Ed.), Advances in Organ Biology (Vol. 10, pp. 83 - 111). New York: Elsevier. El-Gazayerly, O. N., & Hikal, A. H. (1997). Preparation and evaluation of acetazolamide liposomes as ocular delivery systems. International Journal of Pharmaceutics, 158, 121– 127. El Gamal, S., Naggar, V., & Allam, A. N. (2008). Formulation and evaluation of acyclovir ophthalmic inserts. Asian Journal of Pharmaceutical Sciences, 3, 58-67. El Maghraby, G. M. (2010). Self-microemulsifying and microemulsion systems for transdermal delivery of indomethacin: effect of phase transition. Colloids and Surfaces B: Biointerfaces, 75, 595-600. El Maghraby, G. M., & Williams, A. C. (2009). Vesicular systems for delivering conventional small organic molecules and larger macromolecules to and through human skin. Expert Opinion on Drug Delivery, 6, 149-163. El Maghraby, G. M., Williams, A. C., & Barry, B. W. (2004). Interactions of surfactants (edge activators) and skin penetration enhancers with liposomes. International Journal of Pharmaceutics, 276, 143-161. Elsayed, M. M., Abdallah, O. Y., Naggar, V., & Khalafallah, N. M. (2007). Lipid vesicles for skin delivery of drugs: Reviewing three decades of research. International Journal of Pharmaceutics, 332, 1-16. Essa, A. E. (2010). Effect of formulation and processing variables on the particle size of sorbitan monopalmitate niosomes. Asian Journal of Pharmaceutics, 4, 227-233. Fang, J. Y., Yu, S. Y., Wu, P. C., Huang, Y. B., & Tsai, Y. H. (2001). In vitro skin permeation of estradiol from various proniosome formulations. International Journal of Pharmaceutics, 215, 91–99. 252 References Fattal, E., & Bochot, A. (2006). Ocular delivery of nucleic acids: antisense oligonucleotides, aptamers and siRNA. Advanced Drug Delivery Reviews, 58, 1203-1223 Fawcett, J. P., Morgan, N. C., & Woods, D. J. (1997). Formulation and stability of naltrexone oral liquid for rapid withdrawal from methadone. The Annals of Pharmacotherapy, 31, 1291-1295. Feman, S. S. (2000). Diabetes and the eye. In W. Tasman (Ed.), Duanes’s Clinical Opthalmology. (Vol. 5, pp. 1-10). Philadelphia: Lippincott Williams & Wilkins. Fink, M., Zaks, A., Sharoff, R., Mora, A., Bruner, A., Levit, S., & Freedman, A. M. (1968). Naloxone in heroin dependence. Clinical Pharmacology and Therapeutics 9, 568-577. Fitzgerald, P., Hadgraft, J., Kreuter, J., & Wilson, C. G. (1987a). A gamma-scintigraphic evaluation of microparticulate ophthalmic delivery systems: liposomes and nanoparticles. International Journal of Pharmaceutics, 40, 81-84. Fitzgerald, P., Hadgraft, J., & Wilson, C. G. (1987b). A gamma scintigraphic evaluation of the precomeal residence of liposomal formulations in the rabbit. Journal of Pharmacy and Pharmacology, 39, 487-490. Florence, A. T., & Attwood, D. (1998a). Drug stability. In A. T. Florence & D. Attwood (Eds.), Physicochemical Principles of Pharmacy (3rd ed., pp. 101-151). London: Macmillan Press, Ltd. Florence, A. T., & Attwood, D. (1998b). Physical properties of drugs in solution. In A. T. Florence & D. Attwood (Eds.), Physicochemical Principles of Pharmacy (3rd ed., pp. 5699). London: McMillan Press Ltd. Florence, A. T., & Attwood, D. (1998c). Properties of the solid state. In A. T. Florence & D. Attwood (Eds.), Physichochemical Principles of Pharmacy (pp. 5-35). London: McMillan press, Ltd. Florence, A. T., & Attwood, D. (1998d). The solubility of drugs. In A. T. Florence & D. Attwood (Eds.), Physichochemical Principles of Pharmacy (pp. 152-197). London: McMillan Press, Ltd. Florence, A. T., & Baillie, A. J. (1989). Non-ionic surfactant vesicles alternatives to liposomes in drug delivery? In L. E. Prescott & W. S. Nimmo (Eds.), Novel Drug Delivery and Its Therapeutic Application (pp. 281-296). San Francisco, California: John Wiley & Sons, Ltd. 253 References Foster, C. S., Lass, J. H., Moran-Wallace, K., & Giovanoni, R. (1981). Ocular toxicity of topical antifungal agents. Archives of Ophthalmology, 99, 1081-1084. Francoeur, M., Ahmed, I., Sitek, S., & Patton, T. F. (1983). Age-related differences in ophthalmic drug disposition III. Corneal permeability of pilocarpine in rabbits. International Journal of Pharmaceutics, 16, 203-213. Francouer, M., Ahmed, I., Sitek, S., & Patton, T. F. (1983). Age-related differences in ophthalmic drug disposition. III. Corneal permeability of pilocarpine in rabbits. International Journal of Pharmaceutics, 16, 203-213. Frank, R. N. (2004). Diabetic retinopathy. The New England Journal of Medicine, 350, 48–58. Freedman, A., Fink, M., Sharoff, R., & Zaks, A. (1968). Clinical studies of cyclaxocine in the treatment of narcotic addiction. The American Journal of Psychiatry, 124, 1499-1504. Freitas, C., & Muller, R. H. (1999). Correlation between long-term stability of solid lipid nanoparticles (SLN) and crystallinity of the lipid phase. European Journal of Pharmaceutics and Biopharmaceutics, 47, 125–132. Fuentes, J. A., Caraballo, I., Boza, A., Llera, J. M., Holgado, M. A., & Arevalo, M. F. (1997). Study of a complexation process between naltrexone and Eudragit®L as an oral controlled release system. International Journal of Pharmaceutics, 148, 219 -230. Fuller, M. A., & Sajatovic, M. (2009). Naltrexone. In M. A. Fuller & M. Sajatovic (Eds.), Drug Information Handbook for Psychiatry (7th ed., pp. 1098-1100). Hudson, Ohio: LexiComp. Furrer, P., Mayer, J. M., & Gurny, R. (2002a). Ocular tolerance of preservatives and alternatives. European Journal of Pharmaceutics and Biopharmaceutics, 53, 263-280. Furrer, P., Mayer, J. M., Plazonnet, B., & Gurny, R. (2002b). Ocular tolerance of absorption enhancers in ophthalmic preparations. AAPS PharmSci 4, 1 - 5. Furrer, P., Plazonnet, B., Mayer, J. M., & Gurny, R. (2000). Application of in vivo confocal microscopy to the objective evaluation of ocular irritation induced by surfactants International Journal of Pharmaceutics, 207, 89-98. Galichet, L. Y., Moffat, A. C., Osselton, M. D., & Widdop, B. (2006). Naltrexone. In L. Y. Galichet, A. C. Moffat, M. D. Osselton & B. Widdop (Eds.), Clarke's aAnalysis of Drugs and Poisons. New York: Pharmaceutical press. Gan, L., Han, S., Shen, J., Zhu, J., Zhu, C., Zhang, X., & Gan, Y. (2010). Self-assembled liquid crystalline nanoparticles as a novel ophthalmic delivery system for 254 References dexamethasone:improving preocular retention and ocular bioavailability. International Journal of Pharmaceutics, 396, 179-187. Gasset, A. R., Braverman, L. E., Fleming, M. C., Arky, R. A., & Alter, B. R. (1968). Tear glucose detection of hyperglycemia. American Journal of Ophthalmology, 65, 414-420. Gautheron, P., Dukic, M., Alix, D., & Sina, J. F. (1992). Bovine corneal opacity and permeability test: an in vitro assay of ocular irritancy. Fundamental and Applied Toxicology, 18, 442449. German, E., & Siddiqui, N. (1970). Atropine toxicity from eye drops. The New England Journal of Medicine, 282, 689. Geroski, D. H., & Edelhauser, H. F. (2001). Transscleral drug delivery for posterior segment disease. Advanced Drug Delivery Reviews, 52, 37-48. Ghinelli, E., Johansson, J., Ríos, J., Chen, L., Zoukhri, D., Hodges, R. R., & Dart, D. A. (2003). Presence and localization of neurotrophins and neurotrophin receptors in rat lacrimal gland. Investigative Ophthalmology & Visual Science, 44, 3352–3357. Giardino, L., Giuliani, A., Battaglia, A., Carfagna, N., Aloe, L., & Calza, L. (2002). Neuroprotection and aging of the cholinergic system: a role for the ergoline derivative nicergoline (Sermion). Neuroscience, 109, 487– 497. Girigoswami, A., Das, S., & De, S. (2006). Fluorescence and dynamic light scattering studies of niosomes-membrane mimetic systems. Spectrochimica Acta A, 64, 859-866. Gleditsch, E., & Waaler, P. J. (2001). Accelerated stability studies of morphine injections in plastic ampoules. International Journal of Pharmaceutics, 212, 275-287. Gobbels, M., Spitznas, M., & Oldendoerp, J. (1989). Impairment of corneal epithelial barrier function in diabetics. Graefe's Archive for Clinical and Experimental Ophthalmology, 227, 142-144. Gockerman, A., Prevette, T., Jones, J., & Clemmons, D. (1995). Insulin-like growth factor (IGF)binding proteins inhibit the smooth muscle cell migration responses to IGF-1 and IGF-II. Endocrinology, 136, 41168–41173. Gonzalez-Mira, E., Egea, M. A., Souto, E. B., Calpena, A. C., & Garcia, M. L. (2011). Optimizing flurbiprofen-loaded NLC by central composite factorial design for ocular delivery. Nanotechnology, 22, 1-15. Goodman, A. J., Bourdonnec, B. L., & Dolle, R. E. (2007). Mu opioid receptor antagonists: recent developments. Chem Med Chem, 2, 1552 – 1570. 255 References Grant, G. J., Barenholz, Y., Piskoun, B., Bansinath, M., Turndorf, H., & Bolotin, E. M. (2001). DRV liposomal bupivacaine: preparation, characterization, and in vivo evaluation in mice. Pharmaceutical Research, 18, 336-343. Greaves, J. L., & Wilson, C. G. (1993). Treatment of diseases of the eye with mucoadhesive delivery systems. Advanced Drug Delivery Reviews, 11, 349-383. Greaves, J. L., Wilson, C. G., Rozier, A., Grove, J., & Plazonnet, B. (1990). Scintigraphic assessment of an ophthalmic gelling vehicle in man and rabbit. Current Eye Research, 9, 415-420. Green, K., & Downs, S. J. (1974). Prednisolone phosphate penetration into and through the cornea. Investigative Ophthalmology & Visual Science, 13, 16--319. Gregoriadis, G., & Florence, A. T. (1993). Liposomes in drug delivery: clinical, diagnostic and ophthalmic potential Drugs, 45, 15–28. Greish, K., Herd, H., Thiagarajan , G., Price, R., Bauer, H., Yu, T., Anwar, A., & Ghandehari, H. (2010). In Vivo Evaluation of Nanomaterial Biosafety: Evidence of Disseminated Intravascular Coagulopathy (DIC) in Response to Cationic Dendrimer Administration. Paper presented at the 37th Annual Meeting & Exposition of the Controlled Release Society, Portland, Oregon, USA. Griffith, J. F., Nixon, G. A., Bruce, R. D., Reer, P. J., & Bannan, E. A. (1980). Dose-response studies with chemical irritants in the albino rabbit eye as a bsis for selecting optimum conditions for predicting hazards to human eye. Toxicology and Applied Pharmacology, 55, 501-513. Groos, E. (1997). Cornea: Fundamentals of Corneal and External Disease. In J. H. Krachmer, M. J. Mannis & E. J. Hooland (Eds.), Neurotrophic keratitis (pp. 1339–1362). St. Louis: Mosby. Guillory, J. K., & Poust, R. I. (2002). Chemical kinetics and drug stability. In G. S. Banker & C. T. Rhodes (Eds.), Modern Pharmaceutics (pp. 135-162). New York: Marcel Dekker, Inc. Guinedi, A. S., Mortada, N. D., Mansour, S., & Hathout, R. M. (2005). Preparation and evaluation of reverse-phase evaporation and multilamellar niosomes as ophthalmic carriers of acetazolamide. International Journal of Pharmaceutics, 306, 71-82. Gupta, V. D. (2008). Chemical stability of naltrexone hydrochloride injection. International Journal of Pharmaceutical Compounding, 12, 274-275. 256 References Gurny, R. (1981). Preliminary study of prolonged acting drug delivery system for the treatment of glaucoma. Pharmaceutica Acta Helvetiae, 56, 130-132. Gurny, R., Ibrahim, H., Aebi, A., Buri, P., Wilson, C. G., Washington, N., Edman, P., & Camber, O. (1987). Design and evaluation of controlled release systems for the eye. Journal of Controlled Release, 6, 367-373. Gurtler, F., & Gurny, R. (1995). Patent literature review of ophthalmic inserts. Drug Development and Industrial Pharmacy, 21, 1-18. Habib, F. S., Fouad, E. A., Abdel-Rhaman, M. S., & Fathalla, D. (2010). Liposomes as an ocular delivery system of fluconazole: in-vitro studies. Acta Ophthalmologica, 88, 901-904. Harbell, J. W., Mun, G., & Curren, R. D. (2006). Application of histological evaluation to enhance the bovine opacity and permeability (BCOP) assay. Paper presented at the The 45th Annual Society of Toxicology Meeting, San Diego, California, USA. Harbell, J. W., Raabe, H. A., Evans, M. G., & Curren, R. D. (1999). Histopathology associated with opacity and permeability changes in bovine corneas in vitro. The Toxicologist, 48, 336-337. Hariprasad, S. M., Shah, G. K., Chi, J., & Prince, R. (2005). Determination of aqueous and vitreous concentration of moxifloxacin 0.5% after delivery via a dissolvable corneal collagen shield device. Journal of Cataract & Refractive Surgery, 31, 2142-2146. Hashim, F., El-Ridy, M., Nasr, M., & Abdallah, Y. (2010). Preparation and characterization of niosomes containing ribavirin for liver targeting. Drug Delivery, 17, 282-287. Hathout, R. M., Mansour, S., Mortada, N. D., & Guinedi, A. S. (2007). Liposomes as an ocular delivery system for acetazolamide: In vitro and in vivo studies. AAPS PharmSciTech, 8 (1), 1-12. Haugen, B., Werner, L., Romaniv, N., Haymore, J., Kleinmann, G., Mamalis, N., & Olson, R. J. (2008). Prevention of endophthalmitis by collagen shields presoaked in fourth-generation fluoroquinolones versus by topical prophylaxis. Journal of Cataract & Refractive Surgery, 34, 853-858. Henkind, P., & Wise, G. N. (1961). Descemet's wrinkles in diabetes. American Journal of Ophthalmology, 52, 371-374. Herse, P. R. (1988). A review of manifestations of diabetes mellitus in the anterior eye and cornea. American Journal of Optometry and Physiological Optics, 65, 224-230. 257 References Hofland, H. E., Bouwstra, J. A., Verhoef, J. C., Buckton, G., Chowdry, B. Z., Ponec, M., & Junginger, H. E. (1992). Safety aspects of non-ionic surfactant vesicles- a toxicity study related to the physichochmical characteristics of non-ionic surfactants. Journal of Pharmacy and Pharmacology, 44, 287-294. Hofland, H. E. J., Bouwstra, J. A., Ponec, M., Bodde, H. F., Spies, F., Verhoef, J. C., & Junginger, H. E. (1991). Interactions of non-ionic surfactant vesicles with cultured keratinocytes and human skin in vitro: a survey of toxicological aspects and ultrastructural changes in stratum corneum. Journal of Controlled Release, 16, 155 - 168. Hongo, M., Itoi, M., Yamaguchi, K., & Imanishi, J. (1992). Distribution of epidermal growth factor (EGF) receptors in rabbit corneal epithelial cells, keratocytes and endothelial cells, and the changes induced by transforming growth factor-beta 1. Experimental Eye Research, 54, 9-16. Hope, M. J., Bally, M. B., Mayer, L. D., Janoff, A. S., & Cullis, P. R. (1986). Generation of multilamellar and unilamellar phospholipid vesicles. Chemistry and Physics of Lipids, 40, 89-107. Hornof, M., Toropainen, E., & Urtti, A. (2005). Cell culture models of the ocular barriers. European Journal of Pharmaceutics and Biopharmaceutics, 60, 207-225. Hosotani, H., Ohashi, Y., Yamada, M., & subota, K. (1995). Reversal of abnormal corneal epithelial cell morphologic characteristics and reduced corneal sensitivity in diabetic patients by aldose reductase inhibitor, CT-112. American Journal of Ophthalmology, 119, 288–294. Huang, H. S., Schoenwald, R. D., & Lach, J. L. (1983). Corneal penetration behavior of Beta blocking Agents II: assessment of barrier contributions. Journal of Pharmaceutical Sciences, 72, 1272-1279. Hughes, B., Feiz, V., Flynn, S. B., & Brodsky, M. C. (2004). Reversible amantadine-induced corneal edema in an adolescent. Cornea, 23, 823-824. Hunter, C. A., Dolan, T. F., Coombs, G. H., & Baillie, A. J. (1988). Vesicular system (niosome & liposome) for delivery of sodium stibogluconate in experimental murine visceral leishmaniasis. Journal of Pharmacy and Pharmacology, 40, 161-165. Hyndiuk, R. A., Kazarian, E. L., Schultz, R. O., & Seideman, S. (1977). Neurotrophic corneal ulcers in diabetes mellitus. Archives of Ophthalmology, 95, 2193-2196. 258 References Ibrahim, A. A., Bosela, A. A., Ahmed, S. M., & Mahrous, G. M. (2005). Proniosomes as a drug carrier for transdermal delivery of ketorolac. European Journal of Pharmaceutics and Biopharmaceutics, 59, 485–490. ICH. (1996). Guidance for Industry: Q2B Validation for Analytical Procedures: Methodology US Department of Health and Human Services Food and Drug Administration. USA: Center for Evaluation and Research. Im, J. H., Pillion, D. J., & Meezan, E. (1986). Comparison of insulin receptors from bovine retinal blood vessels and nonvascular retinal tissue. Investigative Ophthalmology & Visual Science, 27, 1681–1690. Ingvarsson, P. T., Yang, M., Nielsen, H. M., Rantanen, J., & Foged, C. (2011). Stabilization of liposomes during drying. Expert Opinion on Drug Delivery, 8, 375-388. Ishida, I., Rao, G. N., Cerro, M., & Aquavella, J. V. (1984). Corneal nerve alterations in diabetes mellitus. Archives of Ophthalmology, 102, 1380 - 1384. Israelachvili, J. N., Marcelja, S., & Horn, R. G. (1980). Physical principles of membrane organization. Quarterly Reviews of Biophysics, 13, 121-200. Israelachvili, J. N., & Mitchell, D. J. (1975). A model for the packing of lipids in bilayer membranes. Biochimica et Biophysica Acta, 389, 13-19. Israelachvili, J. N., Mitchell, D. J., & Ninham, B. W. (1976). Theory of self-assembly of hydrocarbon amphiphiles into micelles and bilayers. Journal of the Chemical Society, Faraday Transactions II, 72, 1525-1568. Iwane, M., Kitamura, Y., Kaisho, Y., Yoshimura, K., Shintani, A., Sasada, R., Nakagawa, S., Kawahara, K., Nakahama, K., & Kakinuma, A. (1990). Production, purification and characterization of biologically active recombinant human nerve growth factor. Biochemical and Biophysical Research Communications, 171, 116–122. Iyer, S. S., Barr, W. H., & Karnes, T. (2007). Characterization of a potential medium for ‘biorelevant’ in vitro release testing of a naltrexone implant, employing a validated stability-indicating HPLC method. Journal of Pharmaceutical and Biomedical Analysis, 43, 845–853. Jain, C. P., & Vyas, S. P. (1995). Preparation and characterisation of niosomes containing rifampicin for lung targeting. Journal of Microencapsulation, 12, 401–407. Jain, N. k., & Asthana, A. (2007). Dendritic systems in drug delivery applications. Expert Opinion on Drug Delivery, 4, 495-512. 259 References Jain, S., & Vyas, S. P. (2006). Mannosylated niosomes as adjuvant-carrier system for oral mucosal immunization. Journal of Liposome Research, 16, 331–345. Jarvinen, K., Jarvinen, T., Thompson, D. O., & Stella, V. J. (1994). The effect of a modified Pcyclodextrin. SBE4-P-CD, on the aqueous stability and ocular absorption of pilocarpine. Current Eye Research, 13, 897-905. Jarvinen, K., Jarvinen, T., & Urtti, A. (1995). Ocular absorption following topical delivery. Advanced Drug Delivery Reviews, 16, 3-19. Jester, J., Barry, L. P., & Petroll, W. (1997). Inhibition of corneal fibrosis by topical application of blocking antibodies to TGF-beta in the rabbit. Cornea, 16, 177-187. Jiao, J. (2008). Polyoxyethylated nonionic surfactants and their applications in topical ocular drug delivery. Advanced Drug Delivery Reviews, 60, 1663-1673. Johnson, D. M., & Gu, L. C. (1988). Autoxidation and antioxidants. In J. Swarbrick & J. C. Boylan (Eds.), Encyclopedia of Pharmaceutical Technology (pp. 415-449). New York: Marcel Dekker. Joyce, N., & Zieske, J. (1997). Transforming growth factor-beta receptor expression in human cornea. Investigative Ophthalmology & Visual Science, 38, 1922-1928. Junginger, H. E., Hofland, H. E. J., & Bouwstra, J. A. (1991). Liposomes and niosomes: interactions with human skin. Cosmetics & toiletries, 106, 45–50. Junyaprasert, V. B., Teeranachaideekul, V., & Supaperm, T. (2008). Effect of charged and nonionic membrane additives on physichochemical properties and stability of niosomes. AAPS PharmSciTech, 9, 851-859. Kaila, T., Huupponen, R., & Salminen, L. (1986). Effects of eyelid closure and nasolacrimal duct occlusion on the systemic absorption of ocular timolol in human subjects. Journal of Ocular Pharmacology, 2, 365-369. Kaiserman, I., Kaiserman, N., Nakar, S., & Vinker, S. (2005). Dry eye in diabetic patients. American Journal of Ophthalmology, 139, 498-503. Kaji, Y. (2005). Prevention of diabetic keratopathy. British Journal of Ophthalmology, 89, 254255. Kanaya, E., Higashizaki, T., Ozawa, F., Hirai, K., Nishizawa, M., Tokunaga, M., Tsukui, H., Hatanaka, H., & Hishinuma, F. (1989). Synthesis and secretion of human growth factor by Saccharomyces cerevisiae. Gene, 83, 65-74. 260 References Kastrup, E. K., Johnson, P. B., Meives, C. A., & Kennedy, M. (2009). Naltrexone hydrochloride Drug Facts and Comparisons (2010 ed., pp. 585-591). Missouri: Wolters Kluwer Health. Kaufman, J. J., Semo, N. M., & Koski, W. S. (1975). Microelectrometric titration measurement of the pKa's and partition and drug distribution coefficients of narcotics and narcotics antagonists and their pH and temperature dependence. Journal of Medicinal Chemistry, 18, 647-655. Kaur, I. P., Aggarwal, D., & Singh, H. (2010). Improved ocular absorption kinetics of timolol maleate loaded into a bioadhesive niosomal delivery system. Graefe's Archive for Clinical and Experimental Ophthalmology, 248, 1467–1472. Kaur, I. P., Garg, A., & Aggarwal, D. (2004). Vesicular systems in ocular drug delivery: an overview. International Journal of Pharmaceutics, 269, 1–14. Kaur, I. P., & Kanwar, M. (2002). Ocular preparations: the formulation approach. Drug Development and Industrial Pharmacy, 28, 473-493. Kaur, I. P., Rana, C., & Singh, H. (2008). Development of effective ocular preparations of antifungal agents [Review]. Journal of Ocular Pharmacology and Therapeutics, 24, 481493. Kaur, I. P., & Smitha, R. (2002). Penetration enhancers and ocular bioadhesives: two new avenues for ophthalmic drug delivery. Drug Development and Industrial Pharmacy 28, 353-369. Kayser, O., Lemke, A., & Hernández-Trejo, N. (2005). The impact of nanobiotechnology on the development of new drug delivery systems. Current Pharmaceutical Biotechnology, 6, 35. Kennon, L. (1964). Use of models in determining chemical pharmaceutical stability. Journal of Pharmaceutical Sciences 53, 815-818. Kenny, S. J., Aubert, R. E., & Geiss, L. S. (1995). Prevalence and incidence of non-insulindependent diabetes Diabetes in America (2nd ed.). Washington D C: Government Printing Office. Khan, K. A. (1975). The concept of dissolution efficiency. Journal of Pharmacy and Pharmacology, 27, 48-49. Kim, J. G., & Kim, J. D. (1991). Vesicle to micelle transitions of egg phosphatidylcholine liposomes induced by nonionic surfactants, poly(oxyethylene) cetyl ethers. The Journal of Biochemistry, 110, 436-442. 261 References Kim, S. Y., Choi, J. S., & Joo, C. K. (2009). Effects of nicergoline on corneal epithelial wound healing in rat eyes. Investigative Ophthalmology & Visual Science, 50, 621-625. Kim, S. Y., Yang, J., & Lee, Y. C. (2011). The effects of nicergoline on corneal nerve regeneration in rat corneas after photorefractive keratectomy. Current Eye Research, 36(1), 29-33. Kingsley, R., & Marfurt, C. (1997). Topical substance P and corneal epithelial wound closure in the rabbit. Investigative Ophthalmology & Visual Science, 38, 388-395. Kinouchi, Y., Takeichi, Y., & Yata, N. (1996). A novel method for the preclinical assessment of rectal irritation. Journal of Pharmacy and Pharmacology, 48, 310 - 315. Kirby, C., & Gregoriadis, G. (1984a). Dehydration-rehydration vesicles: a simple method for high yield drug entrapment in liposomes. Biotechnology, 2, 979-984. Kirby, C., & Gregoriadis, G. (1984b). A simple procedure for preparing liposomes capable of high encapsulation effeciency under mild conditions. In G. Gregoriadis (Ed.), Liposome Technology (Vol. 1, pp. 19-27). Florida: CRC Press. Kiritoshi, A., SundarRaj, N., & Thoft, R. A. (1991). Differentiation in cultured limbal epithelium as defined by keratin expression. Investigative Ophthalmology & Visual Science, 32, 3073-3077. Kiwada, H., Niimura, H., Fujisaki, Y., Yamada, S., & Kato, Y. (1985). Application of synthetic alkyl glycoside vesicles as drug carriers. I preparation and physical properties. Chemical & Pharmaceutical Bulletin, 33, 753–759. Klocek, M. A., Sassani, J. W., McLaughlin, P. J., & Zagon, I. S. (2007). Topically applied naltrexone restores Corneal reepithelialization in diabetic rats. Journal of Ocular Pharmacology and Therapeutics, 23, 89-102. Klocek, M. S., Sassani, J. W., McLaughlin, P. J., & Zagon, I. S. (2009). Naltrexone and insulin are independently effective but not additive in accelerating corneal epithelial healing in type I diabetic rats. Experimental Eye Research 89, 686–692. Klyce, S. D., & Beuerman, R. W. (1998). Structure and function of the cornea. In H. E. Kaufman, B. A. Barron & G. B. McDonald (Eds.), The Cornea (2nd ed., pp. 3-50). Boston: Butterworth-Heinemann. Korsmeyer, R. W., Gurnya, R., Doelkera, E., Buria, P., & Peppas, N. A. (1983). Mechanisms of solute release from porous hydrophilic polymers. International Journal of Pharmaceutics, 15, 25-35. 262 References Krull, I., & Szulc, M. (1997). Detection sensitivity and selectivity. In L. R. Snyder, J. J. Kirkland & J. L. Glajch (Eds.), Practical HPLC Method Development (2nd ed., pp. 59-99). Hoboken, New Jersey John Wile & Sons, Inc. Kruse, F., & Tseng, S. (1993). Growth factors modulate clonal growth and differentiation of cultured rabbit limbal and corneal epithelium. Investigative Ophthalmology & Visual Science, 34, 1963-1976. Kumar, S., Haglund, B. O., & Himmelstein, K. J. (1994). In situ-forming gels for ophthalmic drug delivery. Journal of Ocular Pharmacology and Therapeutics, 10, 47-56. Kyyrönen, K., & Urtti, A. (1990). Effects of epinephrine pretreatment and solution pH on ocular and systemic absorption of ocularly applied timolol in rabbits. Journal of Pharmaceutical Sciences, 79, 688-691. Lachman, L., DeLuca, P., & Akers, M. J. (1986). Kinetic principles and stability testing. In L. Lachman, H. A. Lieberman & J. L. Kanig (Eds.), The Theory and practice of Industrial Pharmacy (pp. 760-802). Philadelphia: Lea & Febiger. Lambiase, A., Bonini, S., Bonini, S., Micera, A., Margini, L., Bracci-Laudiero, L., & Aloe, L. (1995). Increased plasma levels of nerve growth factor in vernal keratoconjunctivitis and relationship to conjunctival mast cells. Investigative Ophthalmology & Visual Science, 36, 2127-2132. Lambiase, A., Bonini, S., Micera, A., Rama, P., Bonini, S., & Aloe, L. (1998a). Expression of nerve growth factors on the ocular surface in healthy subjects and during manifestation of inflammatory diseases. Investigative Ophthalmology & Visual Science, 39, 1272-1275. Lambiase, A., Manni, L., Rama, P., Bonini, S., Micera, A., & Aloe, L. (2000). Nerve Growth Factor Promotes Corneal Healing: Structural, Biochemical, and Molecular Analyses of Rat and Human Corneas. Investigative Ophthalmology & Visual Science, 14, 1063-1069. Lambiase, A., Rama, P., Aloe, L., & Bonini, S. (1999). Management of neurotrophic keratopathy. Current Opinion in Ophthalmology, 10, 270-276. Lambiase, A., Rama, P., Bonini, S., Caprioglio, G., & Aloe, L. (1998b). Topical treatment with nerve growth factor for corneal neurotrophic ulcers. The New England Journal of Medicine, 338, 1174-1180. Lang, J. C. (1995). Ocular drug delivery conventional ocular formulations. Advanced Drug Delivery Reviews, 16, 39-43. 263 References Lang, J. C., Roehrs, R. E., Rodeheaver, D. P., Missel, P. J., Jani, R., & Chowhan, M. A. (2002). Design and evaluation of ophthalmic pharmaceutical products. In G. S. Banker & C. T. Rhodes (Eds.), Modern Phatmaceutics (pp. 626-717). New York: Marcel Dekker. Law, S. L., Huang, K. J., & Chiang, C. H. (2000). Acyclovir-containing liposomes for potential ocular delivery Corneal penetration and absorption. Journal of Controlled Release, 63, 135–140. Lawrence, M. J. (2003). Sorbitan esters (sorbitan fatty acid esters). In R. C. Rowe, P. L. Sheskey & P. L. Weller (Eds.), Handbook of Pharmaceutical Excipients (4th ed., pp. 591 - 595). London: Pharmaceutical Press. Lawrence, M. J., & Rees, G. D. (2000). Microemulsion-based media as novel drug delivery systems. Advanced Drug Delivery Reviews, 45, 89–121. Lawrence, R. S., Groom, M. H., Ackroyd, D. M., & Parish, W. E. (1986). The chorioallantoic membrane in irritation testing. Food and Chemical Toxicology, 24, 497-502. Le Bourlais, C., Acar, L., Zia, H., Sado, P., Needham, T., & Leverge , R. (1998). Ophthalmic drug delivery systems- recent advances. Progress in Retinal and Eye Research, 17, 33-58. Le Bourlias, C. A., Treupel-Acar, L., Rhodes, C. T., Sado, P. A., & Leverge, R. (1995). New ophthalmic drug delivery systems. Drug Development and Industrial Pharmacy, 21, 19– 59. Lee, B. L., Matoba, A. Y., Osato, M. S., & Robinson, N. M. (1992). The solubility of antibiotic and corticosteroid combinations. American Journal of Ophthalmology, 114, 212-215. Lee, H., Lee, J., Kim, M., Karya, Y., Miyazaki, K., & Kim, E. (2006). Insulin-like growth factor1 induces migration and expression of laminin-5 incultutured human corneal epithelial cells. Investigative Ophthalmology & Visual Science, 47, 873-882. Lee, V. H. (1990). Review: new directions in the optimization of ocular drug delivery. Journal of Ocular Pharmacology and Therapeutics, 6, 157-164. Lee, V. H., Carson, L. W., & Takemoto, K. A. (1986). Macromolecular drug absorption in the albino rabbit eye. International Journal of Pharmaceutics, 29, 43-51. Lee, V. H., Chien, D. S., & Sasaki, H. (1988). Ocular ketone reductase distribution and its role in metabolism of ocularly applied levobunolol in the pigmented rabbit. Journal of Pharmacology and Experimental Therapeutics, 246, 871-878. Lee, V. H., & Li, V. H. (1989). Prodrug for improved ocular drug drug delivery. Advanced Drug Delivery Reviews, 3, 1-38. 264 References Lee, V. H., Morimoto, K. W., & Stratfor, R. E. (1982). Esterases distribution in the rabbit cornea and its implications in ocular drug bioavailability. Biopharmaceutics and Drug Dispostion, 68, 673-684. Lee, Y. C., Simamora, P., Pinsuwan, S., & Yalkowsky, S. H. (2002). Review on the systemic delivery of insulin via the ocular route. International Journal of Pharmaceutics, 233, 118. Lehr, C. M., Yong, H., & Lee, V. H. (1994). Improved ocular penetration of gentamicin by mucoadhesive polymer polycarbophil in the pigmented rabbit. . Investigative Ophthalmology & Visual Science, 35, 2809-2814. Leighton, J., Nassauer, J., & Tachao, R. (1985). The chick embryo in toxicology: an alternative to the rabbit eye. Food and Chemical Toxicology, 23, 293-298. Leo, A., Hansch, C., & Elkins, D. (1971). Partition coefficients and their uses. Chemical Reviews, 71, 525-616. Lerk, C. F., Lagas, M., Boelstra, J. P., & Broersma, P. (1977). Contact angles of pharmaceutical powders. Journal of Pharmaceutical Sciences, 66, 1480-1481. Levi-Montalcini, R. (1987). The nerve growth factor 35 years later. Science, 237, 1154-1162. Li, D. Q., & Tseng, S. C. (1996). Differential regulation of cytokine and receptor transcript expression in human corneal and limbal fibroblasts by epidermal growth factor, transforming growth factor-alpha, platelet derived growth factor B, and interleukin-1 beta. Investigative Ophthalmology & Visual Science, 37, 2068-2080. Li, J. P., Tripathi, R. C., & Tripathi, B. J. ( 2008). Drug-induced ocular disorders. Drug Safety, 31, 127-141. Loftssona, T., & Jarvinen, T. (1999). Cyclodextrins in ophthalmic drug delivery. Advanced Drug Delivery Reviews, 36, 59-79. Lonnerholm, G. (1974). Carbonic anhydrase in the cornea. Acta Phyiologica Scandinavica, 90, 143-152. Lu, L., Reinach, P., & Kao, W. (2001). Corneal epithelial wound healing. Experimental Biology and Medicine, 226, 653–664. Ludwig, A. (2005). The use of mucoadhesive polymers in ocular drug delivery. Advanced Drug Delivery Reviews, 57, 1595– 1639. Luepke, N. P. (1985). Hen's egg chorioallantoic membrane test for irritation potential. Food and Chemical Toxicology, 23, 287-291. 265 References Macha, S., Hughes, P. M., & Mitra, A. K. (2003). Overview of ocular drug delivery. In A. K. Mitra (Ed.), Ophthalmic Drug Delivery Systems (2nd ed., pp. 1-11). New York: Marcel Dekker. Mackie, I. A. (1995). Current Ocular Therapy. In F. Fraunfelder & F. H. Roy (Eds.), Neuroparalytic keratitis (4th ed., pp. 506–508). Philadelphia: WB Saunders. Maestrelli, F., Capasso, G., Gonzalez-Rodriguez, M., Rabasco, A., & Mura, P. (2009). Effect of preparation technique on the properties and in vivo efficacy ethosome -loaded benzocaine. Journal of Liposome Research, 19, 253-260. Mainardes, R. M., Urban, M. C., Cinto, P. O., Khalil, N. M., Chaud, M. V., Evangelista, R. C., & Gremião, M. P. (2005). Colloidal carriers for ophthalmic drug delivery. Current Drug Targets, 6, 363-371. Maldonado, B., & Furcht, L. (1995a). Epidermal growth factor stimulates integrin mediated cell migration of cultured human corneal epithelial cells on fibronectin and arginine-glycineaspartic acid peptide. Investigative Ophthalmology & Visual Science, 36, 2120–2126. Maldonado, B., & Furcht, L. (1995b). Involvement of integrins with adhesion promoting, heparin-binding peptides of type IV collagen in cultured human corneal epithelial cells. Investigative Ophthalmology & Visual Science, 36, 364-372. Manconi, M., Valenti, D., Sinico, C., Lai, F., Loy, G., & Fadda, A. M. (2003). Niosomes as carriers for tretinoin II. Influence of vesicular incorporation on tretinoin photostability. International Journal of Pharmaceutics, 260, 261–272. Manosroi, A., Chutoprapat, R., Abe, M., & Manosroi, J. (2008a). Characteristics of niosomes prepared by supercritical carbon dioxide (scCO2) fluid. International Journal of Pharmaceutics, 352, 248–255 Manosroi, A., Jantrawuta, P., & Manosroi, J. (2008b). Anti-inflammatory activity of gel containing novel elastic niosomes entrapped with diclofenac diethylammonium. International Journal of Pharmaceutics, 360, 156-163. Manosroi, A., Wongtrakul, P., Manosroi, J., Sakai, H., Sugawara, F., Yuasa, M., & Abe, M. (2003). Characterization of vesicles prepared with various non-ionic surfactants mixed with cholesterol. Colloids and Surfaces B: Biointerfaces, 30, 129-138. Marsh, R. J., & Maurice, D. M. (1971). The influence o f non-ionic detergents and other surfactants on human corneal permeability. Experimental Eye Research, 11, 43-48. 266 References Martin, A., Bustamante, P., & Chun, A. H. C. (1993). Rheology. In A. Martin (Ed.), Physical Pharmacy (4th ed., pp. 453-476). Baltimore, Maryland: Lippincott Williams & Wilkins. Martini, L. G., Embleton, J. K., Malcolmson, R. J., & Wilson, C. G. (1997). The use of small volume ocular sprays to improve the bioavailability of topically applied ophthalmic drugs. European Journal of Pharmaceutics and Biopharmaceutics, 44, 121-126. Maurer, N., Zhigaltsev, I., & Cullis, P. R. (2007). Encapsulation of nucleic acid–based therapeutics. In G. Gregoriadis (Ed.), Liposome Technology: entrapment of drugs and other materials into liposomes (3rd ed., Vol. 2, pp. 149-168). New York: Informa Healthcare. Mayer, L. D., Hope, M. J., Cullis, P. R., & Janoff, A. S. (1985). Solute distributions and trapping efficiencies observed in freeze-thawed multilamellar vesicles. Biochimica et Biophysica Acta, 817, 193-196. Mayer, L. D., Tai, L. C., Ko, D. S., Masin, D., Ginsberg, R. S., Cullis, P. R., & Bally, M. B. (1989). Influence of vesicle size, lipid composition, and drug-to-lipid ratio on the biological activity of liposomal doxorubicin in mice. Cancer Research, 49, 5922–5930. Mayers, J., Gomes, J., & Siepser, S. (1997). Effect of transforoming growth factor beta 1 on stromal haze following excimer laser photorefractive keratectomy in rabbits. Journal of Refractive Surgery, 13, 356-361. McAvoy, J. W., & Chamberlain, C. G. (1990). Growth factors in the eye. Progress in Growth Factor Research, 2, 29-43. McClure, D. A. (1975). In Y. Higuchi & V. Stella (Eds.), Prodrugs as Novel Drug Delivery Systems (pp. 224-235). Washington: American Chemical Society. McLaughlin, P. J. (1994). Opioid antagonist modulation of rat heart development. Life Sciences, 54, 1423-1431. McLaughlin, P. J., Sassani, J. W., Klocek, M. S., & Zagon, I. S. (2010). Diabetic keratopathy and treatment by modulation of the opioid growth factor (OGF)–OGF receptor (OGFr) axis with naltrexone: A review. Brain Research Bulletin, 81, 236–247. Mehanna, M. M., Elmaradny, H. A., & Samaha, M. W. (2009). Ciprofloxacin liposomes as vesicular reservoirs for ocular delivery: formulation, optimization, and in vitro characterization. Drug Development and Industrial Pharmacy, 35, 583-593. 267 References Mehanna, M. M., Elmaradny, H. A., & Samaha, M. W. (2010). Mucoadhesive liposomes as ocular delivery system: physical, microbiological and in vivo assessment. Drug Development and Industrial Pharmacy, 36, 108-118. Meisnera, D., & Mezeib, M. (1995). Liposome ocular delivery systems. Advanced Drug Delivery Reviews, 16, 75-93. Meriney, S. D., For, M. J., Oliva, D., & Pilar, G. (1991). Endogenous opioids modulate neural survival in the developing avian ciliary ganglion. The Journal of Neuroscience, 11, 37053717. Micera, A., Lambiase, A., Aloe, L., Bonini, S., Levi, Schaffer, F., & Bonini, S. (2004). Nerve growth factor involvement in the visual system: implications in allergic and neurodegenerative diseases. Cytokine & Growth Factor Reviews, 15, 411–417. Mikkelson, T. J., Chrai, S. S., & Robinson, J. R. (1973a). Altered bioavailability of drugs in the eye due to drug-protein interaction. Journal of Pharmaceutical Sciences, 62, 1648-1653. Mikkelson, T. J., Chrai, S. S., & Robinson, J. R. (1973b). Competitive inhibition of drug-protein interaction in the eye fluids and tissues. Journal of Pharmaceutical Sciences, 62, 19421945. Miller, S. C., & Donovan, M. D. (1982). Effect of poloxamer 407 gel on the miotic activity of pilocarpine nitrate in rabbits. International Journal of Pharmaceutics, 12, 147-152. Misra, A. L. (1981). Narcotic antagonists: naltrexoe pharmacochemistry and sustained-release preparations. In R. E. Willette & G. Barnett (Eds.), Natlional Institute on Drug Abuse Research Monograph Series 28 (pp. 1-276). Maryland: National Institute On Drug Abuse. Miyazaki, S., Oshiba, M., & Nadai, T. (1981). Precaution on use of hydrochloride salts in pharmaceutical formulation. Journal of Pharmaceutical Sciences, 70, 594-596. Miyazaki, S., Suzuki, S., Kawasaki, N., Endo, K., Takahashi, A., & Attwood, D. (2001). In situ gelling xyloglucan formulations for sustained release ocular delivery of pilocarpine hydrochloride. International Journal of Pharmaceutics, 229, 29-36. Moazeni, E., Gilani, K., Sotoudegan, F., Pardakhty, A., Najafabadi, A. r., Ghalandari, R., Fazeli, M. R., & Jamalifar, H. (2010). Formulation and in vitro evaluation of ciprofloxacin containing niosomes for pulmonary delivery. Journal of Microencapsulation 27, 618-627. Mockel, J. E., & Lippold, B. C. (1993). Zero-Order Release from Hydrocolloid Matrices. Pharmaceutical Research, 90, 1066-1070. 268 References Mokhtar, M., Sammour, O. A., Hammad, M. A., & Megrab, N. A. (2008). Effect of some formulation parameters on flurbiprofen encapsulation and release rates of niosomes prepared from proniosomes. International Journal of Pharmaceutics, 361(1-2), 104-111. Mosteller, M. W., Gebhardt, B. M., Hamilton, A. M., & Kaufman, H. E. (1985). Penetration of topical cyclosporine into the rabbit cornea, aqueous humor, and serum. Archives of Ophthalmology, 103, 101-102. Muir, C. K. (1985). Opacity of bovine cornea in vitro induced by surfactants and industrial chemicals compared with ocular irritancy in vivo Toxicology Letters, 24, 157-162. Mukherjee, B., Patra, B., Layek, B., & Mukherjee, A. (2007). Sustained release of acyclovir from nano-liposomes and nano-niosomes: an in vitro study. International Journal of Nanomedicine, 2, 213-225. Muller, L. J., Vrensen, G. F., Pels, L., Cordozo, B. N., & Willekens, B. (1997). Architecture of human corneal nerves. Investigative Ophthalmology & Visual Science, 38, 985-994. Muller, R. H., Mader, K., & Gohla, S. (2000). Solid lipid nanoparticles (SLN) for controlled drug delivery-review of the state of the art. European Journal of Pharmaceutics and Biopharmaceutics, 50, 161-177. Mura, S., Pirot, F., Manconi, M., Falson, F., & Fadda, A. M. (2007). Liposomes and niosomes as potential carriers for dermal delivery of minoxidil. Journal of Drug Targeting, 15(2), 101108. Muzzalupo, R., Tavano, L., Cassano, R., Trombino, S., Ferrarelli, T., & Picci, N. (2011). A new approach for the evaluation of niosomes as effective transdermal drug delivery systems. European Journal of Pharmaceutics and Biopharmaceutics, In Press. Naeser, P. (1997). Insulin receptors in human ocular tissues. Immunohistochemical demonstration in normal and diabetic eyes. Upsala Journal of Medical Sciences, 102, 3540. Nagano, T., Nakamura, M., & Nakata, K. (2003). Effects of substance P and IGF-1 in cornea epithelial barrier function and wound healing in a rat model of neurotrophic keratopathy. Investigative Ophthalmology & Visual Science, 44, 3810–3815. Nakamura, M., Ofuji, K., Chikama, T., & Nishida, T. (1997). The NK1 receptor and its participation in the synergistic enhancement of corneal epithelial migration by substance P and insulin-like growth factor- 1. British Journal of Ophthalmology, 120, 547–552. 269 References Nakano, K., Tozuka, Y., Yamamoto, H., Kawashima, Y., & Takeuchi, H. (2008). A novel method for measuring rigidity of submicron-size liposomes with atomic force microscopy. International Journal of Pharmaceutics, 355, 203-209. Namdeo, A., & Jain, N. K. (1999). Niosomal delivery of 5- Fluorouracil. Journal of Microencapsulation, 16, 731- 740. Nanjawade, B. K., Manvi, F. V., & Manjappa, A. S. (2007). In situ-forming hydrogels for sustained ophthalmic drug delivery. Journal of Controlled Release, 122, 119-134. Naresh, R. A. R., & Udupa, N. (1996). Niosome encapsulated bleomycin. STP Pharma Sciences 6, 61–71. Nasr, M., Mansour, S., Mortada, N. D., & Elshamy, A. A. (2008). Vesicular aceclofenac systems: A comparative study between liposomes and niosomes. Journal of Microencapsulation, 25, 499–512. Naveh, N., Muchtar, S., & Benita, S. (1994). Pilocarpine incorporated into a submicron emulsion vehicle causes an unexpectedly prolonged ocular hypotensive effect in rabbits. Journal of Ocular Pharmacology and Therapeutics, 10, 509–520. Nefise, O. S. (2007). Niosomes as nanocarrier systems. In M. R. Mozafari (Ed.), Nanomaterials and Nanosystems for Biomedical Applications (pp. 67–81): Springer. Negro, A., Martini, I., Bigon, E., Cazzola, F., Minozzi, C., Skaper, S. D., & Callegaro, L. (1992). Synthesis of the biologically active β-subunit of human nerve growth factor in Escherichia coli. Gene, 110, 251–256. Nelson, W. L., Fraunfelder, F. T., Sills, J. M., Arrowsmith, J. B., & Kuritsky, J. N. (1986). Adverse respiratory and cardiovascular events attributed to timolol ophthalmic solution. American Journal of Ophthalmology, 102, 606-611. Nesseem, D. I. (2011). Ophthalmic delivery of sparfloxacin from in situ gel formulation for treatment of experimentally induced bacterial keratitis. Drug Testing and Analysis, 3, 106-115. Niemac, S. M., Ramachandran, C., & Weiner, N. (1995). Influence of non-ionic liposomal composition on topical delivery of peptide drugs into pilosebaceous units: an in-vivo study using hamster ear model Pharmaceutical Research, 12, 1184–1188. Ning, M., Guo, Y., Pan, H., Zong, S., & Gu, Z. (2006). Preparation and characterization of EPliposomes and Span 40-niosomes. Die Pharmazie, 208-212. 270 References Nishida, T. (2005). Cornea. In J. H. Krachmer, M. J. Mannis & E. J. Holland (Eds.), Cornea: Fundamentals, Diagnosis and Management (Vol. 1, pp. 3-21). Philadelphia: Elsevier Nishida, T., Nakamura, M., Mishima, H., & Otori, T. (1992). Interleukin 6 promotes epithelial migration by a fibronectin-dependent mechanism. Journal of Cellular Physiology, 153, 1– 5. Nishida, T., Nakamura, M., Ofuji, K., Reid, T., Mannis, M. J., & J, M. C. (1996). Synergistic effects of substance P with Insulin-like growth factor-1 on epithelial migration of the cornea Journal of Cellular Physiology 169, 159–166. Nishio, T., Sunohara, N., Furukawa, S., Akiguchi, I., & Kudo, Y. (1998). Repeated injections of nicergoline increase the nerve growth factor level in the aged rat brain. The Japanese Journal of Pharmcology, 76, 321–323. Nishizawa, M., Ozawa, F., Higashizaki, T., Hirai, K., & Hishinuma, F. (1993). Biologically active humand and mouse nerve growth factors secreted by the yeast Saccharomyces cerevisiae. Applied Microbiology and Biotechnology, 38, 624–630. Norley, S. G., Huang, L., & Rouse, B. T. (1986). Targeting of drug loaded immunoliposomes to herpes simplex virus infected corneal cells: an effective means of inhibiting virus replication in vitro. The Journal of Immunology, 136, 681–685. O’Neil, M. J., Heckelman, P. E., Koch, C. B., & Roman, K. J. (2006). Naltrexone Hydrochloride. In M. J. O’Neil, P. E. Heckelman, C. B. Koch & K. J. Roman (Eds.), The Merck Index: An Encyclopedia of Chemicals, Drugs, and Biologicals (14th ed., p. 6363). New Jersey: Merck & Co., Inc. Ohashi, Y., Matsuda, M., & Hosotani, H. (1988). Aldose reductase inhibitor(CT-112) eye drops for diabetic corneal epitheliopathy. American Journal of Ophthalmology, 105, 233-238. Ohji, M., Mandarino, L., Sundar, R. N., & Thoft, R. (1993). Corneal epithelial cell attachment with endogenous laminin and fibronectin Investigative Ophthalmology & Visual Science, 34, 2487–2492. Ollivon, M., Eidelman, O., Blumenthal, R., & Walter, A. (1988). Micelle-vesicle transition of egg phosphatidylcholine and octylglucoside. Biochemistry, 27, 1695 - 1703. Palmgren, J. J., Toropainen, E., Auriola, S., & Urtti, A. (2002). Liquid chromatographicelectrospray ionization mass spectrometric analysis of neutral and charged polyethylene glycols. Journal of Chromatography A, 976, 165-170. 271 References Pardakhty, A., Varshosaz, J., & Rouholamini, A. (2007). In vitro study of polyoxyethylene alkyl ether niosomes for delivery of insulin. International Journal of Pharmaceutics, 328, 130141. Parthasarathi, G., Udupa, N., Umadevi, P., & Pillai, G. K. (1994). Niosome encapsulated of vincristine sulfate-improved anticancer activity with reduced toxicity in mice. Journal of Drug Targeting, 2, 173–182. Passo, M. S., Palmer, E. A., & Van Buskirk, E. M. (1984). Plasma timolol in glaucoma patients. Ophthalmology, 91, 1361-1363. Pawar, P. K., & Majumdar, D. K. (2006). Effect of formulation factors on in vitro permeation of moxifloxacin from aqueous drops through excised goat, sheep and buffalo corneas. AAPS PharmSciTech, 7(1), 1-6. Peppas, N. A. (1985). Analysis of Fickian and non-Fickian drug release from polymers. Pharmaceutica Acta Helvetiae, 60, 110-111. Perrie, Y., Mohammed, A. U. R., Vangala, A., & McNeil, S. (2007). Environmental scanning electron microscopy offers real-time morphological analysis of liposomes and niosomes. Journal of Liposome Research, 17, 27-37. Pfister, L. A., Alther, E., Papaloizos, M., Merkle, H. P., & Gander, B. (2008). Controlled nerve growth factor release from multi-plyalginate/chitosan-based nerve conduits. European Journal of Pharmaceutics and Biopharmaceutics, 69, 563-572. Pfister, R. R., & Burstein, N. (1976). The effects of ophthalmic drugs, vehicles, preservatives on corneal epithelium: A scanning electron microscope study. Investigative Ophthalmology & Visual Science, 15, 246-259. Pierre, E. J., Barrow, R. E., & Hawkins, H. K. (1998). Effects of insulin on wound healing. The Journal of Trauma, 44, 342–345. Pillion, D. J., Bartlett, J. D., Meezan, E., Yang, M., Crain, R. J., & Grizzle, W. E. (1991). Systemic absorption of insulin delivered topically to the rat eye. Investigative Ophthalmology & Visual Science, 32, 3021-3027. Planas, M. E., Gonzalez, P., Rodriguez, L., Sanchez, S., & Cevc, G. (1992). Noninvasive percutaneous induction of topical analgesia by a new type of drug carrier, and prolongation of local pain on sensitivity by anesthetic liposomes. Anesthesia & Analgesia, 75, 615-621. 272 References Polansky, J. R., & Alvarado, J. A. (1985). Isolation and evaluation of target cells in glaucoma research: hormone receptors and drug responses. Current Eye Research, 4, 267-279. Pontiroli, A., Calderara, A., & Pozza, G. (1989). Intranasal drug delivery. Potential advantages and limitations from a clinical pharmacokinetic perspective. Clinical Pharmacokinetics, 17, 299-307. Prinsen, M. K., & Koeter, H. B. (1993). Justification of the enucleated eye test with eyes of slaughterhouse animals as an alternative to the Draize eye irritation test with rabiits. Food and Chemical Toxicology, 31, 69-76. Quintana, A., Raczka, E., Piehler, L., Lee, I., Myc, A., Majoros, I., Patri, A. K., Thomas, T., Mulé, J., & Baker, J. R. (2002). Design and function of a dendrimer-based therapeutic nanodevice targeted to tumor cells through the folate receptor. Pharmaceutical Research, 19, 1310–1316. Rathore, K. S., & Nema, R. K. (2009). Review on ocular inserts. International Journal of PharmTech Research, 1, 164-169. Rathore, M. S., & Majumdar, D. K. (2006). Effect of formulation factors on in vitro transcorneal permeation of gatifloxacin from aqueous drops. AAPS PharmSciTech, 12 (3), 12-17. Rattenholl, A., Lilie, H., Grossmann, A., Stern, A., Schwarz, E., & Rudolph, R. (2001). The prosequence facilitates folding of human nerve growth factor from Escherichia coli inclusion bodies. European Journal of Biochemistry 268, 3296–3303. Reddy, D. N., & Udupa, N. (1993). Formulation and evaluation of oral and transdermal preparations of flurbiprofen and piroxicam incorporated with different carriers. Drug Development and Industrial Pharmacy, 19, 843–852. Renault, P. F. (1981). Treatment of heroin-dependent persons with antagonists: current status. In R. E. Willette & G. Barnett (Eds.), Natlional Institute on Drug Abuse Research Monograph Series 28 (pp. 21-32). Maryland, USA: National Institute on Drug Abuse. Reynolds, D. W., Facchine, K. L., Mullaney, J. F., Alsante, K. M., Hatajik, T. D., & Motto, M. G. (2002). Available guidance and best practices for conducting forced degradation studies. Pharmaceutical Technology 26, 48-56. Ríos, J. D., Ghinelli, E., Gu, J., Hodges, R. R., & Dartt, D. A. (2007). Role of neurotrophins and neurotrophin receptors in rat conjunctival goblet cell secretion and proliferation. Investigative Ophthalmology & Visual Science, 48, 1543–1551. 273 References Ritger, P. L., & Peppas, N. A. (1987). A Simple equation for the description of solute release III. coupling of diffusion and relaxation. Journal of Controlled Release, 5, 23-36. Robinson, J. R., & Mlynek, G. M. (1995). Bioadhesive and phase-change polymers for ocular drug delivery. Advanced Drug Delivery Reviews, 16, 45-50. Rocha, E. M., Cunha, D. A., Carneiro, E. M., Boschero, A. C., Saad, M. J., & Velloso, L. A. (2002). Insulin, insulin receptor and insulin-like growth factor-I receptor on the human ocular surface. Advances in Experimental Medicine & Biology, 506, 607-610. Roda, A., Russo, C., Pasini, P., Piazza, F., Feroci, G., Kricka, L. J., & Baraldini, M. (1998). Antioxidant properties of bile salt micelles evaluated with different chemiluminescent assays: a possible physiological role. Journal of Bioluminescence and Chemiluminescence, 13, 327-337. Roozen, H. G., Waart, R., & Brink, W. (2007). Efficacy and tolerability of naltrexone in the treatment of alcohol dependence: oral versus injectable delivery. European Addiction Research 13, 201–206. Rosenberg, M. E., Tervo, T. M., Immonen, I. J., Muller, L. J., Gronhagen-Riska, C., & Vesaluoma, M. H. (2000). Corneal structure and sensitivity in type 1 diabetes mellitus. Investigative Ophthalmology & Visual Science, 41, 2915–2921. Rosenthal, S. P. (1968). Acceleration of primary wound healing by insulin. Archives of Surgery, 96, 53–55. Rozier, A., Mazuel, C., Grove, J., & Plazonnet, B. (1989). Gelrite®: a novel, ion-activated, insitu gelling polymer for ophthalmic vehicles. Effect on bioavailability of timolol. International Journal of Pharmaceutics, 57, 163-168. Ruben, S. T. (1994). Corneal sensation in insulin dependent and non-insulin dependent diabetics with proliferative retinopathy. Acta Ophthalmologica, 72(5), 576-580. Rubinstein, L. V., Shoemaker, R. H., Paull, K. D., Simon, R. M., Tosini, S., Skehan, P., Scudiero, D. A., Monks, A., & Boyd, M. R. (1990). Comparison of in vitro anticancer-drugscreening data generated with a tetrazolium assay versus a protein assay against a diverse panel of human tumor cell lines. Journal of the National Cancer Institute, 82, 1113-1118. Saettone, M. F., Chetoni, P., Cerbai, R., Mazzanti, G., & Braghiroli, L. (1996a). Evaluation of ocular permeation enhancers: in vitro effects on corneal transport of four fl-blockers, and in vitro/in vivo toxic activity. International Journal of Pharmaceutics, 142, 103-113. 274 References Saettone, M. F., Perini, G., Carafa, M., Santucci, E., & Alhaique, F. (1996b). Non-ionic surfactant vesicles as ophthalmic carriers for cyclopentolate. A preliminary evaluation. STP Pharma Sciences, 6, 94–98. Sahin, N. O. (2006). Niosomes as nanocarrier systems. In M. R. Mozafari (Ed.), Nanomaterials and Nanosystems for Biomedical Applications (pp. 67–81). New York: Springer. Sahoo, S. K., Dilnawaz, F., & Krishnakumar, S. (2008). Nanotechnology in ocular drug delivery. Drug discovery today, 13, 144-151. Saika, S., Okada, Y., & Miyamoto, T. (2004). Role of p38 MAP kinase in regulation of cell migration and proliferation in healing cornealepithelium. Investigative Ophthalmology & Visual Science, 45, 100–109. Sakiyama-Elbert, S., & Hubbell, J. A. (2000). Controlled release of nerve growth factor from a heparin-based cell ingrowth matrix. Journal of Controlled Release, 69, 149-158. Saneto, R. P., Awashti, Y. C., & Srivastava, S. K. (1982). Mercapturic acid pathway enzymes in bovine ocular lens, cornea, retina, retinal pigmented epithelium. Experimental Eye Research, 34, 107-111. Scaife, M. C. (1985). An in vitro cytotoxicity test to predict the ocular irritation potential of detergents and detergent products. Food and Chemical Toxicology, 23, 252-258. Schaeffer, H. E., & Krohn, D. L. (1982). Liposomes in topical drug delivery. Investigative Ophthalmology & Visual Science, 22, 220 - 227. Schoenwald, R. D., & Huang, H. S. (1983). Corneal penetration behavior of beta blocking agents I: physichochemical factors. Journal of Pharmaceutical Sciences, 72, 1266-1272. Schoenwald, R. D., & Stewart, P. (1980). Effect of particle size on ophthalmic bioavailability of dexamethasone suspensions in rabbits. Journal of Pharmaceutical Sciences, 69, 391-394. Schoenwald, R. D., & Ward, R. L. (1978). Relationship between steroid permeability across excised rabbit cornea and octanol-Water partition coefficients. Journal of Pharmaceutical Sciences, 67, 786-788. Schreier, H., & Bouwstra, J. (1994). Liposomes and niosomes as topical drug carriers-dermal and transdermal drug-delivery. Journal of Controlled Release, 30, 1-15. Schultz, G. S., Davis, J. B., & Eiferman, R. A. (1990). Growth factors and corneal epithelium. Cornea, 9, 2-9. 275 References Schultz, R. O., Van Horn, D. L., Peters, M. A., Klewin, K. M., & Schutten, W. H. (1981 ). Diabetic keratopathy. Transactions of the American Ophthalmological Society, 79, 180199. Schulze, S. D., Sekundo, W., & Kroll, P. (2006). Autologous serum for the treatment of corneal epithelial abrasions in diabetic patients undergoing vitrectomy. American Journal of Ophthalmology, 142, 207-214. Schwartz, D. E. (1974). Corneal sensitivity in diabetics. Archives of Ophthalmology, 91, 174-178. Schwartz, S. D., Harrison, S. A., Engstrom, R. E., Bawden, R. E., Lee, D. A., & Mondino, B. J. (1990). Collagen shield delivery of Amphotericin B. American Journal of Ophthalmology, 109, 701-704. Shah, V. P., Midha, K. K., Dighe, S., Mcgilveray, I. L., Skelly, J. P., & Jacobi, A. (1992). Analytical methods validation: Bioavailability, bioequivalence, and pharmacokinetic studies Journal of Pharmaceutical Sciences, 81, 309–312. Shahiwala, A., & Misra, A. (2002). Studies in topical application of niosomally entrapped nimesulide. Journal of Pharmacy & Pharmaceutical Sciences, 5, 220-225. Sharma, G., He, J., & Bazan, H. (2003). p38 and ERK1/2 coordinate cellular migration and proliferation in epithelial wound healing. The Journal of Biological Chemistry, 278, :21989–21997. Shell, J. W. (1982). Pharmacokinetics of topically applied ophthalmic drugs. Survey of Ophthalmology, 26, 207-218. Shooter, E. M. (2001). Early days of the nerve growth factor proteins. Annual Review of Neuroscience, 24, 601–629. Short, B. G. (2008). Safety evaluation of ocular drug delivery formulations: techniques and practical considerations. Toxicologic Pathology, 36, 49-62. Silverstein, R. M., Webster, F. X., & Kiemle, D. J. (2005). Infrared spectrometry. In R. M. Silverstein, F. X. Webster & D. J. Kiemle (Eds.), Spectrometric Identification of Organic Compounds (7th ed., pp. 72-126). New York: John Wiley & Sons, Inc. Sinha, V. R., Monika, A., Kumar, M., Singh, S., & Bhinge, J. R. (2007). Stress studies on acyclovir. The Journal of Chromatographic Science, 45, 319-324. Skehan, P., Storeng, R., Scudiero, D., Monks, A., McMahon, J., Vistica, D., Warren, J. T., Bokesch, H., Kenney, S., & Boyd, M. R. (1990). New colorimetric cytotoxicity assay for anticancer-drug screening. Journal of the National Cancer Institute, 82, 1107-1112. 276 References Smolen, V. F., Clevenger, J. M., Williams, E. J., & Bergdolt, H. W. (1973). Biophasic availability of ophthalmic carbachol. I: Mechanism of cationic polymers- and surfactantpromoted miotic activity. Journal of Pharmaceutical Sciences, 62, 958-961. Solanki, A. B., Parikh, J. R., Parikh, R. H., & Patel, M. R. (2010). Evaluation of different compositions of niosomes to optimize aceclofenac transdermal delivery. Asian Journal of Pharmaceutical Sciences, 5, 87-95. Spicer, P. T. (2005). Progress in liquid crystalline dispersions: Cubosomes. Current Opinion in Colloid & Interface Science, 10, 274 – 279. Stewart, P. J., & Tucker, I. G. (1984). Prediction of drug stability-part 1: mechanisms of drug degradation and basic rate laws. Australian Journal of Hospital Pharmacy 14, 165-170. Stratford, R. E., & Lee, V. H. L. (1985). Ocuar aminopeptidase activity and distribution in the albino rabbit. Current Eye Research, 4, 995-999. Suwakul, W., Ongipattanakul, B., & Vardhanabhuti, N. (2006). Preparation and Characterization of Propylthiouracil Niosomes. Journal of Liposome Research 16, 391–401. Suzuki, K., Saito, J., Yanai, R., Yamada, N., Chikama, T., Seki, K., & Nishida, T. (2003). Cellmatrix and cell-cell interactions during corneal epithelial wound healing. Progress in Retinal and Eye Research 22, 113-133. Sweetman, S. C. (2007). Naltrexone Martindale: the Complete Drug Reference (35th ed., pp. 1310-1311). London: Pharmaceutical Press. Szoka, F., & Papahadjopoulos, D. (1978). Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proceedings of the National Academy of Sciences, 75, 4194-4198. Talsma, H., Steenbergen, M. J. V., Borchert, J. C. H., & Crommelin, D. J. A. (1994). A novel technique for the one-step preparation of liposomes and non-ionic surfactant vesicles without the use of organic solvents. Liposome formation in a continuos gas stream: the bubble method. Journal of Pharmaceutical Sciences, 83, 276–280. Tanaka, M. (1990). Properties of aqueous vesicle dispersion formed with poly(oxyethylene) hydrogenated castor oil. Journal of the American Oil Chemists' Society 67, 55–60. Tanford, C. (1973). Micelles. In C. Tanford (Ed.), The Hydrophobic Effect: Formation of Micelles and Biological Membranes (1st ed., pp. 36-44). New York: John Wiley & Sons, Ltd. 277 References Tang-Liu, D. D., Liu, S. S., & Weinkam, R. J. (1984). Ocular and systemic bioavailability of ophthalmic flurbiprofen. Journal of Pharmacokinetics and Biopharmaceutics, 12, 611626. Taylor, K. M. G., & Craig, D. Q. M. (2003). Physical methods of study: differential scanning calorimetry. In V. Torchilin & V. Weissig (Eds.), Liposomes : a practical approach (2nd ed., pp. 79-101). New York: Oxford University Press. Thoft, R. A., & Friend, J. (1983). The X, Y, Z hypothesis of corneal epithelial maintenance. Investigative Ophthalmology & Visual Science, 24, 1442-1443. Thomopoulos, P., & Pessac, B. (1979). Insulin receptors in cultured mouse retinal cells Diabetologia, 16, 275–279. Tinsley, P. W., Dass, C., Trimble, J. W., & Desiderio, D. M. (1990). Identification of methionine enkephalin in the bovine by fast atom bombardment-mass spectrometry. Experimental Eye Research, 51, 671-674. Tinsley, P. W., Fridland, G. H., Killmar, J. T., & Desiderio, D. M. (1989). Purification, characterization, and localization of neuropeptides in the cornea. Peptides, 9, 1373-1379. Tomalia, D. A., Baker, H., Dewald, J., Hall, M., Martin, S., Roeck , J., Ryder, J., & Smith, P. (1985). A new class of polymers: Starburst-dendritic macromolecules. Polymer Journal, 17, 117–132. Touhami, A., Grueterich, M., & Tseng, S. C. (2002). The role of NGF signaling in human limbal epithelium expanded by amniotic membrane culture. Investigative Ophthalmology & Visual Science, 43, 987–994. Touitou, E., Alkabes, M., & Dayan, N. (1997). Ethosomes: novel lipid vesicular system for enhanced delivery. Pharmaceutical Research, S14, 305–306. Touitou, E., Dayan, N., Bergelson, L., Godin, B., & Eliaz, M. (2000). Ethosomes-novel vesicular carriers for enhanced delivery: characterization and skin penetration properties. Journal of Controlled Release, 65, 403–418. Tous, S. S., & El Nasser, K. A. (1992). Acetazolamide topical formulation and ocular effect. STP Pharma Sciences 2, 125-131. Treblay, C., Barza, M., Skoda, F., Lahav, M., & Baum, J. (1993). Reduced toxicity of liposomassociated amphotericin B injected intravitreally in rabbits. Investigative Ophthalmology & Visual Science, 34, 2737-2742. 278 References Tripathi, B. J., Kwait, J. P. S., & Tripathi, R. C. (1990). Corneal growth factors: a new generation of ophthalmic pharmaceuticals. Cornea, 9, 2-9. Tsubota, K., & Dogru, M. (2006). Increasing tear quantity with artifical tears and reducing evaporation. In P. A. Asbell & M. A. Lemp (Eds.), Dry Eye Disease (pp. 85-94). New York: Thieme. Uchegbu, I. F., Bouwstra, J. A., & Florence, A. T. (1992). Large disk-shaped structures (Discomes) in non-ionic surfactant vesicles to micelle transitions. The Journal of Physical Chemistry, 96, 10548-10553. Uchegbu, I. F., Double, J. A., Turton, J. A., & Florence, A. T. (1995). Distribution, metabolism and tumoricidal activity of doxorubicin administered in sorbitan monostearate (Span 60) niosomes in the mouse. Pharmaceutical Research, 12, 1019-1024. Uchegbu, I. F., & Duncan, R. (1997). Niosomes containing N-(2 hydroxypropyl)methacrylamide copolymer-doxorubicin (PK1): effect of method of preparation and choice of surfactant on niosome characteristics and a preliminary study of body distribution. International Journal of Pharmaceutics, 155, 7–17. Uchegbu, I. F., & Florence, A. T. (1995). Non-ionic surfactant vesicles (niosomes): physical and pharmaceutical chemistry. Advances in Colloid and Interface Science, 58, 1-55. Uchegbu, I. F., Gianasi, E. C., F, Florence, A. T., & Duncan, R. (1996a). Niosome encapsulation of a doxorubicin polymer conjugate. European Journal of Pharmaceutical Sciences, Suppl. 4, S28. Uchegbu, I. F., McCarthy, D., Schatzlein, A., & Florence, A. T. (1996b). Phase-transitions in aqueous dispersions of the hexadecyl diglycerol ether (C16G2) non-ionic surfactant, cholesterol and cholesteryl poly-24-oxyethylene ether-vesicles, tubules, discomes and micelles. STP Pharma Sciences, 6, 33–43. Uchegbu, I. F., Schatzlein, A. G., Vanlerrberghe, G., Morgatini, N., & Florence, A. T. (1997). Polyhedral non-ionic surfactant vesicles. Journal of Pharmacy and Pharmacology 49, 606-610. Uchegbu, I. F., & Vyas, S. P. (1998). Non-ionic surfactant based vesicles (niosomes) in drug delivery. International Journal of Pharmaceutics, 172, 33–70. Ullrich, A., Gray, A., Berman, C., & Dull, T. J. (1983). Human beta-nerve growth factor gene sequence highly homologous to that of mouse. Nature, 303, 821–825. 279 References Unterman, S. R., Rootman, D. S., Hill, J. M., Parelman, J. J., Thompson, H. W., & Kaufman, H. E. (1988). Collagen shield drug delivery: therapeutic concentrations of tobramycin in the rabbit cornea and aqueous humor. Journal of Cataract & Refractive Surgery, 14, 500-504. Urtti, A. (2006a). Challenges and obstacles of ocular pharmacokinetics and drug delivery. Advanced Drug Delivery Review, 58 1131-1135. Urtti, A. (2006b). Ocular drug delivery. Advanced Drug Delivery Reviews, 58, 1129-1130. Urtti, A., & Salminen, L. (1993). Minimizing systemic absorption of topically administered ophthalmic drugs. Survey of Ophthalmology, 37, 435-456. Urtti, A., Salminen, L., & Miinaleinen, O. (1985). Systemic absorption of ocular pilocarpine is modified by polymer matrices. International Journal of Pharmaceutics, 23, 147-161. USP32/NF27. (2009). Naltrexone hydrochloride United State Pharmacopeia 32/ National Formulary 27 (Vol. Volume 3, pp. 3028-3029). Rockville, Maryland: The United States Pharmacopeial Convention, Inc. Van Hal, D., Vanrensen, A., Devringer, T., Junginger, H. E., & Bouwstra, J. A. (1996). Diffusion of estradiol from non-ionic surfactant vesicles through human stratum-corneum in vitro. STP Pharma Sciences, 6, 72–78. Van Ooteghem, M. M. M. (1993). Formulation of Ophthalmic Solutions and Suspensions. Problems and Advantages. In P. Edman (Ed.), Biopharmaceutics of ocular drug delivery (pp. 27-42). Boca Raton, FLa: CRC Press. Van Ort, S. R., & Gerber, R. M. (1976). Topical application of insulin in the treatment of decubitus ulcers: a pilot study. Nursing Research, 25, 9–12. Van Tomme, S. R., Storm, G., & Hennink, W. E. (2008). In situ gelling hydrogels for pharmaceutical and biomedical applications. International Journal of Pharmaceutics, 355, 1-18. Vandamme, T. F. (2002). Microemulsions as ocular drug delivery systems: recent developments and future challenges. Progress in Retinal and Eye Research, 21, 15–34. Vandamme, T. F., & Brobeck, L. (2005). Poly(amidoamine) dendrimers as ophthalmic vehicles for ocular delivery of pilocarpine nitrate and tropicamide. Journal of Controlled Release, 102, 23–38. Vanparys, P. H., Deknudt, G. H., Sysmans, M., Teuns, G., Coussement, W., & Van Cauteren, H. (1993). Evaluation of corneal opacity -permeability assay as an in vitro alternative to Draizeeye irritation test. Toxicology in Vitro, 7, 471-476. 280 References Varshosaz, J., Pardakhty, A., Hajhashemi, V., & Najafabadi, A. R. (2003). Development and physical characterization of sorbitan monoester niosomes for insulin oral delivery. Drug Delivery, 10, 251–262. Vasantha, R., Sehgal, P. K., & Panduranga, R. K. (1988). Collagen ophthalmic inserts for pilocarpine drug delivery system. International Journal of Pharmaceutics, 47, 95-102. Vora, B., Khopade, A. J., & Jain, N. K. (1998). Proniosome based transdermal delivery of levonorgestrel for effective contraception. Journal of Controlled Release, 54, 149–165. Vyas, S. P., Mysore, N., Jaittley, V., & Venkatesan, N. (1998). Discoidal niosome based controlled ocular delivery of Timolol maleate. Die Pharmazie, 53, 466–469. Wallach, D. F. H., & Philippot, J. R. (1993). New type of lipid vesicle: novasome. In G. Gregoriadis (Ed.), Liposome Technology (Vol. 2, pp. 141-156). Florida: CRC Press. Walter, A., Vinson, P. K., Kaplun, A., & Talmon, Y. (1991a). Intermediate structures in the cholate-phosphatidylcholine vesicle-micelle transition. Biophysical Journal, 60, 13151325. Walter, A., Vinson, P. K., Kaplun, A., & Talmon, Y. (1991b). Intermediate structures in the cholate-phosphatidylcholine vesiclemicelle transition. Biophysical Journal, 60, 13151325. Wang, S. B., Chen, A. C., Weng, L. J., Chen, M. Y., & Xie, K. L. (2004). Effect of drug-loading methods on drug load, encapsulation efficiency and release properties of alginate/poly-larginine/chitosan ternary complex microcapsules. Macromolecular Bioscience, 4, 27-30. Waterman, K. C., Adami, R. C., Alsante, K. M., Hong, J., Landis, M. S., Lombardo, F., & Roberts, C. J. (2002). Stabilization of pharmaceuticals to oxidative degradation. Pharmaceutical Development and Technology, 7, 1 – 32. Watsky, M., Jablonski, M., & Edelhauser, H. F. (1988). Comparison of conjunctival and cornral surface areas in rabbit and human. Current Eye Research, 7, 483-486. Waymouth, C. (1970). Osmolality of mammalian blood and of media for culture of mammalian cells. In Vitro, 6, 109-127. Weder, H. G., & Zumbuehl, O. (1984). The preparation of variably sized homogeneous liposomes for laboratory, clinical and industrial use by controlled detergent dialysis. In G. Gregoriadis (Ed.), Liposome Technology (Vol. 1, pp. 79-107). Florida: CRC Press. 281 References Wei, C., Anderson, J. A., & Leopold, I. (1978). Ocular absorption and metabolism of topically applied epinephrine and a dipivalyl ester of epinephrine. Investigative Ophthalmology & Visual Science, 17, 315-321. Weinreb, R. N., & Jani, R. (1992). A novel formulation of an ophthalmic B-adrenoreceptor antagonist. Journal of parenteral science and technology, 46, 51-53. Wells, J. I., & Aulton, M. (2007). Pharmaceutical preformulation. In M. E. Aulton (Ed.), Aulton's Pharmaceutics: the Design and Manufacture of Medicines (3nd ed., pp. 336-360). London: Churchill Livingstone. Weterings, P. J., & Vanerp, Y. H. (1987). Validation of the Becam assay: an eye irritancy screening test. In A. M. Goldberg (Ed.), Alternative Methods in Toxicology (Vol. 5, pp. 515-521). New York: Mary Ann Liebert, Inc. White, C. J., & Byrne, M. E. (2010). Molecularly imprinted therapeutic contact lenses. Expert Opinion on Drug Delivery, 7, 765-780. Wiberg, K., Andersson, M., Hagman, A., & Jacobsson, S. P. (2004). Peak purity determination with principal component analysis of high-performance liquid chromatography-diode array detection data. Journal of Chromatography A, 1029, 13-20. Wild, S., Roglic, G., Sicree, R., Green , A., & King, H. (2004). Global prevalence of diabetes:estimates for the year 2000 and projections for 2030. Diabetes Care, 27, 10471053. Willoughby, C. E., Batterbury, M., & Kaye, S. B. (2002). Collagen Corneal Shields. Survey of Ophthalmology, 47, 174-182. Wilson, S. E., Walker, J. W., Chwang, E. L., & G, H. Y. (1993). Hepatocyte growth factor, keratinocyte growth factor, their receptors, fibroblast growth factor receptor-2, and the cells of the cornea. . Investigative Ophthalmology & Visual Science, 34, 2544–2561. Winfield, A. J., Jessiman, D., Williams, A., & Esakowitz, L. (1990). A study of the causes of non-compliance by patients prescribed eyedrops. British Journal of Ophthalmology, 74, 477-480. Woo, H. M., Bentley, E., Campbell, S. F., CF, M., & Murphy, C. J. (2005). Nerve growth factor and corneal wound healing in dogs. Experimental Eye Research 80, 633-642. Worakul, N., & Robinson, J. R. (1997). Ocular pharmacokinetics/pharmacodynamics. European Journal of Pharmaceutics and Biopharmaceutics, 44, 71 - 83. 282 References Wu, Z., Tucker, I. G., Razzak, M., & Medlicott, N. (2005). Physichochemical characterisation of ricobendazole: I. solubility, lipophilicity and ionization charactersitics. Journal of Pharmaceutical Sciences, 94, 983-993. Xie, Y., Zhang, X., Cui, W., Lou, J., Nagai, T., & Hou, X. (2005). Transport of nerve growth factor encapsulated into liposomes across the blood-brain barrier: in vitro and in vivo studies. Journal of Controlled Release, 105, 106-119. Yamada, M., Ogata, M., Kawai, M., & Mashima, Y. (2000). Decreased substance P concentrations in tears from patients with corneal hypesthesia. American Journal of Ophthalmology, 129, 671-672. Yamada, N., Matsuda, R., Morishige, N., Yanai, R., Chikama, T. I., Nishida, T., Ishimitsu, T., & Kamiya, A. (2008). Open clinical study of eye-drops containing tetrapeptides derived from substance P and insulin-like growth factor-1 for treatment of persistent corneal epithelial defects associated with neurotrophic keratopathy. British Journal of Ophthalmology, 92, 896-900. Yeh, S., & Lach, J. L. (1961). Stability of morphine in aqueous solution III: kinetics of morphine degradation in aqueous solution. Journal of Pharmaceutical Sciences, 50, 30-34. Yokoi, N., Niiya, A., & Komuro, A. (1997). Effects of aldose reductase inhibitor, CT-112, on the corneal epithelial barrier of galactose-fed rats. Current Eye Research, 16, 595-599. Yoshioka, T., B, S., & Florence, A. T. (1994). Preparation and properties of vesicles (niosomes) of sorbitan monoesters (Span 20, Span 40, Span 60 and Span 80) and a sorbitsan triester (Span 85). International Journal of Pharmaceutics, 105, 1-6. You, L., Kruse, F. E., & Volcker, H. E. (2000). Neurotrophic factors in the human cornea. Investigative Ophthalmology & Visual Science, 41, 692-702. Yu, C. (2003). Polyoxyethylene alkyl ethers. In R. C. Rowe, P. L. Sheskey & P. L. Weller (Eds.), Handbook of Pharmaceutical Excipients (4th ed., pp. 469-473). London: Pharmaceutical Press. Zagon, I. S., Isayama, T., & McLaughlin, P. J. (1994a). Preproenkephalin mRNA expressionin the developing and adult rat brain. Molecular Brain Research 21, 85-98. Zagon, I. S., Jenkins, J. B., Sassani, J. W., Wylie, J. D., Ruth, T. B., Fry, J. L., Lang, C. M., & McLaughlin, P. J. (2002a). Naltrexone, an opioid antagonist, facilitates reepithelialization of the cornea in diabetic rat. Diabetes, 51, 3055-3062. 283 References Zagon, I. S., Klocek, M. A., Sassani, J. W., Mauger, D. T., & J, M. P. (2006a). Corneal safety of topically applied naltrexone. Journal of Ocular Pharmacology and Therapeutics, 22, 377387. Zagon, I. S., Klocek, M. S., Sassani, J. W., & McLaughlin, P. J. (2007a). Use of topical insulin to normalize corneal epithelial healing in diabetes mellitus. Archives of Ophthalmology, 125, 1082-1088. Zagon, I. S., Klocek, M. S., Sassani, J. W., & McLaughlin, P. J. (2009). Dry eye reversal and corneal sensation restoration with topical naltrexone in diabetes mellitus. Archives of Ophthalmology, 127, 1468-1473. Zagon, I. S., & McLaughlin, P. J. (1991a). Identification of opioid peptides regulating proliferation of neurons and glia in the developing nervous system. Brain Research, 542, 318-323. Zagon, I. S., & McLaughlin, P. J. (1991b). The role of endogenous opioids and opioid receptors in human and animal cancers. In A. J. Plotnikoff, A. J. Murgo, R. E. Faith & J. Wybran (Eds.), Stress and Immunity (pp. 343-356). Caldwell, New Jersy: CRC. Zagon, I. S., & McLaughlin, P. J. (1993a). Opioid growth factor receptor in the developing nervous system. In I. S. Zagon & P. J. McLaughlin (Eds.), Receptors in the Developing Nervous System (Vol. 1, pp. 39-62). London: Chapman and Hall. Zagon, I. S., & McLaughlin, P. J. (1993b). Production and characterisation of polyclonal and monoclonal antibodies to the zeta (ζ) opioid receptor. Brain Research, 630, 295-301. Zagon, I. S., Sassani, J. W., Kane, E. R., & McLaughlin, P. J. (1997). Homeostasis of ocular surface epithelium in the rat is regulated by opioid growth factor. Brain Research, 759(1), 92-102. Zagon, I. S., Sassani, J. W., Malefyt, K. J., & McLaughlin, P. J. (2006b). Regulation of corneal repair by particle-mediated gene transfer of opioid growth factor receptor complementary DNA. Archives of Ophthalmology, 124, 1620-1624. Zagon, I. S., Sassani, J. W., & McLaughlin, P. J. (1995). Opioid growth factor modulates corneal epithelial outgrowth in tissue culture. American Journal of Physiology 268 R942-R950. Zagon, I. S., Sassani, J. W., & McLaughlin, P. J. (1998a). Re-epithelialization of the rabbit cornea is regulated by opioid growth factor. Brain Research, 803(1-2), 61-68. Zagon, I. S., Sassani, J. W., & McLaughlin, P. J. (1998b). Re-epithelialization of the rat cornea is accelerated by blockade of opioid receptors. Brain Research, 798(1-2), 254-260. 284 References Zagon, I. S., Sassani, J. W., & McLaughlin, P. J. (2000). Reepithelialization of the human cornea is regulated by endogenous opioids. Investigative Ophthalmology & Visual Science, 41(1), 73-81. Zagon, I. S., Sassani, J. W., & McLaughlin, P. J. (2006c). Insulin treatment ameliorate impaired corneal reepithelization in diabetic rats. Diabetes, 55, 1141-1147. Zagon, I. S., Sassani, J. W., Myers, R. L., & McLaughlin, P. J. (2007b). Naltrexone accelerates healing without compromise of adhesion complexes in normal and diabetic corneal epithelium. Brain Research Bulletin, 72, 18–24. Zagon, I. S., Verderame, M. F., & McLaughlin, P. J. (2002b). The biology of the opioid growth factor receptor(OGFr). Brain Research Reviews, 38, 351-376. Zagon, I. S., Wu, Y., & McLaughlin, P. J. (1994b). Opioid growth factor (OGF) inhibits DNA synthesis in mouse tongue epithelium in a circadian-dependent manner. American Journal of Ophthalmology, 267 R645-R652. Zambito, Y., & D i Colo, G. (2010). Chitosan and its derivatives as intraocular penetration enhancers. Journal of Drug Delivery Science and Technology, 20, 45-52. Zhigaltsev, I. V., Kaplun, A. P., Kucheryanu, V. G., Kryzhanovsky, G. N., Kolomeichuk, S. N., Shvets, V. I., & Yurasov, V. V. (2001). Liposomes containing dopamine entrapped in response to transmembrane ammonium sulfate gradient as carrier systrem for dopamine delivery into the brain of Parkinsonian mice. Journal of Liposome Research, 11, 55-71. Zignani, M., Tabatabay, C., & Gurny, R. (1995). Topical semi-solid drug delivery: kinetics and tolerance of ophthalmic hydrogels. Advanced Drug Delivery Reviews, 16, 51-60. Zimmer, A. K., & Kreuter, J. (1991). Studies on the transport pathway of PBCA nanoparticles in ocular tissues. Journal of Microencapsulation 8, 497–504. Zimmerman, T. J., Kooner, K., S, Kandarakis, A. S., & Ziegler, L. P. (1984). Improving the therapeutic index of topically applied ocular drugs. Archives of Ophthalmology, 102, 551553. Zuidam, N. J., Lee, S. S., & Crommelin, D. J. A. (1993). Sterilization of liposomes by heat treatment. Pharmaceutical Research, 10, 1591-1596. Zuidam, N. J., Winden, E. V., Vrueh, R., & Crommelin, D. J. A. (2003). Stability, storage, and sterilization of liposomes. In V. Torchilin & V. Weissig (Eds.), Liposomes: a practical approach (2nd ed., pp. 149-165). New York: Oxford University Press. 285 Appendix -Publication Appendix-Publications 286