Enhanced Delivery of Actives Through Skin from Patches

Home , Cabotegravir, Cobicistat, Elvitegravir, Maraviroc

ENHANCED DELIVERY OF ACTIVES THROUGH SKIN FROM PATCHES

AND FORMULATIONS, AND DISTRIBUTION WITHIN AND ACROSS SKIN

ASHANA PURI

A Dissertation Submitted to the Graduate Faculty

of Mercer University College of Pharmacy

in Partial Fulfillment of the

Requirements for the Degree

DOCTOR OF PHILOSOPHY

Atlanta, GA

2018

ENHANCED DELIVERY OF ACTIVES THROUGH SKIN FROM PATCHES

AND FORMULATIONS, AND DISTRIBUTION WITHIN AND ACROSS SKIN

ASHANA PURI

Approved:

……………………………………………………………………………………………… Dr. Ajay K. Banga, Ph.D. Date Thesis Advisor

……………………………………………………………………………………………… Dr. Martin J. D'souza, Ph.D. Date Dissertation Committee Member

……………………………………………………………………………………………… Dr. Ravi Palaniappan, Ph.D. Date Dissertation Committee Member

...... Dr. J. Grady Strom, Ph.D. Date Dissertation Committee Member

……………………………………………………………………………………………… Dr. Vishal Sachdeva, Ph.D. Date Dissertation Committee Member

……………………………………………………………………………………………… Dr. Brian L. Crabtree, Pharm.D. Date Dean, College of Pharmacy

DEDICATION

In memory of my father, Ramesh Kumar Puri

To my mother, Chanchal Puri

With Love and Eternal Appreciation

iii

ACKNOWLEDGEMENT

It is always a tough job to express gratitude in words, for the “beautiful minds” that accompany in each and every step of the entire journey of carrying out the research work. The same perplexity haunts me in penning down sincere thanks to all those to whom this present work owes its existence. First and foremost, I offer my humble obeisance to Almighty God for giving me enough courage, strength, and mental tenacity to accomplish this endeavor.

At this juncture, I think of my parents, whose selfless sacrificial life and great efforts with pain and unceasing prayers enabled me to reach the present position in my life. Unfortunately, my father could not witness the completion of my PhD, but I owe it all to him as it would not have been possible without his loving support, patience, understanding, timely guidance, constant motivation, and blessings. His confidence in me was my greatest strength that enabled me to get through the difficult stages during my research. I also owe special thanks to my uncle, aunt, brother, sister-in-law, brother-in- law, and all my cousins for being extremely supportive during my journey of pursuing

PhD. My heartiest gratitude, especially to my sister, Vanita Puri Gakhar, who persistently motivated me to pursue research in Mercer University and was always with me, through thick and thin.

At this moment of accomplishment, it gives me immense pleasure to express my sincere thanks to my advisor, Dr. Ajay K. Banga, who guided me and gave me the

confidence to go on and never hesitated to help me throughout my research. I deem it as my privilege to have got the opportunity to work under his able guidance. This work would not have been possible without his support, guidance, and valuable suggestions.

Under his guidance, I successfully overcame many difficulties and learnt a lot. His kind and humble words of appreciation always motivated me to work harder and perform to the best of my abilities. I would like to express my gratitude to Mrs. Saveta Banga,

Manisha, and Anshul for all the cordiality and affection.

I would like to express my heartfelt thanks to my committee members: Dr. J.

Grady Strom, Dr. Martin D’Souza, Dr. Ravi Palaniappan, and Dr. Vishal Sachdeva for guiding me through my research and also helping in my professional development. I am also very thankful to my collaborators, Dr. Wei Zhang, Dr. Kevin Murnane, Dr. Onkar

Singh for their valuable suggestions and guidance. A special note of thanks to Dean

Hewitt William Matthews and Dean Brian L. Crabtree for their encouragement and support.

My sincere thanks to all my colleagues and friends, Dr. Pooja Bakshi, Dr. Yamini

Gorantla, Dr. Amruta Indapurkar, Dr. Sindhu Sunkara Ganti, Dr. Mahima Manian, Dr.

Arunprasad Sivaraman, Nne Emmanuel Uko, Dr. Xinyi Gao, Behnam Dasht Bozorg,

Sonalika Arup Bhattaccharjee, Ying Jiang, Yujin Kim, Dr. Yang Song, Jessica Anne

Scott, Dr. Hiep Nguyen, Dr. Sucheta D’Sa, Dr. Kimberly Braz Gomes, Dr. Grace Lovia

Allotey-Babington, Trang Nguyen, Ipshita Jayaprakash Menon, Mohammad Shajid

Ashraf Junaid, and Madhura Kale for their timely help, guidance, company, and support.

More than a word of appreciation is recorded for my friends in India: Arshjyoti,

Navjot, Anjali, Jayati, Ankita ma’am, and Dr. Neeraj Kumar, who selflessly supported me from distance, whenever I needed them.

I would like to specially thank Ms. Vivienne Brown, Ms. Mary Catherine, Ms.

Brenda Austrie-Cannaday, Ms. Cherilyn D’Souza, and Mr. Terry Mernard for all their help and assistance in administrative matters.

Finally, I would like to thank all those whose direct and indirect support helped me in completing my dissertation.

TABLE OF CONTENTS

Page ACKNOWLEDGEMENT ...... iv LIST OF FIGURES ...... xiv LIST OF TABLES ...... xvii CHAPTER 1 ...... 1 Introduction ...... 1 CHAPTER 2 ...... 9 Review of literature ...... 9 Transdermal and Topical Drug Delivery ...... 9 Skin Structure...... 10 Crossing the Stratum Corneum ...... 11 Permeation of Actives across Skin ...... 12 Permeation Enhancement Strategies ...... 13 Chemical Penetration Enhancers ...... 14 Microneedles ...... 18 Laser Microporation...... 20 Iontophoresis ...... 22 Lateral Diffusion of Drugs in Skin ...... 25 Microdialysis...... 27 HIV prophylaxis...... 29 Transdermal Delivery for HIV Prophylactic Agents ...... 36 Transdermal Patches ...... 37

CHAPTER 3 ...... 43 Formulation of epigallocatechin-3-gallate hydrogel and evaluation of its intradermal delivery with the aid of microneedles ...... 43 Abstract ...... 43 Introduction ...... 44 Materials ...... 48 Methods ...... 48

vii

TABLE OF CONTENTS (Continued)

Photodegradation studies ...... 48 Formulation of hydrogels ...... 49 Rheological evaluation...... 50 Flow curves...... 50 Amplitude sweep...... 51 Frequency sweep...... 51 Thixotropy shear and oscillation tests...... 51 Preparation of skin samples ...... 52 Microneedle insertion and visualization of microchannels ...... 52 Pore uniformity...... 52 Transepidermal water loss measurement...... 53 Skin resistance evaluation...... 53 In vitro skin permeation studies ...... 54 Extraction recovery study ...... 55 Extraction of EGCG from skin ...... 56 Quantitative analysis ...... 56 Data analysis ...... 57 Results and Discussion ...... 57 Photodegradation studies ...... 57 Rheological evaluation of hydrogels ...... 59 Flow curves...... 60 Amplitude sweep...... 61 Frequency sweep...... 63 Thixotropy shear and oscillation tests...... 63 Characterization of microchannels ...... 65 Pore uniformity...... 65 Transepidermal water loss measurement...... 66 Skin resistance...... 66 In vitro skin permeation studies ...... 67 Recovery studies ...... 69 Extraction of EGCG from skin ...... 70 Conclusion ...... 74

CHAPTER 4 ...... 76 Effects of chemical and physical enhancement techniques on transdermal delivery of 3-fluoroamphetamine hydrochloride ...... 76

viii

TABLE OF CONTENTS (Continued) Abstract ...... 76 Introduction ...... 77 Materials ...... 82 Methods ...... 82 Solubility studies ...... 82 Skin preparation ...... 83 Evaluation of skin integrity ...... 83 In vitro permeation study paradigm ...... 84 Passive permeation of PAL-353...... 85 Effects of a chemical enhancer on the skin permeation of PAL-353...... 85 Microneedle-mediated transdermal delivery of PAL-353...... 85 Transdermal delivery of PAL-353 through ablative laser treated skin. ... 86 Iontophoresis-mediated transdermal delivery of PAL-353...... 86 Calculation of lag time...... 87 Quantitative analysis ...... 87 Data analysis ...... 88 Characterization of skin exposed to a chemical enhancer, microneedles, laser treatment ...... 88 Attenuated Reflectance Fourier Transform Infrared Spectroscopy...... 88 Scanning electron microscopy...... 88 Dye binding...... 89 Pore uniformity...... 89 Confocal laser scanning microscopy ...... 90 Histological evaluation...... 90 Results ...... 91 Solubility studies ...... 91 In vitro permeation studies...... 91 Passive permeation of PAL-353...... 91 Effects of oleic acid on the skin permeation of PAL-353...... 91 Microneedle-mediated transdermal delivery of PAL-353...... 93 Transdermal delivery of PAL-353 through ablative laser treated skin .... 97 Iontophoresis-mediated transdermal delivery of PAL-353……………….101 Discussion...... 102 Conclusion ...... 108

TABLE OF CONTENTS (Continued) CHAPTER 5 ...... 110 In vitro microdialysis as a tool for quantitative analysis of diffusion rate of diclofenac sodium in human skin ...... 110 Abstract ...... 110 Introduction ...... 111 Materials ...... 116 Methods ...... 117 Skin treatment with microneedles and ablative laser ...... 117 Skin integrity and thickness assessment ...... 117 Probe insertion and depth measurement ...... 118 Probe recovery studies ...... 119 In vitro permeation studies...... 119 Investigation of the barrier and adhesive property of the tape ...... 121 Drug distribution in skin ...... 122 Quantitative analysis ...... 123 Data analysis ...... 123 Results ...... 123 Skin integrity and thickness assessment ...... 123 Probe insertion and depth measurement ...... 124 Probe recovery studies ...... 125 In vitro permeation studies...... 126 Skin distribution studies ...... 130 Discussion...... 130 Conclusion ...... 133

CHAPTER 6 ...... 135 Feasibility investigation for transdermal delivery of elvitegravir formulations ...... 135 Abstract ...... 135 Introduction ...... 136 Materials ...... 139 Methods ...... 139 Solubility studies ...... 139 Skin preparation...... 140 Separation of epidermis...... 140 Skin integrity and thickness assessment...... 140 In vitro permeation study paradigm ...... 141 Passive permeation of EVG...... 142 Effect of different chemical enhancers on the skin permeation of EVG. 142 Effect of combination of enhancers on the skin permeation of EVG...... 142

TABLE OF CONTENTS (Continued) Calculation of permeation flux and lag time...... 143 Skin extraction ...... 143 Quantitative analysis ...... 143 Data analysis ...... 144 Results and Discussion ...... 144 Solubility studies ...... 144 In vitro permeation studies...... 145 Passive permeation of EVG...... 145 Effect of chemical enhancers on skin permeation of EVG...... 147 Effect of combination of chemical enhancers ...... 152 Conclusion ...... 154

CHAPTER 7 ...... 155 Development of a transdermal patch of tenofovir alafenamide for HIV prophylaxis 155 Abstract ...... 155 Introduction ...... 156 Materials ...... 159 Methods ...... 160 Slide crystallization studies...... 160 Preparation of drug in adhesive patches ...... 161 Microscopy of acrylate-based patches ...... 162 Coat weight and drug content of the patches ...... 162 In vitro skin permeation studies ...... 163 Separation of epidermis...... 163 Epidermal integrity and thickness assessment…………………………...... 163 In vitro skin permeation set up...... 165 Quantitative analysis ...... 167 Data analysis ...... 167 Results and Discussion ...... 167 Slide crystallization studies...... 167 Formulation of patches ...... 170 In vitro permeation studies...... 177 Epidermal integrity and thickness assessment...... 177 Selection of receptor solution and in vitro drug permeation studies...... 177 Conclusion ...... 183

TABLE OF CONTENTS (Continued) CHAPTER 8 ...... 185 Development of suspension-based transdermal patch of tenofovir alafenamide ...... 185 Abstract ...... 185 Introduction ...... 186 Materials ...... 188 Methods ...... 189 Formulation of suspension patches ...... 189 Visual observations of the suspension patches ...... 189 Effect of homogenization on particle size of TAF...... 192 Coat weight and drug content of the patches ...... 192 In vitro skin permeation studies ...... 192 Separation of epidermis ...... 193 Epidermal Integrity and Thickness Assessment...... 193 In vitro skin permeation set up...... 194 In vitro drug release studies ...... 195 Coat thickness, coat weight, and drug content of optimized patch ...... 195 Quantitative analysis ...... 196 Physical characterizations of optimized patch ...... 196 Peel adhesion...... 196 Tack properties...... 197 Evaluation of skin irritation potential of optimized patch ...... 197 Data analysis ...... 198 Results and Discussion ...... 198 Formulation of suspension patches ...... 198 Visual observation of patches ...... 201 Effect of homogenization on the particle size of TAF ...... 202 Coat weight and drug content ...... 204 In vitro permeation studies...... 205 Epidermal integrity and thickness assessment...... 205 Determination of permeation flux of TAF...... 205 In vitro drug release studies ...... 210 Coat thickness, coat weight, and drug content of optimized patch ...... 211 Physical characterization of the optimized patch ...... 212 Peel adhesion...... 212 Tack properties...... 212 Evaluation of skin irritation potential of optimized patch ...... 213 Conclusion ...... 213

xii

TABLE OF CONTENTS (Continued)

CHAPTER 9 ...... 215 Summary and Conclusions ...... 215

BIBLIOGRAPHY….………………………………………………………………….219

xiii

LIST OF FIGURES

Figure 1. Types of transdermal patches. (a). Single-layer matrix patches (b). Bi-layer matrix patches (c). Multi-layer matrix patches (d). Reservoir patches ...... 37

Figure 3. (a). Fluorescent image of microneedle-treated skin with (b). Pore permeability index values (c). Histogram generated using Fluoropore software ...... 65

Figure 4. (a). Transepidermal water loss and (b). skin resistance values, pre- and post- microneedle insertion in dermatomed porcine skin ...... 67

Figure 5. Extraction recovery of EGCG from viable epidermis and dermis ...... 70

Figure 6. Distribution of EGCG in different layers of stratum corneum of dermatomed porcine skin using tape stripping technique ...... 71

Figure 7. Average amount of EGCG in viable epidermis and dermis of dermatomed porcine skin after 24 h permeation...... 73

Figure 8. Effect of oleic acid on permeation of PAL-353 through dermatomed human skin ...... 92

Figure 9. The FTIR spectra of human stratum corneum: (a). Untreated (control) (b). Treated with propylene glycol (c). Treated with 5% (w/w) oleic acid in propylene glycol...... 92

Figure 10. Microneedle-mediated transdermal delivery of PAL-353 through dermatomed human skin ...... 94

Figure 11. (a and b). SEM of untreated and microneedle-treated skin, respectively (c). Dye binding (d). Calcein imaging (e). Pore permeability index values (f and g). Histology of untreated and microneedle-treated skin, respectively ...... 95

xiv

LIST OF FIGURES (Continued)

Figure 12. Effect of iontophoresis and ablative laser treatment on transdermal permeation of PAL-353 across dermatomed human skin ...... 97

Figure 13. (a and b). SEM (c). Dye binding (d). Calcein imaging (e). Pore permeability index values (f). Histological evaluation (g). CLSM image showing distribution of calcein in XY plane of laser porated skin ...... 99

Figure 14. Confocal microscopy z-stack of microchannels created by P.L.E.A.S.E.® laser in human dermatomed skin ...... 101

Figure 15. Schematic representation of vertical and lateral diffusion of drug molecules across stratum corneum...... 112

Figure 16. (a). Linear microdialysis probes in dermatomed human skin (b). Skin with probes mounted on Franz diffusion cell (c). Schematic representation of the in vitro microdialysis set up...... 120

Figure 17. Microscopic images of probe in (a). Dermatomed human cadaver skin (b). Dermis layer of human skin; DermaScan images of dermatomed human skin (c) without probe, (d) with probe...... 125

Figure 18. Recovery of diclofenac sodium from linear microdialysis probe ...... 126

Figure 19. Rate of diffusion in (a). vertical direction into the central probe and (b). horizontal direction into the lateral probe ...... 127

Figure 20. Drug levels in (a). receptor fluid and (b) skin layers ...... 128

Figure 21. Fluorescent images of (a). with-tape skin (b). backside of tape ...... 129

Figure 22. Chemical structure of EVG ...... 138

Figure 23. Passive permeation profile of EVG ...... 146

Figure 24. Effect of chemical enhancers on EVG permeation through human epidermis ...... 148

Figure 25. Effect of ethanol and lauric acid on EVG permeation through human epidermis during day 1 ...... 148

Figure 26. Lag times for EVG permeation through human epidermis in different test groups ...... 149

LIST OF FIGURES (Continued)

Figure 27. Amount of EVG retained in human epidermis after 7 days ...... 151

Figure 28. Effect of combination of enhancers on EVG permeation through human epidermis ...... 153

Figure 29. Images of slide crystallization studies, under optical microscope ...... 169

Figure 30. TAF crystals in acrylate-based transdermal patches (at 10x) ...... 173

Figure 31. Permeation profiles of TAF from acrylate-based patches across human epidermis ...... 179

Figure 32. Effect of combination of enhancers and higher patch coat weight on TAF permeation across human epidermis from acrylate patches: a. Seven day permeation profile b. Six hour permeation profile ...... 181

Figure 33. Permeation profile of acrylate-based TAF patches comprising of oleyl alcohol and comparison with the control patch through human epidermis ...... 182

Figure 34. Microscopic images of pure TAF particles suspended in heptane ...... 204

Figure 35. Comparison of permeation profile of TAF from silicone and PIB based patches through human epidermis ...... 206

Figure 36. Comparison of permeation profile of TAF from silicone-based suspension patches through human epidermis ...... 207

Figure 37. Effect of TAF concentration in the silicone-based suspension patches on its permeation through human epidermis ...... 208

Figure 38. In vitro release profiles of TAF from the optimized suspension patch ...... 211

xvi

LIST OF TABLES

Table 1. Flow properties of hydrogels as calculated by Herschel Bulkley model ...... 49

Table 2. Specifications of the test groups evaluated in the in vitro permeation study ..... 68

Table 3. Solubility of EVG in different solvents ...... 145

Table 4. Composition of TAF containing acrylate-based transdermal patches ...... 164

Table 5. Microscopy observations of acrylate-based TAF containing transdermal patches for 6 months ...... 173

Table 6. Coat weight and drug content of acrylate-based clear patches ...... 176

Table 7. TAF stability in different receptor solutions...... 178

Table 8. Composition and manufacturing parameters of the silicone-based TAF containing suspension transdermal patches ...... 190

Table 9. Visual observations of suspension patches ...... 201

Table 10. Coat weight and drug content of silicone-based suspension patches ...... 205

Table 11. Characterization parameters of optimized suspension patch ...... 211

xvii

ABSTRACT

ENHANCED DELIVERY OF ACTIVES THROUGH SKIN FROM PATCHES AND FORMULATIONS, AND DISTRIBUTION WITHIN AND ACROSS SKIN Under the direction of Dr. Ajay K. Banga Professor and Department Chair, Pharmaceutical Sciences Mercer University, Atlanta GA

Transdermal drug delivery, an attractive alternative to other routes of delivery, allows drugs to reach the systemic circulation by traversing the skin barrier. Skin, with a surface area of 1-2 m2, is a readily accessible route for local as well as systemic drug delivery. However, a major challenge encountered while developing a successful topical/transdermal drug delivery system is to ensure the penetration of drugs across the outermost lipophilic and dead layer of skin, the stratum corneum, which acts as a barrier to drug delivery. Moderately lipophilic (log P of 1-3), potent, and unionized drug molecules with a molecular weight of <500 Da and melting point <250 °C, tend to diffuse passively through the skin. However, hydrophilic drugs are unable to permeate through the passive route, and lipophilic drugs have tendency of forming depots in the stratum corneum. Thus, enhancement techniques are applied to facilitate delivery of desired therapeutic or prophylactically relevant doses of such actives, topically or transdermally.

In the present study, chemical strategies such as addition of penetration enhancers in the drug formulations and/or physical technologies such as microneedles, anodal

xviii

iontophoresis, and ablative laser were investigated for enhancement in delivery of cosmetic (epigallocatechin-3-gallate) as well as therapeutic (3-fluoroamphetamine) and prophylactic (antiretroviral agents: elvitegravir and tenofovir alafenamide) actives, into and across skin. Also, in vitro microdialysis was explored for simultaneous quantification of vertical and lateral rate of diffusion of diclofenac sodium as model drug, in intact as well as skin microporated with microneedles and laser. Furthermore, 7-day transdermal patches of tenofovir alafenamide, including various chemical enhancers were formulated and investigated for in vitro permeation across human epidermis.

Overall, the chemical and physical enhancement techniques were found to significantly enhance the skin permeation of different actives, as compared to their respective passive treatments. Also, the microdialysis studies revealed significant enhancement in lateral and vertical diffusion of diclofenac sodium in dermatomed human skin porated with microneedles and ablative laser, as compared to intact skin.

Furthermore, a 7-day suspension-based transdermal patch of tenofovir alafenamide that could successfully deliver the desired HIV prophylactic dose was developed.

xix

CHAPTER 1

INTRODUCTION

Skin is a readily accessible and commonly employed route for local as well as systemic delivery of drugs. It offers a number of advantages over other administration routes such as oral and parenteral. These include preventing first-pass metabolism of drugs, eliminating fluctuations in plasma drug concentrations, avoiding drug degradation in the gastrointestinal tract, enabling sustained delivery for drugs with short half-life and overall, providing patient compliance due to being non-invasive and convenient to administer as well as terminate the treatment in case of any adverse effects (Alkilani,

McCrudden, & Donnelly, 2015; Banga, 2011; Puri, Sivaraman, Zhang, Clark, & Banga,

2017). However, a major challenge encountered while developing a successful topical/transdermal drug delivery system is to ensure the ability of the compounds to penetrate the outermost dead layer of skin, the stratum corneum, which acts as a barrier to drug delivery. It comprises of keratinized corneocytes that are embedded in a lipophilic matrix and thus, hydrophilic drug compounds are unable to traverse it and lipophilic drugs tend to form a depot in this layer (Alkilani et al., 2015; Banga, 2011; Ghosh,

Pfister, & Yum, 1997; Naik, Kalia, & Guy, 2000; Puri, Sivaraman, et al., 2017).

Transdermal drug delivery can be achieved passively as well as with the aid of chemical and physical enhancement techniques. Moderately lipophilic (log P of 1-3), potent, and unionized drug molecules with a molecular weight of <500 Da and melting point <250 °C, tend to permeate passively through skin. However, passive transdermal

1 1 2

delivery of macromolecules, hydrophilic as well as highly lipophilic drugs is difficult to achieve (Puri, Sivaraman, et al., 2017). Therefore, different strategies are widely applied to investigate enhancement in the transdermal permeation of many such drugs. These include the use of chemical penetration enhancers and physical enhancement techniques such as microneedles, iontophoresis, sonophoresis, ablative laser, electroporation, microdermabrasion, radiofrequency, and thermal ablation (Banga, 2011; Puri, Murnane,

Blough, & Banga, 2017; Puri, Sivaraman, et al., 2017).

Chemical penetration enhancers are included in a large number of dermatological, transdermal, as well as cosmetic formulations to help enhance the dermal absorption of lipophilic and hydrophilic active pharmaceutical ingredients, and were thus employed in the present study (Puri, Murnane, et al., 2017; A. C. Williams & Barry, 2004).

Iontophoresis, skin microporation by microneedles, and ablative laser are the physical strategies that were used to investigate the enhancement in topical/transdermal delivery of different cosmetic and therapeutic agents in this study. The choice of enhancement technique for a particular drug was based on its physicochemical properties as well as clinical applicability.

Iontophoresis involves application of low and constant current (<0.5 mA/cm2) that drives charged, neutral or weakly charged drug molecules, into or through the skin and results in high drug fluxes. Transdermal delivery of more than 50 drugs including proteins and peptides has been investigated using iontophoresis to date (Banga, 2011).

Skin microporation with the use of microneedles is an emerging technology that has been reported to improve intra- as well as transdermal delivery of micro- and macromolecules.

Microneedles are micron-sized needles that are available in different lengths (25–2000

µm) and are designed to create pores in the stratum corneum and deliver drugs into deeper skin layers without causing pain or discomfort (Puri, Murnane, et al., 2017). Skin microporation using different microneedles has been shown to enhance the transdermal delivery of nanoencapsulated Rhodamine B dye, amlodipine and verapamil hydrochloride, PEGylated naltrexone prodrug, mannitol, etc (Nguyen & Banga, 2015). A technology, called P.L.E.A.S.E.® (Precise Laser Epidermal System; Pantec Biosolutions

AG, Liechtenstein) has been developed for laser microporation. It comprises of a diode- pumped fractional Erbium-Yttrium-Aluminum-Garnet (Er:YAG) laser that emits light of

2940 nm and creates a matrix of identical micropores that are about 100–150 μm wide apart. Moreover, laser is directed in the form of microbeams and therefore, it fractionally ablates skin with less damage. This strategy has been reported to significantly enhance transdermal permeation of lipophilic as well as hydrophilic drugs, besides its use in clinics for resurfacing of scars, rhytides, and photodamage (W. R. Lee, Shen, Lai, Hu, &

Fang, 2001; Puri, Murnane, et al., 2017).

Microneedles were used for investigation in enhancement in delivery of a cosmetic agent, epigallocatchin-3-gallate (EGCG) and all the above mentioned chemical and physical strategies were employed for enhancing delivery of therapeutic agent, 3- fluoroamphetamine hydrochloride (PAL-353), into and through dermatomed human skin.

Also, the effects of skin microporation via microneedles and ablative laser, on vertical and lateral diffusion of diclofenac sodium (model drug) in human skin were evaluated.

Further, effects of chemical enhancers on the permeation of HIV prophylactic agents, elvitegravir (EVG) and tenofovir alafenamide (TAF) were investigated. Transdermal patches of different types (clear and suspension type), comprising of TAF and suitable

chemical enhancers were finally formulated and investigated for in vitro skin permeation across human epidermis.

Green tea, obtained as an extract from the leaves of Camellia sinensis, is one of the most popular beverage consumed worldwide. Polyphenolic compounds known as catechins, account for 30% dry weight of green tea. Of all catechins, EGCG is the most abundant, constituting about more than 50% polyphenolic fraction. It is also pharmacologically the most significant and potent green tea catechin and has been reported to have antioxidant, anti-inflammatory, and anticancer properties. Many topical green tea extract based skin care cosmetic products are available in the market such as face creams, lotions, soaps, moisturizers, and hydrogel masks but formulations consisting of EGCG as the active ingredient are currently not available in the market. Low skin permeability of EGCG due to the presence of gallate group that results in its high molecular weight and also increases its binding to the lipid bilayers of the stratum corneum, has been reported in literature. An ideal antioxidant must, however, diffuse across the stratum corneum and reach the viable epidermis and dermis, where the UV radiations penetrate and have damaging effects. Thus, maltose microneedles were used for enhancing the intradermal delivery of EGCG in this study. A photostable hydrogel of

EGCG was first formulated and its delivery was then assessed through dermatomed porcine ear skin using maltose microneedles (Puri, Nguyen, & Banga, 2016).

PAL-353 is a novel phenethylamine substrate-based dopamine/norepinephrine releaser that has shown promising efficacy in preclinical models and double-blind placebo controlled clinical trials for cocaine-use disorder (Negus & Henningfield, 2015;

Puri, Murnane, et al., 2017). Transdermal formulations of this drug can provide slow and

sustained drug delivery as compared to the short acting parenteral dosage forms and shall be more patient-compliant as well. Hydrochloride salt of 3-fluoroamphetamine was studied in this investigation. Physical enhancement techniques are preferably applied for the permeation of hydrophilic molecules due to aqueous nature of the channels created by microneedles and ablative laser as well as due to the requirement of the protocol for iontophoresis. Thus, in this study, passive transdermal delivery of the hydrophilic drug molecule was initially investigated and further, the effects of oleic acid as chemical enhancer and physical enhancement techniques such as microneedles, iontophoresis, and ablative laser microporation on its permeation through dermatomed human skin, were investigated (Banga, 2011).

Migration of drug molecules through skin is usually interpreted as vertical diffusion through stratum corneum, viable epidermis, dermis, and into systemic circulation. However, it has been reported that drugs have a tendency to diffuse both vertically and horizontally in skin, depending on their physicochemical properties and concentration gradient (Nayak, Short, & Das, 2015). The process of lateral drug diffusion occurs in conjunction with the process of vertical drug penetration into the skin (Johnson et al., 1996). However, majority of in vitro skin permeation studies using vertical Franz diffusion cells, evaluate the ability of substances to cross the stratum corneum, giving little or no consideration to the phenomenon of lateral diffusion (Nayak et al., 2015). The overall extent of lateral diffusion of drugs in skin may vary depending on their physicochemical properties, but it is an important phenomenon which must be extensively studied and taken into account while evaluating the transdermal permeation of drugs in vitro. Therefore, in this study, vertical and lateral diffusion of diclofenac

sodium (model drug) was simultaneously evaluated using quantitative analytical method of in vitro microdialysis in intact dermatomed human skin. Furthermore, the effect of two physical enhancement techniques used for skin microporation, namely, microneedles and ablative laser, on vertical and lateral diffusion of diclofenac sodium in skin, using the same microdialysis set up, was also investigated (Nguyen, Puri, Bhattaccharjee, & Banga,

2018).

A number of antiretroviral drugs, pre- and post-exposure prophylactic agents, and microbicides are under investigation for HIV prophylaxis. Drug delivery systems such as oral tablets and coitally-dependent vaginal gels require daily application, are short-acting, and are associated with user adherence, while the coitally-independent systems such as implants and injectables are long-acting for a duration of several months, during which the termination of therapy is impractical in case of adverse effects (Gunawardana et al.,

2015; Q. A. Karim et al., 2010; Nelson et al., 2015). An effective drug delivery system to be used for an intermediate duration, if available, would be an attractive alternative option for users in terms of adherence. Transdermal patches, overcoming most of the limitations of the other routes of administration, and aiming to provide sustained delivery of drugs through skin, can be one of the options.

The HIV prophylactic agents, EVG and TAF, that were investigated in this study belong to the category of HIV-1 integrase and nucleotide reverse transcriptase inhibitors, respectively (Ramanathan, Mathias, German, & Kearney, 2011; Schlesinger et al., 2016).

EVG has been reported to have low oral bioavailability of about 30% in rats and dogs, due to extensive pre-systemic first pass effect (Ramanathan et al., 2011). Transdermal delivery system for the same would thus, prevent the premature metabolism of the drug

as the hepatic first-pass effect is by-passed via this route that would further result in improved bioavailability. Passive delivery of EVG solution through human epidermis was investigated. Further, the effect of different chemical enhancers on the same was also explored. However, as the desired HIV prophylactic permeation flux of 25 µg/cm2/h for

EVG could not be achieved even with the use of different enhancers, individually as well as in combination, it was not feasible to proceed with the formulation of transdermal patch for this drug.

In addition to its therapeutic effects, TAF is a potential prophylactic agent for

HIV. Owing to its moderately lipophilic properties, this drug was considered as a better candidate for preparation of transdermal patches than EVG. Clear and transparent patches

(containing the drug in the solubilized form) as well as suspension-type patches

(containing the drug in the dispersed form), comprising of different enhancers were formulated for TAF. Transdermal patches containing the dissolved drug, depicted stability problems in terms of drug crystallization and thus, achieving the targeted permeation flux (7 µg/cm2/h) was not possible. However, a 7-day silicone-based suspension-type patch developed for the same drug, demonstrated the feasibility of achieving the targeted permeation flux through human epidermis, with no issues of drug crystallization.

The novel elements in this study are as follows: Firstly, a topical formulation of

EGCG is not available commercially and significant enhancement in its intradermal delivery with the aid of microneedles has not been demonstrated before. Secondly, this is the first study to investigate and report the transdermal delivery of PAL-353, passively as well as with the use of different chemical and physical enhancement techniques. Thirdly,

the method of in vitro microdialysis has not been used before for quantitative estimation of vertical or lateral diffusion of drugs in intact as well as microporated human skin. The current study demonstrated the use of a novel, effective, and reliable technique of in vitro microdialysis for evaluation of rate of drug diffusion in human skin tissue. Furthermore, transdermal delivery of the HIV prophylactic agents that were investigated has not been reported in the literature till date. Also, development of a 7-day suspension type transdermal patch of TAF for the purpose of HIV prophylaxis, is novel as well.

The six specific aims included in this dissertation are as follows:

I. Formulation of epigallocatechin-3-gallate hydrogel and evaluation of its

intradermal delivery with the aid of microneedles.

II. Effects of chemical and physical enhancement techniques on transdermal delivery

of 3-fluoroamphetamine hydrochloride.

III. In vitro microdialysis as a tool for quantitative analysis of diffusion rate of

diclofenac sodium in human skin.

IV. Feasibility investigation for transdermal delivery of elvitegravir formulations.

V. Development of a transdermal patch of tenofovir alafenamide for HIV prophylaxis.

VI. Development of suspension based transdermal patch of tenofovir alafenamide for

HIV prophylaxis.

CHAPTER 2

REVIEW OF LITERATURE

Transdermal and Topical Drug Delivery

Delivery of drugs via transdermal route is an attractive alternative to oral, parenteral, and other routes of drug administration (Bronaugh & Maibach, 2005; Guy,

2010; M. R. Prausnitz & Langer, 2008; A. Williams, 2003). This can be attributed to the large surface area of skin (1–2 m2), that provides easy access for local and systemic delivery of different actives as well as various other advantages including (1) prevention of premature metabolism of drugs as it bypasses the hepatic first pass effect observed in case of orally administered drugs; (2) avoidance of degradation of drugs in the gastrointestinal tract; (3) elimination of the fluctuations in the plasma drug concentrations that are commonly observed with injectable and oral drug delivery systems; (4) enabling sustained delivery for drugs with short biological half-life that usually require frequent dosing, and (5) enhanced patient compliance due to non-invasiveness, convenience and ease of administration as well as discontinuation of treatment in case of any adverse effects (Banga, 2011; M. R. Prausnitz & Langer, 2008; Puri, Sivaraman, et al., 2017).

However, the greatest challenge in case of transdermal drug delivery is that only a very limited number of drugs/actives can be administered via this route, owing to the structural and barrier properties of skin (Banga, 2011; H. A. E. Benson & Watkinson, 2011; M. R.

Prausnitz & Langer, 2008).

9 10

Skin Structure

Skin comprises of three main layers: epidermis, dermis, and subcutis (Lai-Cheong

& McGrath, 2009). The epidermis consists of constantly differentiating cells and is divided into five layers, namely, stratum corneum, stratum lucidum, stratum granulosum, stratum spinosum, and stratum basale (outermost to innermost). All the layers excluding stratum corneum constitute the viable epidermis. Total skin thickness is 2-3 mm, but that of entire epidermis is 100-150 μm. Stratum corneum, the outermost lipophilic layer, 10-

15 µm in thickness, is the major barrier to permeation of drugs across skin. Lipid-rich matrix with embedded corneocytes constitutes the stratum corneum and is responsible for its barrier properties (Banga, 2011; Ita, 2014; Lai-Cheong & McGrath, 2009). The corneocytes are made up of cross-linked keratin fibres, and are held together by corneodesmosomes that provide structural stability to stratum corneum. Lipids that form the extracellular matrix of this layer and primarily confer barrier properties to this layer include cholesterol, ceramides, and majorly saturated long chain free fatty acids (Ita,

2014; M R Prausnitz, Mitragotri, & Langer, 2004). Epidermis is dynamic in nature as desquamation in stratum corneum is continuously balanced by cell growths occurring in the lower layers. Keratinocytes constantly move from the basal layer, differentiate and migrate upwards through the stratum spinosum and stratum granulosum. Corneocytes are formed as a result of structural changes in the keratinocytes. The lipid phase behavior of stratum corneum is evidently different as compared to other biological membranes, owing to its exceptional lipophilic composition. Although crystalline phases are predominantly found in the stratum corneum, but a subclass of lipids has been reported to form a liquid phase as well. Therefore, both the crystalline nature as well as the presence

of a lamellar phase (13 nm) are considered to be essential for the skin barrier function

(Bouwstra, Honeywell-Nguyen, Gooris, & Ponec, 2003).

The dermis layer is composed of connective tissue and contains nerves, collagen and elastin fibers, blood and lymphatic vessels, and appendages such as hair follicles, sebaceous glands, and sweat glands. Sebum, a mixture of triglycerides, waxes, and phospholipids, secreted by the sebaceous glands, lubricates the skin and aids in maintaining the pH of around 5 (Banga, 2011; Lai-Cheong & McGrath, 2009).

Underlying the dermis, is the subcutaneous layer consisting of lipocytes, arranged as fat lobules. Most of the body fat (80%) is located in this layer. In addition to providing support to skin and insulation from cold, fat also has some endocrine functions, including generation of leptin hormone that regulates appetite and controls metabolic energy (Ita,

2014; Lai-Cheong & McGrath, 2009; Rivera-Gonzalez, Shook, & Horsley, 2014).

Crossing the Stratum Corneum

The barrier properties of stratum corneum allow very few actives to traverse the skin in therapeutically relevant doses (Ita, 2014). Drugs permeate the skin primarily via three pathways: transcellular, intercellular, and trans-appendageal (Banga, 2011; H. A. E.

Benson & Watkinson, 2011). The transcellular pathway is generally regarded as a polar route as the corneocytes consist of polar and relatively hydrated keratin matrix

(Scheuplein, 1965). However, for the drugs to successfully permeate through stratum corneum, their repeated partitioning between the polar environment and lipophilic domains, surrounding the corneocytes, is essentially required. Therefore, it has been very well-reported in literature that transport of drugs through the stratum corneum occurs predominantly by the intercellular route (H. A. E. Benson & Watkinson, 2011). Although

the lipid bilayers present between the corneocytes occupy a small area of stratum corneum, they provide a continuous pathway for drug permeation (Scheuplein & Blank,

1971). The importance of this route has been demonstrated by the solvents that have been found to delipidize the stratum corneum and enhance drug permeation. This has been shown by histological microscopy studies as well (Boddé, van den Brink, Koerten, & de

Haan, 1991; Harada, Murakami, Yata, & Yamamoto, 1992). Trans-appendageal route is considered as the lowest resistance pathway that dominates mainly during the lag phase of the drug diffusion process. Appendages represent only 0.1 – 1% of the total skin surface area, and are targeted for follicular drug delivery. This type of delivery can be accomplished by either modifying the target molecule in terms of size, solubility, charge, and lipophilicity or by manipulation of the formulations by preparing vesicle particles based dosage forms and including excipients that are miscible with sebum (Ita, 2014).

Permeation of Actives across Skin

Only compounds with specific physicochemical properties, in terms of octanol- water partition coefficient (log P), charge, molecular weight, melting point, water solubility, and number of atoms available for hydrogen bonding are able to diffuse passively through skin (Banga, 2011; H. A. E. Benson & Watkinson, 2011; Ita, 2014).

Drugs should be moderately lipophilic (log P: 1-3) in order to be able to significantly permeate passively into stratum corneum and then diffuse out from it into the aqueous blood circulation in the dermis. Hydrophilic drugs with log P<1 are not able to diffuse into the lipophilic stratum corneum that forms the main barrier to drug permeation. On the other hand, very lipophilic drugs with log P > 3, for example, corticosteroids and antifungals, bind to stratum corneum and form a depot or reservoir (Banga, 2011;

Donnelly et al., 2012; Schulz et al., 2017). Aqueous drug solubility of >1 mg/mL and dose of < 20 mg/day is usually considered ideal for developing transdermal delivery systems (Banga, 2011; Chandrashekar & Shobha Rani, 2008; Puri, Sivaraman, et al.,

2017; I. Singh & Morris, 2011). Further, generally unionized drugs with molecular weight of less than 500 Da tend to permeate passively through skin (Banga, 2011; Puri,

Sivaraman, et al., 2017). Also, melting point of less than 200°C is desirable for passive permeation of drugs. Drugs with lower melting point have greater solubility in skin, and therefore, higher permeation flux (Banga, 2011; Puri, Sivaraman, et al., 2017).

Permeation Enhancement Strategies

Transdermal delivery of hydrophilic, highly lipophilic as well as high molecular weight compounds is extremely challenging (Khavari, 1997; Pettit & Gombotz, 1998).

Several different passive and active enhancement strategies have been employed to overcome the stratum corneum barrier and aid in improving the delivery of such actives, into and across skin. Passive techniques include the use of chemical penetration enhancers, eutectic mixtures or prodrugs for formulation development (Banga, 2011).

Also, formulating vesicular or carrier-based delivery systems such as liposomes, niosomes, microemulsions, nanoparticles is a passive approach to enhance drug delivery across skin (Osborne, Ward, & O’neill, 1988; Pierre & dos Santos Miranda Costa, 2011;

Walters & Hadgraft, 1993). Among the physical enhancement techniques, iontophoresis, electroporation, sonophoresis, microneedles, magnetophoresis, ablative laser microporation, microdermabrasion, radiofrequency and thermal ablation are some of the options that are commonly explored (Banga, 1998, 2011; Denet, Vanbever, & Préat,

2004; Kalia, Naik, Garrison, & Guy, 2004; Kalluri et al., 2011; Li, Guo, & Lu, 2013;

Mitragotri & Kost, 2004; M. R. Prausnitz & Langer, 2008; Rajan & Grimes, 2002;

Sivamani, Liepmann, & Maibach, 2007; Wu & Nyborg, 2006).

Chemical Penetration Enhancers

Chemical enhancers are commonly added into transdermal formulations such as patches, creams, and gels to improvise the penetration kinetic profile of drugs, into and across skin. This is one of the most widely applied formulation tools that aids in drug permeation enhancement by reversibly modifying the skin barrier resistance (Puri,

Murnane, et al., 2017; Puri, Sivaraman, et al., 2017). These enhancers act by either perturbing the lipid matrix organization in the stratum corneum or by enhancing the solubility of the actives in the lipophilic barrier layer, eventually resulting in their partitioning into skin. Different compounds that have been extensively investigated for skin penetration enhancement capability, include alcohols and alkanols (e.g., ethanol), fatty acids (e.g., oleic acid, lauric acid), fatty acid alcohols (e.g., oleyl alcohol), azones

(e.g., laurocapram), sulphoxides (e.g., dimethylsulphoxide - DMSO), glycols (e.g., propylene glycol), pyrrolidones (e.g., 2-pyrrolidone), terpenes, chelating agents, cyclodextrins, and surfactants (e.g., Transcutol®), (Balázs et al., 2016; Banga, 2011;

Mohammed, Hirata, Hadgraft, & Lane, 2014; A. C. Williams & Barry, 2004). Oleic acid, ethanol and oleyl alcohol have been used in marketed estradiol or combination patches.

Also, marketed fentanyl and testosterone patches contain ethanol (Banga, 2011).

Penetration enhancers must ideally be nontoxic, nonallergenic, nonirritating, pharmacologically inert, and also compatible with drugs and excipients. However, it is not possible that a single agent would possess all the desirable attributes of an ideal

enhancer. Therefore, generally a combination of different enhancers is included in the formulations.

Different chemical enhancers that were investigated in this project include oleic acid, lauric acid, oleyl alcohol, propylene glycol, ethanol, DMSO, octisalate, and triacetin. Oleic acid has been reported to enhance the diffusivity of skin permeants by delipidizing and creating permeable defects in stratum corneum lipids. This has been demonstrated using calorimetric and spectrometric measurements. Also, only the cis- form of oleic acid has been found to have the penetration enhancing capability. Oleic acid has previously been found to increase the transdermal permeation of various drugs such as piroxicam, salicylic acid, zidovudine, 5-fluorouracil, either alone or in combination with other enhancers (Mitragotri, 2000; Pathan & Setty, 2009; Saini, Baghel, & Agrawal,

2014). Lauric acid has also been extensively studied and its penetration enhancement effects have been reported as well. However, data precisely depicting its effect on the skin barrier function is lacking (Lane, 2013).

Unsaturated fatty alcohols such as oleyl alcohol have also been reported to delipidize stratum corneum and enhance drug permeation (Santoyo, Arellano, Ygartua, &

Martin, 1995). These alcohols are usually dissolved in the concentration range of 1-10% in propylene glycol as co-solvent before application on skin (Lane, 2013). Literature reports that oleyl alcohol has been found to enhance permeation of different drugs such as piroxicam, melatonin, tenoxicam, etc. (Andega, Kanikkannan, & Singh, 2001; Gwak &

Chun, 2002; A. C. Williams & Barry, 2012).

Propylene glycol is commonly used in skin preparations, either as a co-solvent for poorly soluble drugs or as an enhancer for improvising drug permeation through skin

from topical preparations (Barrett, Hadgraft, Caron, & Sarkany, 1965; Coldman, Poulsen,

& Higuchi, 1969; Hoelgaard & Møllgaard, 1985). Pre-treatment of skin with pure propylene glycol has been observed to reduce the pliability of skin barrier, which may be attributed to dehydration (Ostrenga, Steinmetz, Poulsen, & Yett, 1971). Propylene glycol has been observed to act synergistically with different enhancers and has been reported to significantly enhance the permeation flux of lipophilic drugs in combination with lauric acid, possibly by promoting the penetration of the enhancer into the skin and due to their synergistic lipid fluidization effects on stratum corneum (Banga, 2011). However, exact mechanism of action of propylene glycol as an enhancer is not clearly understood till date.

Ethanol is included in many transdermal products and is also the solvent of choice for formulating patches. Different mechanisms explaining the penetration enhancing effect of ethanol have been reported in literature (Lane, 2013; A. C. Williams & Barry,

2004). Firstly, ethanol acts as a solubility enhancer for poorly soluble drugs in the vehicle itself. Secondly, after permeating into stratum corneum, it alters the solubility properties of skin and thus, improves the partitioning of drugs into the membrane. Thirdly, the evaporative loss of ethanol from the formulation results in increasing the drug concentration beyond saturation, providing a supersaturated state with a greater driving force for permeation. Furthermore, ethanol itself rapidly permeates into skin (steady state flux ~ 1 mg/cm2/h) and thus, can carry the permeants along with it (Berner et al., 1989).

As reported in literature, ethanol has been employed to increase the permeation flux of estradiol, levonorgestrel, 5-fluorouracil, and hydro-cortisone across rat skin and of estradiol across human skin in vivo (A. C. Williams & Barry, 2004).

DMSO is also included in transdermal preparations as a co-solvent as well as penetration enhancer (Lane, 2013; Maibach & Feldmann, 1967). Different mechanisms have been explained for its skin penetration enhancing effects including conformational changes in keratin, interaction with the alkyl chains of stratum corneum lipids, displacement of water bound with keratin, and extraction of lipids from skin (Lane,

2013). However, in order to work as an efficacious enhancer, > 60% DMSO is generally required and such high concentration results in skin irritation. Also, with the use of

DMSO, dimethylsulphide is produced from the solvent that has a foul odor. Therefore, due to these reasons, the use of DMSO in commercially available topical products has been very limited (Lane, 2013; A. C. Williams & Barry, 2004).

Octisalate and triacetin have been included in commercial transdermal and topical products. Mechanism of permeation enhancement by octisalate is not clearly understood and is under investigation. However, due to the structural similarity to azone, as well as comparable observations in attenuated total reflectance spectroscopy and differential scanning calorimetry, octisalate is assumed to mechanistically work in the same way as azone, in terms of delipidizing stratum corneum (Nicoli, Penna, Padula, Colombo, &

Santi, 2006). Also, octisalate has been incorporated as an enhancer in metered spray transdermal systems, sunscreens as well as cosmetic products (Fraser et al., 2007).

Oxytrol® (commercially available patch for overactive bladder) contains triacetin as a penetration enhancer (Nicoli et al., 2006). Also, enhancement in permeation of diclofenac through human skin using 10% triacetin has been reported by Patel et al. (Patel, Patel, &

Patel, 2012). Triacetin possesses solubilizing, plasticizing, anti-microbial, and absorption

accelerating properties (Quan & Deshpanday, 2016). However, exact mechanism by which it enhances transdermal permeation of drugs has not been reported in literature.

Microneedles

Microneedles are currently being investigated as an advanced physical enhancement transdermal strategy for the delivery of small drug molecules as well as macromolecular therapeutic agents. These are micron-sized needles that can porate the stratum corneum layer efficiently and aid in delivering the drugs into the deeper skin layers (epidermis and dermis). It is a minimally invasive transdermal drug delivery strategy that forms hydrophilic micro-channels in the upper layers of skin to facilitate the transport of drugs. Drug delivery via microneedles is a hybrid strategy that combines the advantages of both topical-transdermal (non-invasive) and injectable (invasive) routes of drug administration, and overcomes the major drawbacks associated with these delivery strategies. Unlike injectable delivery systems, no clinician is required in case of microneedles and unlike passive transdermal delivery systems, microneedles facilitate controlled delivery of hydrophilic and high-molecular weight compounds into deeper skin layers and systemic circulation (Bhatnagar, Dave, & Venuganti, 2017).

Microneedles of different types such as solid (silicon, metal, polymer, and ceramic), coated, dissolving, and hollow have been designed and investigated for drug delivery (Banga, 2011; Y. C. Kim, Park, & Prausnitz, 2012; Nguyen & Banga, 2015; M.

R. Prausnitz, 2004). However, due to issues of breakage of needles in skin, complex and costly manufacturing processes associated with metal, glass and/or silicon microneedles, development of biocompatible and biodegradable polymeric microneedles has become popular. Some of the biocompatible polymers include sugars (sucrose, trehalose,

maltose), hyaluronic acid, carboxymethyl cellulose, poly lactic acid, poly glycolic acid, poly lactic glycolic acid, and polyvinylpyrrolidone (A. Machekposhti, Soltani,

Najafizadeh, Ebrahimi, & Chen, 2017). Microneedles are fabricated in different lengths, ranging from 25–2000 μm and are thus used to target drug delivery at different depths in skin and to achieve systemic drug delivery. BD Soluvia™ (Becton Dickinson, NJ, USA) is a commercially available hollow microneedle system for the intradermal delivery of influenza vaccine. It consists of a single 1.5 mm stainless steel microneedle in a glass prefilled syringe (Kang, Song, & Kim, 2012).

Microchannels created by microneedles can be well characterized by scanning electron microscopy, dye-binding/imaging studies, histological evaluation, and confocal microscopy (Nguyen & Banga, 2015). Microneedles have been demonstrated previously to significantly enhance the intradermal delivery of recombinant HIV-1 CN54gp (Pattani et al., 2012), nile red-loaded nano-particles (Donnelly et al., 2010), 5-aminolevulinic acid

(Donnelly et al., 2008), DNA vaccine (N. W. Kim et al., 2014), botulinum toxin A

(Torrisi et al., 2013), nano-encapsulated rhodamine B dye (Gomaa et al., 2012),

PEGylated naltrexone prodrug (Milewski, Yerramreddy, Ghosh, Crooks, & Stinchcomb,

2010), amlodipine, verapamil hydrochloride (Kaur, Ita, Popova, Parikh, & Bair, 2014) and mannitol (Badran, Kuntsche, & Fahr, 2009).

In addition to therapeutic drugs and vaccines, microneedles find major application in cosmetology. In fact, cosmetic application is one of the most promising and successful areas for use of microneedles. While the microneedle-based products intended for therapeutic use are currently in exploratory clinical trial phases, microneedles are already in use in the cosmetic clinics and are also commercially available for self-application

(Bhatnagar et al., 2017). In cosmetology, microneedles are either applied to stimulate the process of natural healing of the skin and help in rejuvenation, reduction in pigmentation, wrinkles and skin scarring or to enhance the delivery of cosmetic actives into the skin by forming microchannels through stratum corneum and epidermis. Application of microneedles with 35% glycolic acid peels was found to improve acne scars (Bhatnagar et al., 2017). Further, application of combination therapy of trichloroacetic acid peel

(15%), subcision, and microneedling by Dermaroller®, was very effective for management of atrophic scars. Microneedle therapy has also been reported to be efficacious for restoring scalp hair and treat alopecia (Dhurat et al., 2013; Y. B. Lee et al.,

2013). Thus, in the present study, maltose microneedles were used to enhance the permeation of EGCG, that has a number of cosmetic applications.

Laser Microporation

Application of lasers for compromising stratum corneum and enhancing delivery of drugs, into and through skin, is gaining popularity and seems to be a promising physical enhancement strategy. Microporation using laser creates micron-sized channels in the skin that results in the enhancement of skin permeation of drugs. Non-ablative and ablative are the two types of lasers that have traditionally been employed for skin treatment (Song, Hemmady, Puri, & Banga, 2018).

Non-ablative lasers heat up the skin tissue and have short recovery times as stratum corneum is not ablated in this case. Heat is produced within the dermal tissue without removal of the overlying stratum corneum in contrast to ablative laser. Non- ablative lasers emit IR light with wavelength of 1000-1500 nm that is weakly absorbed by the superficial layers of skin but penetrates the deeper tissues (Tanzi, Lupton, &

Alster, 2003). These are cheaper, safer, less labor-intensive, and have less healing time as compared to the ablative lasers. These lasers have recently been shown to increase the transdermal permeation of diclofenac sodium and sumatriptan succinate (small molecules) (Ganti & Banga, 2016). However, enhancement in the permeation of large molecular weight compounds has not been demonstrated with the use of these lasers.

Ablative lasers have demonstrated positive results in clinical studies as they cause thermal injury by essentially vaporizing the water in skin and create microscopic channels through stratum corneum and epidermis, resulting in enhanced permeation of drugs. However, these lasers are associated with undesired effects such as pain, redness, oozing, swelling, and longer recovery times (Alexiades-Armenakas, Dover, & Arndt,

2008; Aslam & Alster, 2014). In order to reduce the abovementioned side effects associated with the full ablative lasers, fractional ablative lasers were introduced. Two main types of fractional ablative lasers include 10,600 nm carbon-dioxide and 2,940 nm

Er:YAG laser that operate in the IR region and are absorbed by water. These lasers use fractional photothermolysis to create microscopic treatment zones that are defined as zones of ablation or channels formed in the stratum corneum and epidermis. These microchannels possess a zone of coagulation or thermally necrosed cells around them that is further surrounded by normal skin that is not affected by the laser beam (Manstein,

Herron, Sink, Tanner, & Anderson, 2004). These lasers produce similar skin resurfacing effects as full ablative lasers but with shorter recovery times. The main difference between the two types of ablative lasers is the presence of unaffected/normal skin around the microchannels in case of fractional laser, which promotes faster healing. These lasers

have previously been explored to enhance drug delivery but have not been used in the clinical settings (Ganti & Banga, 2016).

P.L.E.A.S.E.® is a fractional ablative laser based technology that has been developed for skin microporation and was used in the present study. It consists of a diode-pumped Er:YAG laser that radiates light with wavelength of 2940 nm. This specific wavelength is absorbed by water in the skin and creates a matrix of micropores that are identical in nature and 100–150 mm apart from each other. It is a patient-friendly and easy to use technology that can be programmed to control the extent of ablation of stratum corneum by varying parameters such as length, number, pulse energy, and repetition rate (Puri, Murnane, et al., 2017).

Ablative lasers have been reported to significantly enhance transdermal permeation of a number of hydrophilic and lipophilic drugs such as 5-fluorouracil nalbuphine, vitamin C, dextran, indomethacin, 5-aminolevulinic acid, oligonucleotides and DNA (W. R. Lee et al., 2006, 2001; Puri, Murnane, et al., 2017). Furthermore, application of ablative laser for intradermal vaccination or transcutaneous immunotherapy has also been investigated previously (Song et al., 2018).

Iontophoresis

Iontophoresis is another physical enhancement strategy that has been used for over a century to enhance the permeation of drug molecules by application of continuous low-voltage current (usually < 0.5 mA/cm2) (M. R. Prausnitz & Langer, 2008). Skin penetration of charged, weakly charged, and even uncharged drug molecules can be facilitated by iontophoresis (Pikal, 2001). The applied electrical potential repels the charged species into and through the skin and this phenomenon is referred to as electro-

migration, electro-repulsion, or the Nernst–Planck effect (Naik et al., 2000). However, in addition to electro-repulsion, there are two other mechanisms that contribute to the enhancement in drug flux through skin. These include electro-osmosis and current- induced increase in skin permeation, referred to as the damage effect. Electro-osmotic flow is explained as the flux of bulk fluid as a result of potential difference applied across the skin and always occurs in the direction of the flow of counter ions. Because the human skin is negatively charged, the direction of the movement of cations determines the electro-osmotic flow, which is always from anode to cathode. As a result, the cathodal delivery of anions is obstructed, whereas the anodal delivery of cations is facilitated, by electro-osmosis. In addition, the delivery of uncharged drug molecules under iontophoresis is due to electro-osmosis (Rawat, Vengurlekar, Rakesh, Jain, &

Srikarti, 2008).

Iontophoresis has been in clinical use for the delivery of lidocaine (a local anesthetic) (Zempsky, Sullivan, Paulson, & Hoath, 2004), extraction of glucose from skin for glucose monitoring (Rao et al., 1995), and delivery of pilocarpine to induce sweat production for the diagnosis of cystic fibrosis (Beauchamp & Lands, 2005). In addition, tap water iontophoresis is the method of choice used clinically to initiate the treatment of hyperhidrosis (Kreyden, 2004).

More than 50 drugs have been investigated in the last 30 years for transdermal delivery using the iontophoretic technique (Banga, 2011). Transdermal delivery of a number of peptides and proteins with a molecular weight of up to ~ 10–15 kDa has also been investigated. Some of these include thyrotropin releasing hormone, vasopressin, luteinizing hormone releasing hormone, nafarelin, cytochrome c, daniplestim, and

ribonuclease A. Because of the phenomenon of electro-osmosis, anodal iontophoresis is overall more effective than cathodal iontophoresis and is thus a preferred technique for the delivery of proteins with a high isoelectric point because they would be positively charged in the formulation maintained at an appropriate pH, as well as at the physiological pH and in the skin, which has a pH of 4–6 (Banga, 2011; Kalluri et al.,

2011). Solubility of drugs in water is a prerequisite for iontophoresis, so this enhancement technique is mostly applicable for hydrophilic drugs. In addition to solubility, pKa of the drugs and pH of the formulation play an important role in determining the charge on the drugs and their extent of ionization, which affects their iontophoretic delivery. The pH of formulations must be within the range of 4–8 so as to be close to the pH of skin surface (4–6) and physiological pH (7.4) and thus avoid any irritation to the skin (Kalluri et al., 2011).

For lipophilic drugs and those with poor solubility in water (< 1 mg/mL), iontophoretic delivery can be investigated by complexation with anionic cyclodextrins

(Juluri & Murthy, 2014) or encapsulating them in charged solid lipid nano-particles or nano-structured lipid carriers (Charoenputtakun, Li, & Ngawhirunpat, 2015). Type of iontophoresis to be applied for lipophilic drugs is governed by the type of charge on the surface of the complexes, nanoparticles, or other carrier systems. The drug flux rate of transdermal delivery via iontophoresis can be modulated by varying the intensity and duration of electrical current being applied. In addition, it is a good technical approach to incorporate iontophoresis into a transdermal patch system design, which has been demonstrated as evidenced by the fentanyl and sumatriptan iontophoretic patch (Mayes &

Ferrone, 2006).

In addition to performing studies applying the physical enhancement techniques for transdermal drug delivery, effect of two of such technologies, microneedles and laser microporation on lateral diffusion of drugs in skin was investigated using in vitro microdialysis.

Lateral Diffusion of Drugs in Skin

Diffusion of drugs through skin is usually believed to only occur in the vertical direction through stratum corneum, viable epidermis, and into the blood circulation.

However, diffusion in the lateral direction occurs in conjunction with the vertical diffusion, and the extent depends on drug’s physicochemical properties as well as concentration gradient (Jacobi et al., 2004; Nayak et al., 2015; Schicksnus & Muller-

Goymann, 2004). Lateral drug diffusion can occur on the stratum corneum surface, along the lipid bilayers within the stratum corneum or from one bilayer into a parallel bilayer, and finally through the lipophilic passages between the corneocytes into the deeper skin layers (Gee, Nicolazzo, Watkinson, & Finnin, 2012). Molecular weight, lipophilicity, and solubility of the drug in the moistened skin surface determines the extent of lateral diffusion (Gee et al., 2012; Lampe et al., 1983).

Both, hydrophobic and hydrophilic drugs have been observed to diffuse faster in the lateral than vertical direction in the stratum corneum because of the continuous nature of the intercellular pathway (Johnson, Blankschtein, & Langer, 1997). Lateral diffusion is however, more pronounced for actives with molecular weight <300 g/mol (Simonsen,

Petersen, Benfeldt, & Serup, 2002) as was demonstrated in a study by Gee et al., where hydrocortisone (362.5 g/mol) exhibited lower lateral diffusion than ibuprofen (206.3 g/mol) and caffeine (194.5 g/mol) (Gee et al., 2012). Furthermore, as stratum corneum is

lipophilic in nature, drugs with high log P are more likely to undergo lateral diffusion

(Simonsen et al., 2002; Weigmann et al., 1999).

The in vitro skin permeation studies conducted using vertical Franz cells traditionally evaluate the ability of drugs to cross the stratum corneum with mostly no consideration to the phenomenon of lateral diffusion (Nayak et al., 2015). Overall, the extent of lateral diffusion of different drugs in skin may vary depending on their physicochemical properties but it is an essential phenomenon that must be extensively investigated and taken into account when evaluating their in vitro transdermal permeation. However, only few techniques have been applied to study this phenomenon of drugs in skin till date. Fluorescence-based methods such as confocal laser scanning microscopy, fluorescence recovery after photobleaching, and raster image correlation spectroscopy have previously been used to study lateral diffusion of actives in isolated stratum corneum as well as in lipid models for stratum corneum (Bai, Sachdeva, Kim,

Friden, & Banga, 2014; Mizutani, Mitsutake, Tsuji, Kihara, & Igarashi, 2009; Park, Kim,

Jeong, Li, & Seo, 2010; van Smeden et al., 2011). Vibrational spectroscopic techniques such as confocal Raman microscopy and Fourier transform infrared microscopy have also been used to study lateral diffusion of dithranol in artificial membrane constructs (Gotter,

Faubel, & Neubert, 2010) and urea in human stratum corneum (Hartmann et al., 2004).

Also, studies using skin biopsy and tape stripping protocols to investigate horizontal permeation of butyl salicylate, lidocaine, salicylic acid and other drugs have been performed earlier and reported in literature (Gee et al., 2012; Simonsen et al., 2002).

The effect of physical enhancement techniques on lateral diffusion of drugs has not been investigated extensively before. Urea and DMSO were reported to significantly

enhance the lateral diffusion of ibuprofen (Schicksnus & Muller-Goymann, 2004). Also, just one study has demonstrated increase in the lateral diffusion of bovine Alexafluor

555-serum albumin conjugate with the use of combination of microneedles and iontophoresis (Schicksnus & Muller-Goymann, 2004). The effect of physical enhancement techniques on the lateral diffusion of drugs in skin needed further exploration and was performed in one of our studies.

The above mentioned methods that have previously been employed for studying lateral diffusion of drugs in skin such as fluorescence microscopy and infrared spectroscopy are mostly qualitative in nature and are applicable for drug analysis in the stratum corneum. Other methods such skin biopsy (for epidermis and dermis) and tape stripping (for stratum corneum) are quantitative but can be performed only towards the end of the permeation studies in an in vitro or in vivo set up. Also, investigation of real- time vertical or lateral diffusion of drugs that are intended for transdermal delivery has not been investigated before using microdialysis in an in vitro set up.

Microdialysis

Microdialysis, an in vivo sampling methodology, enables monitoring of analyte concentrations in situ for a prolonged period of time. It provides remarkably high resolution, in terms of space as well as time. A simple microdialysis experimental set up involves inserting a probe, with an outer diameter of about 1 mm mounted with a semipermeable membrane, in the tissue of interest. The probe is continuously perfused with an aqueous solution, preferably physiological isotonic solution. Substances dissolved in the interstitial fluid then diffuse across semi-permeable membrane of the

probe and can finally be detected and analyzed for drug amount using a suitable analytical technique (Erdo, Hashimoto, Karvaly, Nakamichi, & Kato, 2016).

There are two designs of probes that are usually employed for cutaneous microdialysis: linear and concentric. These probes are flexible and soft, unlike the probes used for brain microdialysis. Linear probes are however, thinner and more flexible and cause milder skin reactions than concentric probes (Holmgaard, Nielsen, & Benfeldt,

2010). The location of the probe in skin depends on the thickness of the integument, dimensions and design of the probe and anatomy of the site being evaluated.

Ultrasonographic imaging using a DermaScan can be used to verify the depth of the probe in skin (Erdo et al., 2016). Resolution of the samples is possible in terms of space in addition to time due to the large surface area of skin (Davies, Cooper, Desmond,

Lunte, & Lunte, 2000). Insertion of multiple microdialysis probes has been performed in vivo to determine the diffusion properties of the drugs as well as to investigate a number of experimental factors in parallel (Erdo et al., 2016).

The insertion of a microdialysis probe in skin may cause minor trauma, similar to needle pricking and trigger skin reactions resulting in release of tissue necrosis factor-α and interleukins (Anderson, Andersson, & Wårdell, 1994; Clausen, Kaastrup, &

Stallknecht, 2009; Sjögren, Svensson, & Anderson, 2002). Also, these effects can impact the extent of analyte recovery and thus, in vivo calibration of probe recovery must be performed for accurate assessment of target analyte concentrations. Dermal microdialysis has been conventionally used for different applications such as assessment of biomarkers for skin diseases, investigation of mechanism of skin irritation, monitoring various endogenous clinical markers, determination the rate and extent of permeation of drugs

into the systemic circulation, evaluation of drug concentrations in the skin after topical application in real-time, measurement of topical bioavailability and for establishing bioequivalence between topical products as well (Shukla, Bashaw, Stagni, & Benfeldt,

2014; Tettey-Amlalo, Kanfer, Skinner, Benfeldt, & Verbeeck, 2009).

Some of the limitations of in vivo dermal microdialysis include requirement of sufficiently sensitive analytical method for accurate measurement of low drug concentrations in the dialysates, mild uneasiness for the subjects in terms of limited mobility during the entire study, requirement of equilibration time for the local trauma due to probe insertion to fade away, and applicability for drug compounds with a log P of

1-5 (Erdo et al., 2016). However, dermal microdialysis in an in vitro settings has not been explored before. Therefore, in one of the projects, we designed an in vitro microdialysis experiment to simultaneously investigate vertical and lateral diffusion of diclofenac sodium through intact as well as microneedle and laser microporated skin (Nguyen et al.,

2018).

HIV prophylaxis1

According to the World Health Organization, HIV infection continues to be the most challenging public health issue, having claimed about 34 million lives to date

(“WHO | Definition of key terms,” 2013). The anti-retroviral therapy (ART) regimens have been found to be successful in preventing the progression of HIV infection to severe

1 Puri, A., Sivaraman, A., Zhang, W., Clark, M. R., & Banga, A. K. (2017). Expanding the domain of drug delivery for HIV prevention: exploration of the transdermal route. Critical Reviews in Therapeutic Drug Carrier Systems, 34(6), 551–587. http://doi.org/10.1615/CritRevTherDrugCarrierSyst.2017020147

manifestations such as AIDS and other opportunistic infections and have been accessed by millions of people, but significant limitations such as long-term administration, poor oral bioavailability, inaccessibility to tissues and reservoirs, patient non-compliance due to side effects, and chances of developing drug resistance persist (Ramana, Anand,

Sethuraman, & Krishnan, 2014). Constant efforts to develop strategies for HIV prevention using antiretroviral (ARV) drugs, pre-exposure prophylactic (PrEP) agents and post-exposure prophylactic agents, microbicides, and HIV vaccines have been made, but so far, with limited success (Veselinovic et al., 2014). Truvada®, a combination of tenofovir disoproxil fumarate (TDF) and emtricitabine, the only drug product approved for oral PrEP, possesses similar limitations as ART (Nelson et al., 2015). Microbicides, the drug products that are topically (vaginally or rectally) applied to reduce the risk of sexual transmission of HIV are under investigation (Antimisiaris & Mourtas, 2015).

More than 60 microbicide candidates have been identified to have in vitro microbicidal activity against HIV and several of them are being currently tested in clinical trials, but an efficacious and safe microbicide product is still not commercially available (Nutan &

Gupta, 2011). In addition, the development of an effective HIV vaccine continues to be challenging due to the genetic diversity of HIV, the difficulty in developing highly immunogenic antigens, and ambiguity about the constituents of protective immunity

(Maartens, Celum, & Lewin, 2014). The goal of preventing HIV transmission remains a big challenge, so the demand is high for the innovative design and development of different drug-delivery systems/devices for some promising ARVs with different mechanisms of action that may be administered by different routes to achieve this goal.

A number of ARVs (fusion inhibitors, non-nucleoside reverse transcriptase inhibitors, nucleoside/nucleotide reverse transcriptase inhibitors, integrase strand transfer inhibitors, and protease inhibitors), are being widely evaluated for possible use as topical and systemic PrEP. These drug candidates are formulated in various dosage forms such as gels, ointments, rings, films, implants, tablets, and injectables and are administered via different routes for research purposes. These include vaginal, oral, subdermal, and parenteral routes. Limited success for HIV prevention with the use of early microbicides

(non-ARV-based products) such as BufferGel, carraguard, nonoxynol-9, PRO2000 gel, and cellulose sulfate has been reported in clinical trials. Due to this, the focus shifted towards investigation of ARV formulations for HIV prevention. TFV and TFV-based prodrugs are the most extensively explored ARVs for this purpose. As reported in a clinical trial on CAPRISA004, 1% TFV vaginal gel microbicide product has been demonstrated to be safe and effective in preventing HIV transmission (Q. A. Karim et al.,

2010; Rohan, Devlin, & Yang, 2014). In addition, clinical trials such as Partner’s PrEP,

IPrEx, TDF2, PROUD, and IPERGAY studies showed the combination of TDF and FTC to be effective in reducing HIV acquisition when administered orally (Baeten et al., 2012;

Grant et al., 2010; McCormack et al., 2016; Thigpen et al., 2012). This combination, called as Truvada®, is the only U.S. FDA-approved and licensed oral tablet for PrEP that is prescribed to be taken once daily (Nelson et al., 2015). Furthermore, for prevention of

HIV transmission from mother to child, the standard combination ART (TDF + FTC+ efavirenz) in pregnant and breastfeeding women was demonstrated to supersede the earlier prescribed antepartum zidovudine monotherapy and single dose of nevirapine at the time of labor (Maartens et al., 2014; Tsague & Abrams, 2014).

A variety of delivery systems such as oral/vaginal tablets, vaginal/rectal gels, vaginal ointments, films, rings, electro-spun fibers, and diaphragm-based devices, subdermal implants, oral/parenteral liposomes, and nano-particles as investigational anti-

HIV agents have been reported in the literature (Ball, Krogstad, Chaowanachan, &

Woodrow, 2012; Dou et al., 2006; Gunawardana et al., 2015; Mallipeddi & Rohan, 2010;

Vedha Hari, Devendharan, & Narayanan, 2013).

Oral administration of an effective prophylactic pill seems an attractive prevention strategy that enables individuals to take the medication outside the settings of sexual intercourse at discreet times, which empowers those who might not be able to negotiate with their partners. After Truvada® was approved by the FDA for HIV prevention, the next generation drugs that are currently being evaluated for oral PrEP include maraviroc, rilpivirine, and cabotegravir (Heneine & Kashuba, 2012). However, the oral use of ARVs for PrEP has been shown to have inconsistent effects in preventing

HIV in different clinical trials due to lack of user adherence. In addition, concerns of toxic adverse effects and the potential for development of ARV resistance are associated with the long-term use of these oral medications (Nelson et al., 2015; Ramana et al.,

2014; Veselinovic et al., 2014). Clinical results revealed that Truvada® showed a 44% reduction in HIV infection in the IPrEx trial (Grant et al., 2010), 63% reduction in the

TDF2 study (Thigpen et al., 2012), 75% reduction in the Partners PrEP study (Baeten et al., 2012) and an 86% reduction in the PROUD and IPERGAY trials (McCormack et al.,

2016). However, FEM-PrEP trial with oral Truvada® was stopped due to lack of protection in the HIV high-risk female population (Van Damme et al., 2012). In-depth analysis taking into account the number of patients adhering strictly to the medication

regimen revealed that the effectiveness of the systemic oral PrEP was affected directly by the extent of pill adherence in the subjects (Nelson et al., 2015). Furthermore, oral PrEP has been reported to result in suboptimal drug concentrations in the target mucosal regions compared with plasma. The concentration of TFV diphosphate (active metabolite) in the vaginal tissue after oral administration of TFV and FTC has been observed to be 1000-fold lower that after application of vaginal TFV gel (S. S. A. Karim,

Kashuba, Werner, & Karim, 2011). In addition, the certainty of efficacy of the oral PrEP remains relatively less for women than men. However, because the occurrence of new

HIV infections is reported to be higher in women, FDA approval requires positive results in at-risk women due to mixed evidence of PrEP effectiveness reported in earlier studies.

Therefore, PrEP addresses the requirement for female-controlled prevention strategies such as topical microbicides that have the potential to increase the effectiveness of chemoprophylaxis in women (Jay & Gostin, 2012).

Topical microbicides that are administered vaginally or rectally provide an opportunity for women to protect themselves, especially those who are unable to negotiate with their partners regarding the use of male or female condoms due to social, cultural, or economic reasons. Therefore, they appear to be a potential alternative to oral

PrEP (Antimisiaris & Mourtas, 2015). However, variable clinical results for effectiveness in HIV prevention have also been observed with the use of vaginal and rectal PrEP formulations. The CAPRISA004 trial showed 44% HIV protection with the use of 1% TFV vaginal gel, the application of which was coitally dependent (one dose 12 hours before sex and one up to 12 hours after sex). Conversely, the TFV gel arm in the

VOICE trial was discontinued due to lack of protection, mainly caused by low adherence

in patients (Q. A. Karim et al., 2010; McEnery, 2011; Van Der Straten, Van Damme,

Haberer, & Bangsberg, 2012). Another phase III clinical trial, FACT001, was conducted with an intent to verify the CAPRISA004 clinical outcome on a larger scale and was completed in 2015. However, the results did not confirm the efficacy observed in

CAPRISA004 and showed that the effectiveness of the TFV gel was affected directly by the extent of patient adherence. Variable patient adherence associated with the use of vaginal or rectal preparations could be affected by factors such as the texture, color, size, packaging, labeling, and perceptibility (sensory perceptions such as messiness/leakiness) of the dosage form (Morrow, Fava, et al., 2014; Morrow, Underhill, et al., 2014).

However, frequent administration and coital dependency of these short-acting topical formulations are the major factors that contribute to poor patient adherence. Furthermore, topical/vaginal products have a barrier to use due to the need of inserting them digitally.

To counter the shortcomings of topical vaginal gels, many studies evaluating the efficacy and acceptability of controlled drug release delivery systems for ARVs such as intravaginal rings (coitally independent) and nano-technology-based microbicide products (biodegradable polymeric nano-particles, nano-fibers, liposomes, and inorganic nano-platforms) are currently being performed in preclinical and clinical trials (Nelson et al., 2015; Sánchez-Rodríguez, Vacas-Córdoba, Gómez, De La Mata, & Muñoz-

Fernández, 2015). In two recently reported phase III clinical trials, the Ring Study and

ASPIRE, a monthly vaginal ring incorporating dapivirine was found to be moderately effective for HIV-1 prevention among African women with relatively greater efficacy in groups of subjects with evidence of better adherence (Baeten et al., 2016). In the Ring study, effectiveness of the ring was found to be only 15% in younger age group

participants (18–21) due to poor adherence (Puri, Sivaraman, et al., 2017). In addition,

TFV and TFV/levonorgestrel multipurpose technology intravaginal ring (designed for protection up to 90 days) are currently being tested in phase I clinical study (Puri,

Sivaraman, et al., 2017), demonstrating that coitally independent intravaginal rings can be used as one of the promising pharmaceutical strategies to reduce the adherence issue by providing continuous release of drugs into the vaginal mucosa for 1–3 months.

In addition to intravaginal rings, formulation of ARVs into long-acting injectable delivery systems is being investigated that has potential to enable a 1- to 3-month protective duration for HIV prevention (Eisingerich et al., 2012). A phase I clinical study showed that nanosuspension-based long-acting rilpivirine injectable administered intramuscularly is well tolerated at single and multiple doses and possesses favorable pharmacokinetic characteristics to provide protection at the genital tract tissues susceptible to HIV infection (A. Jackson & McGowan, 2015). Phase II clinical trials assessing the safety and acceptability of long-acting rilpivirine are being currently conducted (Puri, Sivaraman, et al., 2017). Cabotegravir formulated into a nanosuspension-based long-acting injectable for intramuscular administration is another attractive anti-HIV candidate for PrEP for both men and women that is presently being investigated in phase II studies (Puri, Sivaraman, et al., 2017). In addition, the combination use of long acting rilpivirine and cabotegravir injectables has been demonstrated to show effective antiviral activity in phase IIb studies (Margolis et al.,

2016). These promising clinical results suggest that long-acting injectable formulation of

ARV drugs has the potential to provide protective plasma drug concentration for HIV prevention for at least 1–3 months. However, tolerance, safety, management of

discontinuation, and chances of resistance are concerns associated with the long-acting injectables that need to be taken into account. Furthermore, as reported recently, a subdermal implant containing TAF was developed and found to provide sustained drug release over 40 d in beagle dogs (Ray, Fordyce, & Hitchcock, 2016). A biodegradable thin film polymer device for subcutaneous delivery of TAF has also been designed

(Schlesinger et al., 2016). These novel implant devices and other ARV-releasing implant technologies that are in early development hold promise as a strategy to improve adherence by providing sustained release of the drug candidates for a long duration. It is a minimally invasive technique, but may pose a limitation to use due to the requirement of a clinician for its administration (Gunawardana et al., 2015).

Transdermal Delivery for HIV Prophylactic Agents

Oral tablets and topical gel formulations require frequent dosing regimen

(daily or event-driven) due to their short-acting pharmacological effect, whereas intravaginal rings, injectables, and implants are designed strategically to be long-acting for at least 1 to several months. Therefore, due to the need of a method mixture that would provide a new option to the users, it is imperative to design an effective drug- delivery system with an intermediate duration (few days to 1 week). This would be an alternative option for women and probably a preferred one due to better adherence.

Transdermal drug delivery systems aimed at providing sustained and controlled release of drugs through the skin into the systemic circulation could be one of the possible strategies. Thus, two of the projects discussed in the further chapters focused on developing transdermal patches for potential HIV prophylactic agents.

Transdermal Patches

Transdermal patches are medicated adhesive systems designed to adhere onto skin and provide systemic drug delivery in a safe and controlled release manner, which can be configured as single (Figure 1a), bi-layer (Figure 1b), or multi-layer (Figure 1c) matrix-based or reservoir-based (Figure 1d) transdermal patches (Puri, Sivaraman, et al.,

2017).

Figure 1. Types of transdermal patches. (a). Single-layer matrix patches (b). Bi-layer matrix patches (c). Multi-layer matrix patches (d). Reservoir patches

Reservoir-based transdermal patches contain drug in the form of solution, gel, or suspension usually sandwiched between the backing and the rate-controlling membrane, followed by an adhesive layer and a release liner in that order. However, the major disadvantages of the reservoir-based patches are dose dumping and abuse potential (Puri,

Sivaraman, et al., 2017). Conversely, matrix type patches contain drug either dissolved or

dispersed in an organic-based pressure-sensitive adhesive such as acrylate copolymer, silicone, or polyisobutylene. The adhesive layer containing the drug is sandwiched between an impermeable backing and a release liner. The amount of drug in the adhesive is usually maintained at sub-saturation level to avoid crystallization over time. The basic components of a transdermal patch include pressure sensitive adhesives (PSAs), release liner, backing membrane, and excipients.

The adhesive properties of transdermal patch determine its efficiency to deliver the required dose of the drug across skin. The entire surface area of the applied patch must be in contact with the skin for efficient drug delivery. Thus, PSAs serve the purpose of adhering to a substrate. An intimate contact between skin and patch is a result of inter- atomic as well as intermolecular forces of attraction between the skin and PSA at the interface. An ideal PSA must adhere upon application of finger pressure, possess aggressive tack properties, exert a strong contact force, and should not leave a residue upon removal (Banerjee, Chattopadhyay, Ghosh, Datta, & Veer, 2014). The adhesive properties of PSAs are affected by the thickness and composition of the matrix and surface free energy, flexibility, roughness, and physicochemical properties of the backing membrane. PSAs are characteristically, viscoelastic in nature. The balance between the amount of elastic energy stored and the viscous flow properties determines the utility of a

PSA material. The different properties of the adhesives that must be essentially evaluated while selecting a PSA for patch development include: skin adhesion properties such as tack and shear, stability, compatibility with drug and excipients, safety, toxicological profile, shelf-life, skin irritation potential and impact of mechanical movement (Banerjee,

Chattopadhyay, Ghosh, Datta, et al., 2014).

Three most commonly employed adhesives for formulation of transdermal patches include acrylic, silicone, and polyisobutylene-based adhesives (Banerjee,

Chattopadhyay, Ghosh, Datta, et al., 2014). PSAs usually consist of an elastomeric polymer, resin, filler, antioxidants, cross-linking agents, and stabilizers if needed. The cohesive properties of the adhesive are attributed to the soft (butyl acrylate, 2-ethylhexyl acrylate) or hard (methyl acrylate, vinyl acetate) monomers, whereas the polarity and wettability for adhesion is provided by the functional monomers (acrylic acid, 2- hydroxyethyl acrylate) present in the adhesive. Acrylate PSAs, include various DURO-

TAK® products, such as DURO-TAK® 87-2516, -2287, -900A, -9301, and -4098 that are marketed by Henkel Corporation in the U.S. These adhesives are extremely stable and don’t degrade until exposed to extreme conditions. Oxidative degradation can occur at high pressure and temperature conditions due to the combination of oxygen with free radicals formed in the polymer. Silicone PSAs include silanol (Si–OH) or Si–OH silylated (Si–O–SiMe3) monomers. Three dimensionally, silicone resin has a silicate structure, which is end capped with trimethyl siloxy groups and also contains residual silanol groups (Woodward & Metevia, 1985). It is prepared by a condensation reaction between the silicate resin and silanol groups on the poly(dimethylsiloxane) polymer that results in formation of siloxane bonds (Si–O–Si) that cross-link the reactants (Banerjee,

Chattopadhyay, Ghosh, Datta, et al., 2014). Amine-compatible silicone adhesives including Dow Corning’s BIO-PSA 7-4101, -4201, -4301, -4102, -4202, -4302 are popularly used for patch development. Polyisobutylene adhesives are completely saturated and possess high oxygen resistance. They usually consist of a mixture of high

(molecular weight: 450,000 - 2100,000 g/mol) and low molecular weight (1000 and

450,000 g/mol) fractions. DURO-TAK 87-6908 marketed by Henkel Corporation, New

Jersey, U.S, is one of the commonly used polyisobutylene adhesive (Banerjee,

Chattopadhyay, Ghosh, Datta, et al., 2014).

Backing membranes are selected for occlusion, appearance, and flexibility. Some of the backing membranes that are commercially used include polyester, polyethylene, polyurethane, ethyl vinyl acetate, polyolefin, polyethylene terephthalate (PET), and polyester laminate film (Kandavilli, Nair, & Panchagnula, 2002). Other factors that must be kept into consideration while selecting backing membranes include leaching of additives from the backing membranes, diffusion of drugs or excipients through the backing, oxygen and moisture-vapor transmission rate. However, higher chemical resistance may result in highly occlusive environment that may eventually cause the patch to possibly lift up and cause skin irritation as well. An ideal backing may be the one with high flexibility or lowest modulus, high moisture–vapor transmission rate as well as good oxygen transmission. For e.g. Co-Tran 9702, 9706, 9720, 9722, Foam tape 9772L,

ScotchpakTM 1109, 1006, 9732, 9733, and 9723 (Banerjee, Chattopadhyay, Ghosh, Datta, et al., 2014).

The release liner is a protective liner/membrane that is removed before applying the patch to the skin. It does not come in direct contact with skin, but material properties of the release liner need to be considered for patch preparation as it is in direct contact with the drug formulation. Release liners are considered as part of primary packaging material and must ideally be inert, impermeable to water and drug formulation, resistant to cracking, etc. (Santoro, Rovati, Lanzini, & Setnikar, 2000). Another very important property to consider while selecting release liners is to check the probability of cross-

linking between the liner and adhesive to be used for formulation (Nakajima-Kambe,

Shigeno-Akutsu, Nomura, Onuma, & Nakahara, 1999). Release liners can be silicone, fluorosilicone, or fluoropolymer coated. Silicone coated liners are preferred when using acrylate adhesives and fluoropolymers are usually employed when working with silicone- based adhesives so as to avoid cross-linking issues and provide easy peel-off. Some examples of the release liners include ScotchpakTM 1022 (fluoropolymer coated),

ScotchpakTM 9709, 9954 (fluorosilicone coated), and silicone-release liner (4999 and

4986). These are all manufactured by 3M Drug Delivery Systems, St. Paul, MN, USA

(Banerjee, Chattopadhyay, Ghosh, Datta, et al., 2014).

Patches may also include excipients such as permeation enhancers, for example: oleic acid, oleyl alcohol, isopropyl myristate, lauric acid; plasticizers such as mineral oil, diethyl or dibutyl phthalate, propylene glycol; colloidal silicon dioxide as adsorbent and viscosity builder/thickener, and crystallization inhibitors such as polyvinyl pyrrolidone (

PVP 360, Kollidon® VA64, Kollidon® 90F, and Kollidon® 30LP), poloxamers, triacetin, oleic acid, etc. As discussed earlier under the section of chemical penetration enhancers, these excipients are added to modulate the permeation properties of the skin and aid in improving drug delivery into and through the skin. Crystallization inhibitors are included in the patches in order to incorporate drugs in amounts higher than its saturation solubility in the adhesive alone. Mechanistically, these additives stabilize the systems against crystallization, either due to enhancing the solubility of the drug or by aiding adsorption of drug crystals in the adhesive matrix. PVP has previously been demonstrated as one of the most effective additives in inhibiting drug crystallization in patch formulations acting by the formation of amorphous co-precipitates with the drugs (K.

Gong, Viboonkiat, Rehman, Buckton, & Darr, 2005; Sekikawa, Nakano, & Arita, 1978).

Also, it has been previously shown that PVP inhibits crystallization of drugs by getting adsorbed on the growing crystal surfaces (Jain & Banga, 2013).

Transdermal patches are usually evaluated for matrix structure formation, rheology, crystallization, cold flow, tack, shear, peel adhesion, skin irritation, and in vitro drug permeation using human cadaver or porcine ear skin and drug pharmacokinetic profile using animal models such as hairless rats, guinea pigs, and dogs (Maillard-Salin,

Bécourt, & Couarraze, 2000; Mehdizadeh, Toliate, Rouini, Abashzadeh, & Dorkoosh,

2004).

CHAPTER 32

FORMULATION OF EPIGALLOCATECHIN-3-GALLATE HYDROGEL AND EVALUATION OF ITS INTRADERMAL DELIVERY WITH THE AID OF MICRONEEDLES

Abstract

Epigallocatechin-3-gallate (EGCG) is the physiologically most active and abundant flavanol, accounting for 50-80% of green tea catechins. It is an anti- inflammatory, anti-oxidant, chemopreventive and skin photoprotective agent. However, light sensitivity and low permeability of EGCG across the stratum corneum due to its high molecular weight as well as strong binding to the lipid bilayers in the skin makes it difficult to be used as a key ingredient in cosmetic products. The present study aimed to formulate a photostable hydrogel of EGCG with good rheological properties for topical application and investigate the effect of skin microporation using maltose microneedles on its permeation through dermatomed porcine skin.

Effect of L-glutathione on photodegradation of EGCG was investigated by exposing samples to ultraviolet irradiation for 1 h using a solar simulator. Hydrogels with varying concentrations of Carbopol 980 (0.5 - 2% w/v) as gelling agent were prepared

2 Puri, A., Nguyen, H. X., & Banga, A. K. (2016). Microneedle-mediated intradermal delivery of epigallocatechin-3-gallate. International Journal of Cosmetic Science, 38(5), 512–523. http://doi.org/10.1111/ics.12320

43 44

and their rheological properties were evaluated using a rheometer. Skin microporation was confirmed by assessing the skin resistance, transepidermal water loss, and calcein imaging of the microchannels created by the microneedles. Permeation of EGCG from aqueous solution as well as the rheologically optimized hydrogel through the dermatomed porcine skin (untreated and microneedle treated) was evaluated using static vertical Franz diffusion cells.

L-glutathione acting as a co-antioxidant and photostabilizer, significantly

(p<0.05) reduced the degradation of EGCG from 21.53 ± 2.78 to 1.0 ± 0.68% after 1 h of ultraviolet irradiation. Rotational and oscillatory rheological tests indicated that the hydrogel containing 1.5% Carbopol 980 is acceptable for topical application in terms of flow behavior, elasticity, spreadability, structural stability, and thixotropy. Microneedle treated skin showed significant enhancement (p<0.05) in the delivery of EGCG to viable epidermis and dermis from the aqueous solution (38.67 ± 2.96 μg/cm2) as well as hydrogel (24.60 ± 2.62 μg/cm2) in comparison to the untreated skin (24.16 ± 2.11 and

15.62 ± 0.24 μg/cm2 for aqueous solution and hydrogel, respectively). In conclusion, addition of glutathione in EGCG formulations significantly reduced its photodegradation.

Skin microporation with maltose microneedles facilitated penetration of EGCG across the stratum corneum into the deeper skin layers - viable epidermis and dermis.

Introduction

Green tea, obtained as an extract from the leaves of Camellia sinensis, is one of the most popular beverage consumed worldwide (Kondo, Park, Watanabe, Yamamoto, &

Akashi, 2004; Min & Kwon, 2014; B. N. Singh, Shankar, & Srivastava, 2011; C. S.

Yang, Lambert, Ju, Lu, & Sang, 2007). Polyphenolic compounds known as catechins,

account for 30% dry weight of green tea. Of all catechins, EGCG is the most abundant, constituting about more than 50% polyphenolic fraction (Katiyar, Afaq, Azizuddin, &

Mukhtar, 2001; Kondo et al., 2004; Kuroda & Hara, 1999; Nagle, Ferreira, & Zhou,

2006; B. N. Singh et al., 2011; Soriani, Rice-Evans, & Tyrrell, 1998). It is also pharmacologically the most significant and potent green tea catechin and has been reported to have antioxidant, anti-inflammatory, and anticancer properties (Batchelder,

Calder, Thomas, & Heard, 2004; Lecumberri, Dupertuis, Miralbell, & Pichard, 2013;

Nagle et al., 2006).

Topical application of EGCG has been reported to protect the skin against ultraviolet radiations and prevent photo aging due to inhibitory effect on apoptosis of epidermal keratinocytes (Kondo et al., 2004), attenuation of UV-B induced hydrogen peroxide production that further inhibits the phosphorylation of mitogen-activated protein kinase signaling pathways (Katiyar et al., 2001), reduction in expression and activity of matrix metalloproteinases preventing collagen degradation (Scalia, Trotta, & Bianchi,

2014), production of antioxidant enzymes and depletion of oxidative enzymes (Tanuja

Yadav, Mishra, Das, Aggarwal, & Rani, 2015). Topical use of EGCG has also been demonstrated to reduce the number of skin tumors (malignant as well as non-malignant) in animal models (Batchelder et al., 2004; Hsu, 2005; Kuroda & Hara, 1999; C. S. Yang et al., 2007). Moreover, EGCG has also been found to have exceptionally high preventive effect against photo carcinogenesis in terms of tumor incidence, number and size when applied topically (Katiyar, 2011). It inhibits inflammation (C.-C. Lu & Yen, 2015; Scalia et al., 2014) and immunosuppression caused by carcinogenic agents. Mechanistically, this occurs due to induction of interleukin- 12, an immunoregulatory cytokine (Katiyar,

2011), apoptosis or cell growth arrest by activation of killer caspases, suppression of nuclear factor kappa Beta activation and overall, by altered expression of cell-cycle regulating proteins by EGCG (B. N. Singh et al., 2011). Besides photoprotective and chemopreventive effects, topical EGCG has also been explored for other indications such as vitiligo (Zhu, Wang, Lin, Li, & Xu, 2014), rosacea (Syed, 2007), antibacterial infections (A. Sharma, Gupta, Sarethy, Dang, & Gabrani, 2012), and telangiectasias

(Domingo et al., 2010).

Many topical green tea extract based skin care cosmetic products are available in the market such as face creams, lotions, soaps, moisturizers, and hydrogel masks

(Yoshino, Mitoma, Tsuruta, Todo, & Sugibayashi, 2014). VEREGEN® is a 15% green tea ointment, approved by the USFDA for the treatment of external perianal and genital warts (Yoshino et al., 2014). However, topical formulations or products of EGCG, the most active and potent catechin in green tea, as the active ingredient are presently not available in the market.

Many studies in literature report low skin permeation of EGCG (soluble in water, log P of 3.1, unstable in light) due to the presence of gallate group that results in its high molecular weight (458.40 Da) and also increases its binding to the lipid bilayers of the stratum corneum (Fang et al., 2007; Fang, Hwang, Huang, & Fang, 2006; Scalia et al.,

2014; Yoshino et al., 2014). An ideal antioxidant must, however, diffuse across the outermost barrier of skin i.e. stratum corneum and reach the viable epidermis and dermis where the ultraviolet radiations penetrate and have damaging effects (Fang et al., 2006;

Scalia et al., 2014; Yoshino et al., 2014). An in vivo study performed in human volunteers to investigate the percutaneous penetration of EGCG from o/w emulsion and

hydrogel formulations showed diffusion of EGCG into the stratum corneum (Scalia et al.,

2014). Various strategies have been investigated to improve the penetration of EGCG across stratum corneum such as emulsification (Yoshino et al., 2014), use of terpenes as penetration enhancers (Fang et al., 2007), and by encapsulation in liposomes (Fang et al.,

2006).

Microneedles are micron-sized structures designed to perforate the stratum corneum and enable delivery of drugs into the epidermis or dermis (Pattani et al., 2012).

Microporation using microneedles is a minimally invasive technique that creates aqueous microchannels in the upper layers of skin, enabling drugs to bypass the stratum corneum and reach the deeper layers (Donnelly et al., 2008). Microneedles range from 25 – 2000

µm in length and are a patient-compliant means of delivering therapeutic agents, both micro and macromolecules, intra- as well as transdermally (Tuan-Mahmood et al., 2013).

Different kinds of microneedles have been previously demonstrated to enhance the cutaneous delivery of 5-aminolevulinic acid (Donnelly et al., 2008), botulinum toxin A

(Torrisi et al., 2013), DNA vaccine (N. W. Kim et al., 2014), recombinant HIV-1

CN54gp140 (Pattani et al., 2012), nanoparticles (Donnelly et al., 2010).

The overall objective of this study was to formulate a photostable hydrogel of

EGCG and assess its delivery through dermatomed porcine ear skin using maltose microneedles. EGCG has been reported to degrade markedly in presence of light

(Bianchi, Marchetti, & Scalia, 2011; Scalia, Marchetti, & Bianchi, 2013). However, as photostability is an essential requirement for an effective topical anti-oxidant, one of our objectives was to reduce the degradation of EGCG in light and formulate a photostable hydrogel. Further, since evaluation of the rheological properties is essential to formulate

a topical product with good elasticity, spreadability, structure stability, adhesiveness and thixotropic behavior and is also a determinant of diffusion and release of the drug from the formulation, various rheological tests were conducted while formulating the hydrogel.

Skin was treated with microneedles to create microchannels before the hydrogel was applied to assess the permeation and retention of EGCG in skin. To our knowledge, this was the first study to be published, that reported the topical delivery of EGCG through skin, microporated using maltose microneedles and provided a detailed picture of the distribution of EGCG in the different skin layers (Puri et al., 2016).

Materials

EGCG and L-glutathione (reduced) were purchased from Medchem Express LLC

(Princeton, NJ, USA) and Sigma – Aldrich (St. Louis, MO, USA), respectively. Citric acid, disodium hydrogen phosphate and triethanolamine were obtained from Fisher

Scientific (Fair Lawn, NJ, USA). Carbopol 980NF was procured from Lubrizol

Corporation (Wickliffe, Ohio, USA). Methanol and ethanol were obtained from

Pharmco-aaper (Brookfield, CT, USA) and Fluoresoft® (0.35%) from Holles Laboratories

Inc. (Cohasset, MA, USA). A local slaughter house (Atlanta, GA, USA) provided the porcine ear skin. Maltose microneedles (array of 27 identical needles, 500 µm long) were purchased from Elegaphy Inc. (Tokyo, Japan).

Methods

Photodegradation studies

Effect of ultraviolet irradiation on 1% (w/v) solution of EGCG in citrate phosphate buffer (CPB) prepared using disodium hydrogen phosphate (Na2HPO4·7H2O,

126 mM), citric acid (49 mM) and pH adjusted to 5.5 with o-phosphoric acid was studied.

Photodegradation was evaluated with and without addition of 0.5% w/v L-glutathione to the EGCG solution using solar simulator (Newport, Oriel Instruments, Stratford, CT,

USA) equipped with a xenon–ozone-free 150-Watt lamp. Solutions (50 µL) were placed in wells (surface area, 0.5 cm2) and exposed to solar simulator for 1 h at 150 W. After exposure, samples were diluted 10 times with CPB, vortexed and analyzed using high performance liquid chromatography (HPLC) method. Results were reported as mean ±

SD (n = 3).

Formulation of hydrogels

Hydrogels were prepared using different concentrations of Carbopol 980 NF as gelling agent. Carbopol, in varying concentrations (0.5, 1, 1.5, 2% w/v) was added to the solution of EGCG (1% w/v) and glutathione (0.5% w/v) in CPB. The resulting mixture was slowly stirred overnight at room temperature on a magnetic stirrer (VWR®, Radnor,

PA, USA). After 24 h, the mixture was neutralized with triethanolamine to form a transparent gel matrix with pH of about 5.5 (Liu et al., 2008). The pH was measured using a Denver Instrument™ Model 215 pH meter (Fisher Scientific, NJ, USA).

Triethanolamine was preferred over other bases such as sodium hydroxide for neutralization as it forms gels with better transparency (Wolf, Dungan, McCarthy, Lim,

& Phillips, 2014). The symbols used for the different hydrogel formulations have been mentioned in Table 1.

Table 1. Flow properties of hydrogels as calculated by Herschel Bulkley model

Formulation Carbopol Yield Flow Flow symbols concentration stress consistency index (% w/v) (Pa) index (Pa.sn ) (n) F0.5 0.5 8.05 2.71 0.42 F1 1 32.92 7.98 0.40 F1.5 1.5 39.21 56.29 0.23 F2 2 192.46 335.24 0.12

Rheological evaluation

Rheological analysis was performed using a rheometer (Rheoplus/32 V3.62,

Anton Paar Germany GmbH, Germany). Parallel-plate spindle (PP 25/S) with the diameter of 24.987 mm was used and a gap of 100 µm was maintained between the plates. Rotational and oscillatory tests were then performed on the hydrogels at a temperature of 32 °C to assess the flow behavior and structural stability of the gel formulations.

Flow curves. Gels were sheared at an increasing rate of 0 to 100 s-1 and rheograms of viscosity versus shear rate were plotted. Shear stress values were also recorded. The data was fitted by the Herschel–Bulkley model (Kelessidis, Poulakakis, & Chatzistamou,

2011):

n τ = τy + Kγ

Where, τ represents shear stress (Pa), τy the yield stress (Pa), K the flow consistency index (Pa.sn), γ the shear rate (up to 100 s-1) and n the ﬂow behavior index

(dimensionless). The values of yield stress, flow consistency index and flow behavior index were recorded for all the hydrogels.

Amplitude sweep. The samples were exposed to an increasing strain of 1–100% at a constant angular frequency of 10 rad/s. The values of the oscillatory shear responses, storage modulus (G’) and loss modulus (G”) were plotted in logarithmic scale against the increasing strain (log scale). This test enabled determination of the linear viscoelastic region that was subsequently used to choose a strain value to be applied for the frequency sweep and thixotropy oscillation tests.

Frequency sweep. Hydrogels were exposed to an increasing angular frequency

(0.1-100 rad/s) at a constant strain (1% for F0.5 and 5% for F1, F1.5 and F2) as determined from the linear viscoelastic region of amplitude sweep curves. The values of storage modulus (G’) and loss modulus (G”) were plotted against the increasing frequency using log-log plot.

Thixotropy shear and oscillation tests. In order to evaluate the stability and structure-recovery ability of the gel after deformation at higher shear rates, the hydrogels were subjected to varying shear rates for different time intervals, referred to as the thixotropy shear test. Shear rate of 0.25 s-1 was applied for 25 s (pre-shear interval) followed by a rest shear interval of 20 s at the same rate. Further, a high shear rate of

3000 s-1 was applied for 15 s (high-shear interval) again followed by the rest shear interval of 65 s at 0.25 s-1. The changes in viscosity of the gels with time upon application of the above parameters were plotted and structure recovery ratios were recorded.

An oscillatory thixotropy test was used to evaluate the structure breakdown and rebuilding behavior of the hydrogel formulations upon application of high shear. Setting a constant strain (1% for F0.5 and 5% for F1, F1.5 and F2) as per the linear viscoelastic region of respective amplitude sweep curves, the gels were exposed to cyclic changes in

shear rate; 10 s-1 for 100 s followed by 3000 s-1 for 15 s and finally at 10 s-1 for 100 s.

The structure-recovery ratio was recorded for all the gel formulations.

Preparation of skin samples

Full thickness porcine ear skin, obtained from a local slaughter house, was cleaned thoroughly using DI water and stored at – 80 °C till further use. For the experimental study, skin was allowed to thaw at room temperature, washed with DI water and hair were cut carefully with scissors. It was then dermatomed using Dermatom 75 mm (Nouvag AG, Switzerland) to obtain skin pieces, approximately 0.6 mm thick.

Thickness was measured using a thickness gauge (Cedarhurst, NY, USA). The dermatomed skin was then cut into appropriate sizes (2 x 2 cm2) for mounting on the vertical Franz diffusion cells.

Microneedle insertion and visualization of microchannels

An array of 500 µm long and sharp tipped solid maltose microneedles was used for treatment in the skin permeation studies. Each array comprised of 3 straight parallel rows of 27 identical microneedles (Gujjar & Banga, 2014; Nguyen & Banga, 2015). Pore uniformity, transepidermal water loss and skin resistance were measured in order to confirm successful microporation of skin and characterize the microchannels before using the microneedles for in vitro permeation studies.

Pore uniformity. Microneedles were inserted for 1 min in the skin and microchannels were visualized by calcein imaging and pore uniformity was determined using Fluoropore image analysis tool. Fluoresoft® solution (0.35%) was applied on the microporated skin for 1 min after which the dye was wiped off with alcohol swabs and kimwipes. The protocol described previously by us was followed (Kolli & Banga, 2008).

A fluorescent image was then taken two-dimensionally that showed the distribution of fluorescent intensity within as well as around each pore. This was further converted into pore permeability index which represented the calcein flux for each channel.

Transepidermal water loss measurement. The objective of transepidermal water loss measurement was to assess the effect of microneedle treatment on the barrier integrity of the porcine ear skin as microneedles disrupt the stratum corneum and increase the transepidermal water loss in comparison to untreated skin (Kolli & Banga, 2008;

Nguyen & Banga, 2015). Transepidermal water loss was thus measured before and after microneedle treatment, non-invasively using a VapoMeter (Delfin Technologies Ltd,

Kuopio, Finland) in the ambient conditions of humidity and temperature after mounting the skin on vertical Franz diffusion cells (PermeGear, Hellertown, PA, USA) containing

CPB in the receptor compartment. The probe was placed still on the skin for 10 s and a stable value was displayed on the screen (Nguyen & Banga, 2015). The transepidermal water loss values obtained pre- and post- microneedle treatment were recorded and compared (n=4).

Skin resistance evaluation. Skin resistance was measured before and after microneedle treatment using silver/silver chloride electrodes, Agilent 33220A, 20 MHz

Function/arbitrary waveform generator and 34410A 6 ½ digit multimeter (Agilent

Technologies, CA, USA). Skin was mounted on the vertical Franz diffusion cells and

CPB was added in the receptor as well as in the donor. Silver chloride electrode was placed in the donor and silver in the receptor. Load resistor (RL) was connected in series with the skin, and the drop in voltage across the complete circuit (VO) and across the skin

(VS) was recorded as displayed on the multimeter. Skin resistance (RS) was calculated using the formula (Murthy, Sen, Zhao, & Hui, 2003):

RS = VS RL / (VO – VS)

Where, RL and VO were 100 kΩ and 100 mV, respectively.

Also, measurement of skin resistance of the intact skin (before microneedle treatment) enabled us to assess the integrity of skin barrier and select suitable skin samples for the permeation study. Skin pieces with resistance lower than 10 kΩ were not included in the study.

In vitro skin permeation studies

Permeation of EGCG (n = 4), through dermatomed porcine ear skin, with and without microneedle treatment was studied in vitro using static vertical Franz diffusion cells that provided an effective diffusion area of 0.64 cm2. Temperature of the receptor compartment was maintained at 37 °C using a water bath and an in-built water circulation jacket around the receiver cells. This enabled maintaining the temperature of the skin surface at 32 °C throughout the study. The cells were washed prior to the experiment and filled with 5 mL CPB. Skin was mounted on the Franz cells with the dermis facing the receptor and stratum corneum towards the donor. For the microporation study, microneedles were inserted and transepidermal water loss as well as skin resistance were evaluated after mounting the skin as described earlier. Formulations applied in the donor chamber included 100 µL of EGCG solution in CPB (10 mg/mL) and 100 mg of 1%

EGCG (w/v) rheologically optimized hydrogel. The donor chambers were covered with parafilm to avoid any chances of oxidation of EGCG during the study (Fang et al., 2006;

Scalia et al., 2014). The receptor buffer (300 µL) was withdrawn at pre-determined time

points and replaced immediately with 300 µL fresh buffer solution. Samples were analyzed using HPLC.

Extraction recovery study

The objective of this study was to determine the efficiency of the procedure to be employed for extraction of EGCG from dermatomed porcine ear skin after separation of stratum corneum as achieving 100% drug extraction is not practically feasible due to chemical interactions between the drug and skin components. Dermatomed porcine skin was tape stripped twenty times using D-Squame Stripping DiscsD (Dallas, TX, USA) to separate the stratum corneum from epidermis and dermis. After removal of stratum corneum, epidermis and dermis were cut into small pieces. Skin pieces were weighed and

50 mg of skin was placed in separate vials. EGCG standard solutions (100, 200, 500,

1000 and 2000 µg/mL) were prepared in ethanol and 50 µL of each of the solutions was added into respective vials containing the weighed amount of skin (n = 3). The vials were first centrifuged at 14000 rpm for 15 min at room temperature so as to ensure that all the skin pieces were at the bottom of the vial and in complete contact with ethanol to maximize drug absorption into skin layers and then kept overnight in an incubator at

37° C. Ethanol (500 µL) was then added to each vial and vortexed for 30 s to remove any drug present on the skin surface or adhering to the walls of the vials. Ethanol solutions were analyzed by HPLC to calculate the amount of EGCG actually taken up by the skin.

The skin pieces were then removed and placed individually from each vial in a 6 well plate (Becton Dickinson, Franklin Lakes, NJ, USA) and the following extraction protocol was performed. Methanol (2 mL) was used as the extraction solvent and added to each well. The plates were kept on the shaker (New Brunswick Scientific Co. Inc., Edison, NJ,

USA) at 150 rpm for 4 h. Samples were then filtered through 0.22 µm syringe filters

(Celltreat Scientific Products, Shirley, MA, USA) and analyzed using HPLC.

Extraction of EGCG from skin

After the 24 h permeation study, donor formulations were removed first with two dry Q tips and the skin was then cleaned properly with two CPB soaked Q tips and kimwipes. In order to measure the amount of EGCG retained in the skin layers, stratum corneum was separated from the underlying epidermis and dermis using tape stripping.

Adhesive tapes, D101 - D-Squame Stripping DiscsD were applied onto the permeation area of skin, one by one, pressed five times manually with a finger, removed quickly with forceps and collected in 6-well plates. The first two tapes were discarded to prevent overestimation of EGCG in stratum corneum due to any drug remaining on the skin surface. Subsequent tapes 1-5, 6-10, 11-15 and 16-20 were collected separately. The remaining skin (viable epidermis and dermis) was cut into small pieces using a surgical scissors and placed in 6-well plate. Further, the extraction protocol as described in the recovery study was followed. Similar procedure was performed using blank dermatomed porcine skin (n=4) in order to investigate any background noise as well as interference at the retention time of EGCG due to components of skin.

Quantitative analysis

Reverse phase HPLC based on UV detection was used for quantitative estimation of EGCG. Analysis was carried out using Waters Alliance 2695 separation module

(Milford, MA, USA) coupled with a 2996 photodiode array detector. Gradient elution was performed on Waters XbridgeTM C18, 50 x 4.6 mm, 3.5 μm (Waters Corporation,

Milford, MA, USA) column at a flow rate of 1.0 mL/min and column temperature of 35

°C after injecting 30 µL of sample. The chromatographic conditions were: methanol

(phase A) and 0.1% v/v trifluoroacetic acid in DI water (phase B) from 5 to 45% A in 6 min, 45 to 80% A (6 to 6.5 min), 80% to 5% A (6.5 to 8.5 min) and at 5% A upto 12 min for equilibration. The run time was 12 min and the retention time of EGCG was around

5.3 min. Drug standards were prepared in CPB and the detection wavelength was 274 nm. The limit of detection and quantification were 0.40 μg/mL and 1.22 μg/mL, respectively and linearity was observed in the concentration range of 0.1 - 50 µg/mL (R2

= 0.998).

Data analysis

Data was analyzed using Microsoft Excel Worksheets. Statistical analysis was carried out using Student’s t-test and p<0.05 was considered for significant difference.

Results and Discussion

Photodegradation studies

An ideal cutaneous photoprotective agent must possess adequate stability upon exposure to solar irradiation (Scalia et al., 2013). Photodecomposition of such agents would depreciate their protective activity against ultraviolet radiations (Bianchi et al.,

2011). Photochemical behavior of EGCG has been explored previously and literature reports marked degradation of EGCG when exposed to sunlight (Bianchi et al., 2011;

Scalia et al., 2013, 2014). Photodegradation of EGCG was tested in CPB (pH 5.5) which was further used as the vehicle for its hydrogel formulation as EGCG has been reported to exhibit sufficient chemical stability in aqueous solutions of low pH and it was also close to the pH of skin that would render it safe for dermatological application

(Batchelder et al., 2004; Bianchi et al., 2011; Scalia et al., 2013, 2014). The study

investigated the effect of addition of 0.5% w/v L-glutathione on the photodecomposition of EGCG.

Exposure of 1% EGCG solution to solar radiations (150 W/0.5 cm2) was observed to degrade 21.53 ± 2.78% EGCG after 1 h. Formation of EGCG dimers, the major oxidation products (Sang, Lee, Hou, Ho, & Yang, 2005) upon photolysis of EGCG have been reported earlier (Scalia et al., 2013). However, addition of 0.5% L-glutathione significantly reduced the decomposition of EGCG to 1.0 ± 0.68%, when exposed to similar conditions of solar irradiation (p<0.05).

Different strategies have been previously investigated for photostabilization of

EGCG. These include addition of co-antioxidants (Bianchi et al., 2011; Scalia et al.,

2013) and UV filters or sunscreen agents (Bianchi et al., 2011) to the EGCG formulations. Vitamin E and butylated hydroxy toluene were found to enhance photodegradation of EGCG (Bianchi et al., 2011), whereas ascorbic acid and alpha-lipoic acid successfully imparted photostabilization (Scalia et al., 2013). Also, addition of a

UV-B filter, benzophenone-4 was found to significantly stabilize EGCG in presence of light (Bianchi et al., 2011).

The present study investigated the potential of glutathione, a water soluble antioxidant, as a photostabilizing agent for EGCG and it was found to significantly reduce its photolysis. A possible explanation for the same can be attributed to the lower reactivity and redox potential of glutathione that ranges from -0.35 V to 0.04 V (Rost &

Rapoport, 1964) as compared to that of EGCG (0.43 V) (Scalia et al., 2013) which prevents oxidation of EGCG and provides photoprotection. The free sulfhydryl group present in glutathione provides nucleophilicity due to which it acts as a reducing agent

and itself transforms to the oxidized form (Touitou, Alkabes, Memoli, & Alhaique,

1996), protecting EGCG from oxidative degradation. Glutathione is a physiologic antioxidant (Pinnell, 2003) and its presence in the formulation might provide additional antioxidant and photoprotective activity as well besides stabilizing EGCG in presence of light.

Rheological evaluation of hydrogels

Aqueous polymers are commonly employed as thickening agents in topical pharmaceutical and cosmetic products to improve their rheological properties (Islam,

Rodríguez-Hornedo, Ciotti, & Ackermann, 2004; J. Y. Kim, Song, Lee, & Park, 2003).

Evaluation and knowledge of rheological parameters of topical dosage forms is essential as they tend to influence the microstructure responsible for the diffusion and release of drugs from the formulation and can also be an indirect measure of the same. Rheology can thus be used as an optimization tool for topical drug delivery of dermatological formulations in terms of physical stability, spreadability, thixotropy, time for which the formulation stays on skin and drug release rates (Islam et al., 2004).

In this study, we used a carbomer polymer, Carbopol 980 NF as the gelling agent which is an acrylic acid polymer (anionic) cross-linked with divinyl glycol (Das, Nayak,

& Nanda, 2013). It forms a gel in aqueous medium upon neutralization with organic amines or other bases such as sodium hydroxide. Carbomers are widely used in the pharmaceutical products due to their thermal stability, mucoadhesive property, lower concentrations resulting in high viscosity, good compatibility with many active ingredients and the resultant gels produced have a characteristic flow behavior, are aesthetically elegant and patient compliant as well (Islam et al., 2004; Tamburic & Craig,

1995). Carbomer grades containing no residual benzene are toxicologically preferred and thus, their use is encouraged in topical formulations (Bonacucina, Cespi, Misici-Falzi, &

Palmieri, 2008). Therefore, Carbopol 980NF, that is polymerized in cosolvent system of cyclohexane and ethyl acetate, was used as gelling agent in this study. Our aim of characterizing the flow behavior of EGCG hydrogels as a function of Carbopol concentration was to determine an optimum concentration that would produce physically stable gels with good spreadability as well as structure-recovery potential after being sheared at high shear rates.

Flow curves. The flow curves obtained for different hydrogels are shown in

Figure 2a. Data was fitted and compared using the well-known Herschel-Berkley model applied using Rheoplus software. The parameters (yield stress, consistency and flow index) obtained for all the hydrogels are mentioned in Table 1. All the gels showed pseudoplastic or shear thinning behavior as the values of flow index were observed to be less than 1 and the viscosities decreased with increasing shear rate. Pseudoplasticity is a desirable characteristic for topical semisolid formulations as it eases application over larger area utilizing lesser dosage amount (V. K. Singh et al., 2014). Further, with the increase in percentage of Carbopol, viscosity, yield stress and consistency values were found to increase. This depicted better entanglement of the polymeric chains and thus, gel strength with higher percentage of gelling agent. Minimum stress that should be applied for the material to actually begin to flow is referred to as yield stress and is an important rheological parameter for topical gels. It signifies the extent of cross-linking, strength and stability of the microgel structure, which was observed to increase with the increase in percentage of Carbopol in this study (Islam et al., 2004).

Figure 2. Comparative rheograms of hydrogels with different concentrations of Carbopol 980 corresponding to the following rheological tests: (a). Flow curves (b). Amplitude sweep (c). Frequency sweep (d). Thixotropy oscillation (e). Thixotropy shear F0.5: EGCG hydrogel with 0.5% (w/v) Carbopol; F1: EGCG hydrogel with 1% (w/v) Carbopol; F1.5: EGCG hydrogel with 1.5% (w/v) Carbopol; F2: EGCG hydrogel with 2% (w/v) Carbopol; η: viscosity (cp); G’: storage modulus (Pa); G”: loss modulus (Pa)

Amplitude sweep. Amplitude or strain sweep experiments were carried out at a constant oscillating frequency of 10 rad/s and increasing strain of 1 to 100%. The oscillatory shear responses, storage modulus (G’) and loss modulus (G”) were monitored.

The storage or elastic modulus is representative of the elastic storage of energy and is

therefore, a measure of rigidity of the material. The viscous or loss modulus, on the other hand, is a measure of dissipation of viscous energy and is large for predominantly viscous samples (Bonacucina, Martelli, & Palmieri, 2004; Tamburic & Craig, 1995). As shown in

Figure 2b, both storage and loss modulus curves were observed for all the formulations, confirming their viscoelastic nature. Moreover, at each concentration of Carbopol, storage modulus was observed to be greater than the loss modulus, indicating the samples being more elastic than viscous which is a characteristic property of an ideal gel (Puri et al., 2016). It was also observed that an increase in concentration of Carbopol from 0.5 to

2% resulted in an increase in G’ and G”. Thus, the gels F1.5 and F2 were better structured, more elastic and viscous as compared to F0.5 and F1. A cross over point (G’ =

G” = 2.42 Pa) at strain of 90% was observed for F0.5 that showed its structural breakdown. However, no structural breakdown was observed for F1, F1.5 and F2.

Amplitude sweep curves were also used to determine the linear viscoelastic region for all the formulations so as to obtain the strain values at which the other investigative tests could be conducted without disruption of the gel structure. Strain of 1% for F0.5 and 5% for F1, F1.5 and F2 was obtained beyond which changes in the structure of the gel were evident as depicted in Figure 2b. Polymers used in the formulations must also possess good adhesiveness in order to be able to make contact with the surface and spread.

Adhesiveness is a combination of liquid-like properties that enables formation of good molecular contact when pressure is applied as well as solid-like properties that provide resistance to any applied stress after the bond between polymer and surface is formed.

Elastic modulus of less than 105 Pa is an indicator of good adhesiveness of a formulation as it indicates ability to dissipate energy due to viscous contributions as well as to deform

to make a good contact with the surface (Puri et al., 2016). In our study, all the gels had an elastic modulus of less than 1000 Pa, thus fulfilling the criterion for good adhesion to skin surface.

Frequency sweep. These tests are useful to investigate the variations in responses of the polymeric gels with the changes in applied frequency and are mostly applicable for polymeric gel networks with imperfections as these gels show variable performance with changes in frequency (Puri et al., 2016). As depicted in Figure 2c, the measured storage and loss moduli were observed as almost constant for all the formulations, independent of the increase in frequency from 0.1 to 100 rad/s. This ensured formation of gels with no imperfections in the network. Moreover, in concordance with the amplitude sweep tests, the rheograms clearly showed an increase in G’ and G” values with increase in concentration of Carbopol (0.5 - 2%). Thus, it was again confirmed that the gels with 1.5 and 2% Carbopol were better structured, more elastic and viscous in comparison to 0.5% and 1% Carbopol gels. However, the highest elasticity and viscosity was observed for F2.

Thixotropy shear and oscillation tests. Thixotropic property is an important as well as desirable feature for topical products. It refers to the ability of the formulations to break down their structure upon application of high shear in order to flow smoothly across the skin surface, followed by rebuilding and recovering the structure once the strain is removed. This enables uniform application as well as enhances the formulation’s contact time with skin, further providing prolonged action of the drug incorporated in it at the topical site (C. H. Lee, Moturi, & Lee, 2009). Thixotropy shear test was performed and change in viscosity of the samples was monitored with time on being subjected to cyclic changes in applied shear rate (0.25 s-1 to 3000 s-1 and again 0.25 s-1) as shown in

Figure 2e. The viscosities observed at shear rate of 0.25 s-1 were constant with time for all the gels followed by a significant reduction when sheared at 1000 s-1. Further, the structures of F1, F1.5 and F2 formulations recovered when the shear rate was reduced back to 0.25 s-1 with the structure-recovery ratios of 85.27 ± 0.54, 93.93 ± 1.78, and

95.94 ± 1.68%, respectively. However, F0.5 showed a significant structure breakdown with the recovery ratio of 50.94 ± 15.38%. Also, structural recovery of F1.5 and F2 was found to be significantly (p< 0.05) higher than F0.5 and F1 gels but insignificant difference was observed between them (p>0.05).

In the thixotropy oscillation test, the hydrogels were exposed to varying oscillatory shear rates of 10 s-1 followed by 3000 s-1 and finally 10 s-1 at constant strain

(1% for F0.5 and 5% for F1, 1.5 and F2). The changes in the storage and loss moduli were recorded and have been illustrated in Figure 2d. Complete structural breakdown was observed for F0.5 while the structure-recovery ratios for F1, F1.5 and F2 were

100.52 ± 0.44, 108.30 ± 1.91 and 111.15 ± 1.62%, respectively. Statistically insignificant difference (p>0.05) was observed between F1.5 and F2. However, structural recovery of

F1.5 and F2 gels was found to be markedly (p<0.05) higher than F0.5 and F1.

According to the amplitude and frequency sweep tests, F2 exhibited the highest elasticity and gel strength in comparison to F0.5, F1 and F1.5. However, thixotropy shear and oscillation test provided a better picture and depicted insignificant difference between F1.5 and F2 formulations in terms of thixotropic behavior and structure-recovery ability after being sheared at high rate. Thus, overall, due to being well-structured, possessing high elasticity, physical stability, better spreadability, and structure recovery ability after breakdown, both F1.5 and F2 were found to be rheologically acceptable for

application on skin. However, according to the inactive ingredients guide provided by the

US FDA, 1.75% Carbomer polymer has been incorporated in the approved topical gel formulations. Hence, F1.5 was selected for the in vitro permeation studies (FDA, 2013).

Characterization of microchannels

Pore uniformity. After insertion of maltose microneedles, calcein was applied and due to its hydrophilic nature, it was taken up by the aqueous microchannels. The fluorescence hence produced was captured by the fluorescence microscope as shown in

Figure 3 a,b.

Figure 3. (a). Fluorescent image of microneedle-treated skin with (b). Pore permeability index values (c). Histogram generated using Fluoropore software

Fluoropore software was used to analyze the fluorescent images and it provided the pore permeability index value of 12.7 ± 6.14 for the 73 micropores created by the microneedles on the basis of the fluorescent intensity in and around each pore. A histogram was also generated depicting relatively uniform distribution of pores (Figure

3c). The results obtained were similar to those reported previously by us for pores created using maltose microneedles (Gujjar & Banga, 2014).

Transepidermal water loss measurement. As shown in Figure 4a, an increase in transepidermal water loss was observed after treatment with microneedles (22.2 ± 2.24 g/m2/h) as compared to the base value of pre-treatment (8.78 ± 1.22 g/m2/h) (p<0.05).

Normal, intact or unbreached skin generally loses very small amount of water. However, the loss is more in case of compromised skin (disrupted stratum corneum) (Kolli &

Banga, 2008). Thus, increase in transepidermal water loss is a measure to assess the formation of microchannels using microneedles as it is an indicator of the barrier integrity of skin. Our results for transepidermal water loss, substantiated with the visualization of pores by calcein imaging confirmed the formation of micropores in the dermatomed porcine skin that was to be used for in vitro permeation studies.

Skin resistance. Intact skin offers resistance to microneedle insertion. Therefore, greater force is required to overcome the resistive force and porate the skin. Skin disruption by microneedles results in significant drop in electrical resistance. The resistance of dermatomed porcine skin as measured before microneedle treatment was found out to be 19.24 ± 8.98 kΩ. However, it reduced about 37 times to 0.52 ± 0.12 kΩ after microneedle treatment (p<0.05) as shown in Figure 4b, confirming the formation of microchannels and also corroborating the results of calcein imaging and transepidermal water loss measurement. Similar results have also been reported earlier in literature

(Wermeling et al., 2008; Wing, Prausnitz, & Buono, 2013).

Figure 4. (a). Transepidermal water loss and (b). skin resistance values, pre- and post- microneedle insertion in dermatomed porcine skin * represents statistical difference between groups (Student’s t-test, p<0.05).

In vitro skin permeation studies

Permeation of EGCG from aqueous solution as well as rheologically optimized hydrogel (F1.5) through dermatomed porcine skin, with and without microneedle treatment was evaluated (n=4). The codes for different test groups have been defined below in Table 2.

Table 2. Specifications of the test groups evaluated in the in vitro permeation study

GROUP CODE EGCG FORMULATION MICRONEEDLE TREATMENT

SOL Solution No

SOL MN Solution Yes

GEL hydrogel – F1.5 No

GEL MN hydrogel – F1.5 Yes

No EGCG was detected in the receptor compartment even after 24 h in all the test groups. This indicated the inability of EGCG to permeate across the skin passively as well as through the microneedle treated skin. The results were consistent with the previous studies carried out to investigate the skin permeation of EGCG (Batchelder et al., 2004; Fang et al., 2007, 2006; Yoshino et al., 2014). Fang et al. encapsulated EGCG in a liposomal formulation consisting of anionic surfactants and ethanol and evaluated its permeation with aqueous solution as control using mice skin (Fang et al., 2006). The authors reported “zero flux” of EGCG from the aqueous solution which was explained due to strong chemical interaction of EGCG with the lipid bilayers of the skin. A slight increase in flux (0.17 ± 0.04 nmol/cm2/h) was observed with the liposomal formulation which might have been observed as mice skin is highly permeable (Banga, 2011). This could hence be one of the reasons for no permeation of EGCG observed through microneedles as porcine skin is less permeable than the mice skin used in other studies

(Fang et al., 2006). Also, it may be possible that any small amount of EGCG if permeated could have been below the limit of detection of the HPLC method (0.40 μg/mL).

Another study was performed to explore the effect of α- terpineol on skin permeation of EGCG in rats using subcutaneous microdialysis (Fang et al., 2007). No

EGCG was found to permeate through α- terpineol treated skin also that was explained due to its high molecular weight, presence of galloyl groups rendering it lipophilic and resulting in strong binding to the lipid bilayers of stratum corneum and retention in skin

(Fang et al., 2007). Yoshino et al. reported 90% binding of EGCG to the skin tissues and further explained that the gallate moiety in its structure increases the number of phenolic groups as well as its molecular weight and enhances its binding to skin (Yoshino et al.,

2014). Our observations of undetectable levels of EGCG in the receptor compartment were thus in concordance with the above mentioned studies. However, we also determined the amount of EGCG in the stratum corneum and underlying layers (viable epidermis and dermis).

Recovery studies

As shown in Figure 5, amount of EGCG added to the skin (epidermis and dermis) was plotted against the amount of EGCG extracted/recovered with the method used for extraction as explained earlier. The amounts of EGCG obtained in the epidermis and dermis after the permeation experiment were corrected using the equation obtained from the correlation plot in Figure 5. The extraction efficiency determined was low (24-29%) for the method used in the study. However, these results were similar to those reported by

Zillich et al., where 19% extraction efficiency for EGCG from porcine skin was observed using a 90:10 (% v/v) mixture of methanol–water mixture containing 0.2 g/L vitamin C as the extraction solvent. This method showed maximum recovery for all the tea catechins except for EGCG that was explained due to its oxidative instability or strong

chemical binding with the skin tissues – lipids and proteins (Zillich, Schweiggert-Weisz,

Hasenkopf, Eisner, & Kerscher, 2013) which might also be the same reason for low recovery of EGCG observed in this study as well.

30.00 y = 0.2449x - 1.2797

SD R² = 0.9919

± 25.00

g) g) μ

20.00

15.00

10.00

5.00

AVERAGE OF AMOUNT AVERAGE EGCG EXTRACTED ( SKIN EXTRACTED FROM 0.00 0.00 20.00 40.00 60.00 80.00 100.00 AMOUNT OF EGCG ADDED IN SKIN (µg)

Figure 5. Extraction recovery of EGCG from viable epidermis and dermis .

Extraction of EGCG from skin

The amount of EGCG was analyzed separately in stratum corneum and underlying layers (viable epidermis and dermis) using tape stripping technique. Twenty tape strips have been demonstrated to completely remove the stratum corneum from porcine skin in literature (Abla & Banga, 2013; Sekkat, Kalia, & Guy, 2002; Venuganti,

Sahdev, Hildreth, Guan, & Perumal, 2011). No background noise as well as interfering peaks at the retention time of EGCG were observed in the blank skin.

SOL SOL MN GEL GEL MN

± ) )

2 40 g/cm

μ 35

AMOUNT OF EGCG ( EGCG OF AMOUNT 15

5 AVERAGE 0 TAPES 1-5 TAPES 6-10 TAPES 11-15 TAPES 16-20 TOTAL

TAPE STRIPS

Figure 6. Distribution of EGCG in different layers of stratum corneum of dermatomed porcine skin using tape stripping technique SOL: EGCG solution without microneedle treatment; SOL MN: EGCG solution with microneedle treatment; GEL: EGCG hydrogel containing 1.5% (w/v) Carbopol without microneedle treatment; GEL MN: EGCG hydrogel containing 1.5% (w/v) Carbopol with microneedle treatment

The total amount of EGCG retained in the stratum corneum of the different test groups, SOL, SOL MN, GEL, and GEL MN was found to be 28.59 ± 12.40, 29.62 ±

2.37, 8.71 ± 1.07, 8.67 ± 0.21 μg/cm2, respectively (Figure 6). There was no significant difference between the total amount of EGCG deposited in the stratum corneum of the intact skin and microneedle treated groups for both the solution (SOL vs SOL MN) and gel formulation (GEL vs GEL MN) (p>0.05). Distribution of EGCG in the different layers of stratum corneum has also been illustrated in Figure 6. The tapes 1-5, 6-10, 11-

15 and 16-20 were pooled separately so as to achieve a quantifiable amount of drug. As shown in the bar graph representation, the amount of EGCG gradually reduced from the

outer layers (tapes 1-5) to the deeper layers (tapes 16-20) of the stratum corneum with statistically significant difference between the drug amount in these layers (p<0.05). This trend was observed for all the test groups. However, no significant difference was observed in the drug distribution in different layers of stratum corneum between SOL and

SOL MN as well as GEL and GEL MN groups. Therefore, microneedles did not affect the retention or distribution of EGCG in the stratum corneum.

Figure 7 shows the amount of EGCG analyzed in viable epidermis and dermis for all the test groups after being corrected using the equation obtained from the recovery study (Figure 5) accounting for the efficiency of the extraction procedure. The amount of

EGCG observed in the passive permeation test groups, SOL and GEL was 24.16 ± 2.11 and 15.62 ± 0.24 μg/cm2, respectively. Significant (p<0.05) increase in the deposition of

EGCG in epidermis and dermis was observed in the microneedle treated groups, SOL-

MN (38.67 ± 2.96 μg/cm2) as compared to SOL and GEL MN (24.60 ± 2.62 μg/cm2) in comparison to GEL group. This showed enhancement in delivery of EGCG to the deeper skin layers with the use of microneedles which are ideally the sites of action for its antioxidant, photoprotective and chemopreventive activity.

* ) ) *

2 45.00 g/cm

μ 40.00 35.00 * * 30.00

25.00 SD ± 20.00 15.00 10.00 5.00

AVERAGE AMOUNT OF EGCG ( EGCG OF AMOUNT AVERAGE 0.00 SOL SOL MN GEL GEL MN

TEST GROUPS

Figure 7. Average amount of EGCG in viable epidermis and dermis of dermatomed porcine skin after 24 h permeation SOL: EGCG solution without microneedle treatment; SOL MN: EGCG solution with microneedle treatment; GEL: EGCG hydrogel containing 1.5% (w/v) Carbopol without microneedle treatment; GEL MN: EGCG hydrogel containing 1.5% (w/v) Carbopol with microneedle treatment, * represents statistical difference from other group (Student’s t-test, p<0.05).

The enhanced skin delivery of EGCG by maltose microneedle treatment was supported by measurement of transepidermal water loss, skin resistance, and pore permeability index that depicted successful piercing of the stratum corneum by the microneedles into the underlying layers and creating channels for movement of EGCG across the barrier layer of the skin. Also, as reported in our earlier studies, maltose microneedles with 500 μm length have been observed to create about 150 – 160 μm deep microchannels (Kolli & Banga, 2008; Nguyen & Banga, 2015). Due to aqueous solubility, EGCG was able to pass through the hydrophilic microchannels but at the same time due to lipophilicity rendered by the gallate group and very high binding capacity to the skin tissues (Fang et al., 2007, 2006; Yoshino et al., 2014), it was restricted to

epidermis, dermis and stratum corneum and was not observed to reach the receptor.

Moreover, Jackson et al. reported high binding affinity of EGCG for collagen due to hydrogen bonding as well as hydrophobic bonding of the galloyl group with the collagen network (J. K. Jackson, Zhao, Wong, & Burt, 2010). This could thus, be one of the possible reasons of EGCG binding to the dermis.

Similar results have been reported earlier where encapsulation in liposomes

(Fang et al., 2006) and α-terpineol (Fang et al., 2007) enhanced the skin deposition of

EGCG mainly due to its binding to lipid bilayers of stratum corneum. However, these studies did not analyze the drug separately in stratum corneum and underlying layers. The present investigation highlighted specifically the delivery of EGCG to deeper skin layers besides stratum corneum. However, significantly lower amount of EGCG was observed in the stratum corneum, epidermis and dermis from the gel formulation when compared to the solution (GEL vs SOL and GEL MN vs SOL MN) (p < 0.05). This would have possibly been due to high viscosity and entrapment of drug particles in the three dimensional gel matrix that tends to decrease the drug release from the formulation itself.

Our results for enhancement in skin deposition of EGCG with the use of microneedles are consistent with various other studies in which microneedle devices have been reported to enhance intradermal delivery of 5-aminolevulinic acid, botulinum toxin

A, DNA vaccine, recombinant HIV-1 CN54gp140, nanoparticles (Donnelly et al., 2008,

2010; N. W. Kim et al., 2014; Pattani et al., 2012; Torrisi et al., 2013).

Conclusion

The present study showed significant enhancement in the delivery of EGCG from photostable aqueous solution as well as hydrogel to deeper skin layers (viable epidermis

and dermis) with the use of maltose microneedles. L-glutathione (0.5% w/v) was used as the photostabilizer as photostability is a pre-requisite for an ideal and efficient topical antioxidant and photoprotective agent. EGCG hydrogel formulated using 1.5% w/v

Carbopol as gelling agent was found to be rheologically stable and acceptable for topical application. In the in vitro permeation studies using dermatomed porcine skin, EGCG was not detected in the receptor compartment. However, its delivery to the deeper skin layers from the solution as well as gel formulation was significantly enhanced with the use of maltose microneedles in comparison to the untreated skin. As epidermis and dermis are the ideal sites for the pharmacological effects of EGCG, enhancement in its intradermal delivery mediated by microneedles seems to have a bright future for its use as an anti-photoaging and chemopreventive agent in cosmetic products.

CHAPTER 43

EFFECTS OF CHEMICAL AND PHYSICAL ENHANCEMENT TECHNIQUES ON TRANSDERMAL DELIVERY OF 3-FLUOROAMPHETAMINE HYDROCHLORIDE

Abstract

The present study investigated the passive transdermal delivery of 3- fluoroamphetamine hydrochloride (PAL-353) and evaluated the effects of chemical and physical enhancement techniques on its permeation through human skin. In vitro drug permeation studies through dermatomed human skin were performed using Franz diffusion cells. Passive permeation of PAL-353 from propylene glycol and phosphate buffered saline as vehicles was studied. Effect of oleic acid, maltose microneedles, ablative laser, and anodal iontophoresis on its transdermal permeation was investigated.

Infrared spectroscopy, scanning electron microscopy, calcein imaging, confocal laser microscopy, and histology studies were used to characterize the effects of chemical and physical treatments on skin integrity.

Passive permeation of PAL-353 (propylene glycol) after 24 h was found to be

1.03 ± 0.17 µg/cm2. Microneedles, oleic acid, and laser significantly increased the

3 Puri, A., Murnane, K. S., Blough, B. E., & Banga, A. K. (2017). Effects of chemical and physical enhancement techniques on transdermal delivery of 3-fluoroamphetamine hydrochloride. International Journal of Pharmaceutics, 528(1-2), 452–462. http://doi.org/10.1016/j.ijpharm.2017.06.041

76 77

permeation to 7.35 ± 4.87 µg/cm2, 38.26 ± 5.56 µg/cm2, and 523.24 ± 86.79 µg/cm2

(p<0.05), respectively. A 548-fold increase in drug permeation was observed using iontophoresis as compared to its passive permeation from phosphate buffered saline

(p<0.05). The characterization studies depicted disruption of the stratum corneum by microneedles and laser treatment. Overall, transdermal permeation of PAL-353 was significantly enhanced by the use of chemical and physical enhancement techniques.

Introduction

Psychostimulant abuse is a major public health issue, with 1.5 million Americans reporting current cocaine use and 1.6 million Americans reporting current non-medical use of other stimulants, including methamphetamine, in 2014 (Center for Behavioral

Health Statistics and Quality, 2015). Currently, there are no FDA-approved pharmacotherapies to treat cocaine-use disorder (Vocci, Acri, & Elkashef, 2005; Vocci &

Appel, 2007; Volkow & Li, 2004). Medication development efforts have focused on antagonist versus substitute agonist therapies with considerable debate as to the best approach. Antagonist therapy relies on stopping cocaine from entering the brain or from reaching its protein targets, and thereby preventing cocaine from stimulating downstream signaling through increases in brain monoamine neurotransmitters. However, substance- dependent individuals, including those with cocaine-use disorder, show unique problems with treatment compliance and antagonist approaches exacerbate such problems.

Substitute-agonist therapies mimic key aspects of the abused drug to reduce craving and withdrawal and promote abstinence. FDA approved substitute-agonist therapies include methadone, buprenorphine, varenicline, and transdermal and buccal formulations of nicotine. Strengths of this approach include the clinical success of these agents, better

compliance, reduced withdrawal and craving, and excellent efficacy profiles in preclinical models. Weaknesses include the risk of toxic drug interactions during relapse and diversion for abuse. Transdermal formulations of the substitute agonists can help in mitigating the weaknesses as they 1) provide slow and sustained drug delivery to increase safety and 2) are abuse deterrent as it is more time consuming and difficult to extract drug from a patch than a pill or tablet and the use of patches can be more easily monitored than the use of pills or tablets.

PAL-353 is a phenethylamine substrate-based dopamine/norepinephrine (DA/NE) releaser. Substitute agonists that function as substrate-based DA/NE releasers have demonstrated promising efficacy in preclinical models (non-human primates) and double-blind placebo controlled clinical trials for treatment of cocaine-use disorder or cocaine dependency (Grabowski, Shearer, Merrill, & Negus, 2004; Negus &

Henningfield, 2015). Efficacy of these drugs is assessed using a strategy, in which the same subjects (or a parallel group) that have been trained to self-administer cocaine, respond for a nondrug reinforcer (e.g., food) or cocaine, after administration of drug under investigation, thereby enabling comparisons (Banks, Blough, Fennell, Snyder, &

Negus, 2013; Banks, Blough, & Negus, 2011; Grabowski et al., 2004; Negus & Mello,

2003a, 2003b). DA/NE selective releasers such as PAL-353, phenmetrazine, phendimetrazine, and 4-benzylpiperidine have been shown to be effective. In particular, one of the most efficacious compounds was observed to be PAL-353 (Banks et al., 2011).

Because of the excellent preclinical efficacy of PAL-353 in human-relevant animal models and the therapeutic benefits of transdermal delivery of substitute agonists for

cocaine-use disorder, in the present study, we evaluated the transdermal delivery of PAL-

353.

Skin, with a surface area of 1-2 m2 seems an accessible route for local and systemic administration of drugs. However, as described in chapter 2, a major challenge encountered while developing a successful transdermal drug delivery system is to ensure the penetration of drugs across the outermost dead lipophilic layer of skin, the stratum corneum, which acts as a barrier to delivery of hydrophilic drugs (Banga, 2011; Brown,

Martin, Jones, & Akomeah, 2006; Scheuplein & Blank, 1971). Moderately lipophilic (log

P of 1-3), potent, and unionized drug molecules with a molecular weight of <500 Da and melting point <250 °C, tend to pass passively through the skin (Banga, 2011). However, passive delivery of high molecular weight as well as hydrophilic drugs and macromolecules is difficult to achieve (Banga, 2011; Khavari, 1997). Therefore, different strategies such as the use of chemical penetration enhancers and physical enhancement techniques such as microneedles, iontophoresis, sonophoresis, ablative laser, electroporation, microdermabrasion, etc. have been widely applied to investigate enhancement in the transdermal permeation of many such drugs (Banga, 1998; Kalluri et al., 2011). Effect of oleic acid as chemical enhancer, iontophoresis, skin microporation by microneedles and ablative laser were used in the present study to investigate the enhancement in transdermal delivery of PAL-353.

Chemical penetration enhancers are included in a large number of dermatological, transdermal, as well as cosmetic products to help enhance the dermal absorption of lipophilic and hydrophilic active pharmaceutical ingredients (Y. Yang, Kalluri, & Banga,

2011). Some compounds that possess penetration enhancing properties include ethanol,

propylene glycol, azones, oleic acid, DMSO, terpenes, and surfactants such as transcutol®. These act by either disrupting the stratum corneum or by increasing the partitioning of drugs into skin by solubilizing them in the membrane (A. C. Williams &

Barry, 2004). Oleic acid, a commonly used chemical enhancer, that has previously been found to enhance transdermal permeation of many drugs such as salicylic acid (Pathan &

Setty, 2009), 5-fluorouracil (Pathan & Setty, 2009), piroxicam (Saini et al., 2014), zidovudine (Mitragotri, 2000) etc., either alone or in combination with other enhancers, was selected for this study.

Iontophoresis involves application of low and constant current (<0.5 mA/cm2) that drives charged, neutral or weakly charged drug molecules into or through the skin and results in enhanced drug permeation (Pikal, 2001). Transdermal delivery of more than 50 drugs including proteins and peptides has been investigated using iontophoresis to date (Banga, 2011).

Skin microporation with the use of microneedles is an emerging technology that has been reported to improve intra- as well as transdermal delivery of micro and macromolecules such as nanoencapsulated Rhodamine B dye, amlodipine and verapamil hydrochloride, PEGylated naltrexone prodrug, mannitol, etc (Donnelly et al., 2010;

Nguyen & Banga, 2015).

Also, as explained earlier in chapter 2, the ablative Er:YAG laser emits light of

2940 nm wavelength that corresponds to the main absorption peak of water and results in ablation of the stratum corneum with minimal thermal damage. It has been reported to significantly enhance transdermal permeation of lipophilic as well as hydrophilic drugs such as indomethacin, 5-aminolevulinic acid, nalbuphine, vitamin C, 5-fluorouracil,

dextran, oligonucleotides, and DNA (W. R. Lee et al., 2006, 2001). A technology, called

P.L.E.A.S.E.®, as described earlier in the literature review has been developed for laser microporation. It comprises of a diode-pumped fractional Er:YAG laser that creates a matrix of identical micropores that are about 100–150 μm wide apart. It is a patient- friendly technology that is programmed to control the extent of disruption of the stratum corneum by varying parameters such as pulse energy, length, number, and repetition rate

(Li et al., 2013).

To our knowledge, transdermal delivery of PAL-353 had not been investigated before and this was the first study that was published reporting the same. Physical enhancement techniques are preferably applied for the permeation of hydrophilic molecules due to aqueous nature of the channels created by the microneedles and ablative laser as well as due to the requirement of the protocol for iontophoresis. Therefore, hydrochloride salt of 3-fluoroamphetamine was used for this study and has been denoted as PAL-353. The distribution coefficient (log D) of the hydrophilic salt form varies from

-1.086 to -0.498 in the pH range of 3.5-7.4. It has a molecular weight of 189.66 g/mol and pka of 9.97, as calculated using chemicalize software (MarvinSketch: version 6.2.2,

ChemAxon, Hungary, Europe).

The aim of the present study was to investigate passive delivery of PAL-353 through dermatomed human skin and study the effect of oleic acid as chemical penetration enhancer, maltose microneedles, anodal iontophoresis, and skin ablation using P.L.E.A.S.E.® technology on its transdermal permeation. Also, the effects of the enhancement strategies on skin were characterized by scanning electron microscopy

(SEM), Fourier transform infrared spectroscopy (FTIR), dye binding studies, confocal

laser scanning microscopy (CLSM), calcein imaging, histological evaluation, and skin resistance measurement.

Materials

PAL-353 was generously provided to us by Dr. Bruce Blough, Senior Research

Chemist, Research Triangle Institute (Research Triangle Park, NC, USA). Phosphate buffered saline, pH 7.4, 10X and sodium chloride were obtained from Fisher Scientific

(Fair Lawn, NJ, USA). Propylene glycol and oleic acid were purchased from Ekichem

(Joliet, IL, USA) and Croda Inc. (Edison, NJ, USA), respectively. Methanol was obtained from Pharmco-aaper (Brookfield, CT, USA) and trifluoroacetic acid from EMD Millipore

(Billerica, MA, USA). Fluoresoft® (0.35%) and methylene blue were procured from

Holles Laboratories Inc. (Cohasset, MA, USA) and Eastman Kodak Co. (Rochester, NY,

USA), respectively. Maltose microneedles (array of 27 identical needles with 500 µm length) were purchased from Elegaphy Inc. (Tokyo, Japan). Silver wire (0.5 mm diameter, 99.99%) and silver/silver chloride electrodes (2 mm x 4 mm) were obtained from Fisher Scientific (Fair Lawn, NJ, USA) and A-M systems (Sequim, WA, USA), respectively. Dermatomed human cadaver skin (abdominal, female, 63 y, ~900 μm thick) was purchased from Science Care (Phoenix, AZ, USA).

Methods

Solubility studies

Solubility of PAL-353 in propylene glycol, oleic acid, 5% w/w oleic acid in propylene glycol, and 1X phosphate buffered saline, pH 7.4 (composition: 10 mM phosphate ions, 137 mM sodium chloride, and 2.7 mM potassium chloride: PBS) containing additional 25 mM sodium chloride was investigated. An excess amount of

drug was added to 1 mL of each solvent and kept at room temperature on a shaker for 24 h at 150 rpm. After 24 h, solutions were centrifuged at 13,400 rpm for 10 min. The supernatant was diluted 1000 times with methanol, filtered through 0.22 µm syringe filters (Cell treat Scientific Products, Shirley, MA, USA), and analyzed using HPLC.

Skin preparation

Dermatomed human abdomen skin was stored in a deep freezer at -80 °C. For permeation studies, it was thawed in PBS at 37 °C and, thereafter was cut into pieces of suitable sizes.

Evaluation of skin integrity

Evaluation of the integrity of the skin samples used in permeation studies is necessary in order to assess any damage caused to skin during surgical removal, technical preparation or storage that would eventually affect drug permeation. The electrical properties of skin have been reported to be related to the physical state of stratum corneum and hence, electrical conductivity has previously been demonstrated as a tool for the assessment of the barrier integrity for epidermal membranes as well as full- thickness abdominal skin (Lawrence, 1997). Therefore, in order to select skin pieces with good barrier integrity for permeation study, resistance imposed by the skin to electrical current was measured and used as the criteria for selection. This was done using silver/silver chloride electrodes, an arbitrary waveform generator (Agilent 33220A, 20

MHz Function), and a 34410A 6 ½ digital multimeter (Agilent Technologies, CA, USA).

PBS (1X, 300 µL) was added in the donor and 5 mL in the receptor. Skin was mounted on vertical Franz diffusion cells and allowed to equilibrate for 30 min. After equilibration, silver chloride electrode and silver wire were placed in the donor and

receptor, respectively. Load resistor (RL) was coupled in series with skin, and the voltage drop (VS) across the entire circuit (VO) and skin was displayed on the multimeter. Skin resistance (RS) was calculated using the formula (Puri et al., 2016):

Rs = VS RL/ (VO - VS)

Where, VO and RL were 100 mV and 100 kΩ, respectively.

Skin samples with resistance greater than 10 kΩ were used for the permeation studies (Puri et al., 2016). The range of the resistance of the skin samples used in this study was 30-80 kΩ.

In vitro permeation study paradigm

Delivery of PAL-353 through dermatomed human skin was investigated by performing in vitro permeation studies using vertical static Franz diffusion cells

(PermeGear, Hellertown, PA, USA). The effective diffusion area provided in this set up was 0.64 cm2. The receptor compartment was maintained at 37 °C using a recirculating water bath system in order to keep the skin temperature as 32 °C. Cells were washed and filled with 5 mL of PBS to maintain sink conditions. Selected skin pieces (untreated or treated with physical enhancement techniques) were mounted on the Franz cells with the dermis side facing downwards. Drug solution was added in the donor chamber and sampling of the receptor solution (300 µL) was done with replacement with equal volume of fresh buffer solution at pre-determined time points. Analysis of the samples was done using HPLC. Results were reported as mean ± SD (n = 4) for each test group.

Specifications of the protocol followed to assess the passive permeation of PAL-

353 as well as to investigate the effect of physical enhancement delivery techniques on its permeation have been elaborated in the following sections.

Passive permeation of PAL-353. For the evaluation of passive permeation of

PAL-353, 100 μL of drug solution in propylene glycol (6.80 mg/mL, corresponding to

90% saturation solubility) was used as the donor (control). The receptor was withdrawn and analyzed at 0 h, 1 h, 2 h, 4 h, 6 h, 8 h, 22 h, and 24 h.

Effects of a chemical enhancer on the skin permeation of PAL-353. The effect of oleic acid on permeation of PAL-353 was investigated in this study. The donor solution consisted of 100 µL of PAL-353 in 5% (w/w) oleic acid in propylene glycol (6.80 mg/mL; corresponding to 82% saturation solubility). The sampling time points were similar to the passive permeation study. Also, the effects of oleic acid on the stratum corneum were evaluated by Attenuated Reflectance Fourier Transform Infrared (ATR-

FTIR) spectroscopy.

Microneedle-mediated transdermal delivery of PAL-353. Maltose microneedles were investigated as a physical enhancement technique for the transdermal permeation of

PAL-353. An array comprising of 81 sharp tipped and 500 µm long solid maltose microneedles was used in this study. The needles in the array were arranged in 3 straight parallel rows with 27 needles in each row (Nguyen & Banga, 2015). For microporation, skin samples were placed on parafilm (Parafil “M” Laboratory film, Neenah, WI, USA) in order to mimic the soft tissue that lies beneath skin and prevent breakage of needles.

Microneedles were placed perpendicular to the skin and inserted manually (using fingers) for 1 min. No applicator was used. Creation of microchannels using the manual insertion method was confirmed by SEM, dye binding, calcein imaging, and histological studies as elaborated further under characterization studies. The donor solution and sampling time points were similar to the passive permeation study.

Transdermal delivery of PAL-353 through ablative laser treated skin. Ablative laser, another physical enhancement technique, was evaluated for its effect on transdermal permeation of PAL-353. For microporation, skin samples were first placed on a flat platform (four layers of parafilm) and treated with P.L.E.A.S.E.® laser.

Treatment specifications included: fluence of 41.5 J/cm2, 1.4 W, 10% density, array size of 8, and 3 pulses/pores. The creation of micropores in skin by laser was also confirmed using SEM, dye binding studies, calcein imaging, CLSM, and histological evaluation.

The donor solution and sampling time points were similar to the passive permeation study.

Iontophoresis-mediated transdermal delivery of PAL-353. Iontophoretic delivery of PAL-353 across dermatomed human skin was also investigated. The donor chamber was filled with 500 μL of PAL-353 solution (10 mg/mL; 54% of saturation solubility) in

PBS containing 25 mM sodium chloride (n=4). The pH of the donor formulation as measured using glass electrode was around 4.0. Anodal iontophoresis was conducted where silver (anode) and silver chloride electrodes (cathode) were placed in the donor chamber and sampling port of the receptor chamber, respectively. It was ensured that there was no contact between the anode and skin in order to avoid skin damage due to high local voltage. The electrodes were then coupled in series to a source of constant current supply (Keithley 2400 Source Meter®, Keithley Instruments Inc., Cleveland, OH,

USA). A current density of 0.5 mA/cm2 was applied for 4 h. However, the total duration of the permeation study was 24 h and sampling was done at 0, 1, 2, 3, 4, 5, 6, 8, 22 and

24 h.

This study was repeated keeping all the parameters same except that no current was applied (n=4). This group served as the passive control for comparison of the iontophoretic delivery of PAL-353 across human skin. Also, in a separate study, for these two groups, after 4 h, the formulations were removed and current was stopped in the iontophoresis group. Skin resistance was then measured using the procedure explained above. Results have been presented as mean ±SE.

Calculation of lag time. Lag time was calculated as the x-intercept of the extrapolated linear portion of the permeation profiles (cumulative drug permeated/cm2 plotted against the time).

Quantitative analysis

A UV detection based RP-HPLC was used for quantitative estimation of PAL-

353. Waters Alliance 2695 separation module (Milford, MA, USA) coupled with a 2996 photodiode array detector was used. Isocratic elution was performed on Kinetex 5µ EVO

C18 100A, 250*4.6 mm column (Phenomenex, CA, USA) at a flow rate of 1.0 mL/min and column temperature of 35 °C after injecting 30 µL of sample. The chromatographic conditions were: methanol (phase A) and 0.1% v/v trifluoroacetic acid in DI water (phase

B) in the ratio of 30:70. The run time was 10 min and the retention time of PAL-353 was around 4.7 min. Drug standards were prepared in PBS and detected at wavelength of 262 nm. The precision limit of detection and quantification were 0.02 μg/mL and 0.06 μg/mL, respectively and linearity was observed in the concentration range of 0.1 - 50 µg/mL (R2

= 0.9999). No interference due to the components leaching from the skin into the receptor was observed with the drug peak, while quantifying the amount of PAL-353 in the receptor using the above mentioned HPLC method.

Data analysis

Data analysis was performed using Microsoft Excel. Student’s t-test was used for statistical analysis and p value of less than 0.05 was considered for significant difference between the test groups.

Characterization of skin exposed to a chemical enhancer, microneedles, or laser treatment

Attenuated Reflectance Fourier Transform Infrared Spectroscopy. The effect of propylene glycol and 5% (w/w) oleic acid in propylene glycol on stratum corneum was studied using ATR-FTIR, IRAffinity-1S model (Shimadzu Scientific Instruments,

Columbia, MD, USA). To separate stratum corneum from skin, dermatomed human skin was incubated with 1% (w/v) trypsin solution at 37 °C for 4 h with the stratum corneum facing upwards. The tissue was then placed on a parafilm and stratum corneum was removed using a moistened cotton-tipped applicator. The obtained transparent stratum corneum was wetted with water, blotted dry, and then finally vacuum dried before storing in a desiccator for 2 days (Y. Yang et al., 2011). Stratum corneum pieces were treated separately with propylene glycol and 5% (w/w) oleic acid in propylene glycol for 24 h at

32 °C. Untreated as well as propylene glycol and oleic acid treated stratum corneum samples were then analyzed using ATR-FTIR equipped with a deuterated L-alanine doped triglycine sulphate detector (20 scans, 0.5 cm-1 resolution). The samples were placed on the diamond crystal and ATR-FTIR spectra was scanned and recorded from

4000 to 800 cm-1 at room temperature.

Scanning electron microscopy. The Phenom™ field emission SEM system

(Nanoscience Instruments, Inc., Phoenix, AZ, USA) was used to investigate the surface morphology of the microchannels created by microneedles as well as ablative laser.

Dermatomed human skin was treated with microneedles and laser as explained earlier and fixed for SEM using 2.5% glutaraldehyde and washed with water. Excess water was removed by blotting followed by drying in oven at 50 °C. Untreated (control) and treated skin samples were then individually mounted on a metal stub with a double-sided carbon sticky tape (Ted Pella, Inc., Redding, CA, USA) and sputter coated using Denton

Vacuum Desk V sputter coater consisting of gold target (Denton Vacuum LLC,

Moorestown, NJ, USA). The sputter coater was set at 30 mA for 30 s. Samples were then examined using field emission SEM (Hitachi, S4100) equipped with a critical dimension measurement system. Accelerating voltage of the primary beam was 10 kV and secondary ion images were observed and collected at different magnifications (Kolli &

Banga, 2008). Calculation of the diameter and surface area of micropores (n=6) was done from the microscopic images using ImageJ 1.41o software (National Institutes of Health,

USA).

Dye binding. Microchannels created by microneedles as well as ablative laser in skin were confirmed by staining with methylene blue solution (1% w/v in DI water).

Pieces of dermatomed human skin with the stratum corneum facing upwards were placed on parafilm and treated with microneedles and laser as explained earlier. After microporation, the treated sites were stained with methylene blue for 1 min. The excess dye was then removed and the skin was cleaned using Kimwipes and alcohol swabs. The stained sites were visualized using a ProScope HR Digital USB Microscope (Bodelin

Technologies, OR, USA) (Nguyen & Banga, 2015).

Pore uniformity. Calcein imaging was performed to visualize the microchannels created by microneedles and ablative laser. Pore permeability index was obtained using

Fluoropore image analysis tool. Microporated dermatomed human skin was stained with

Fluoresoft® solution (0.35%) for 1 min after which the dye was removed using kimwipes and alcohol swabs. A two-dimensional fluorescent image was captured that depicted the fluorescent intensity distribution within and around each pore and the data was converted into pore permeability index that denoted the calcein flux for each channel. Also, the histogram depicting the distribution of the pore permeability index values for each pore as generated by the Fluoropore software was indicative of pore uniformity (Kolli &

Banga, 2008; Puri et al., 2016).

Confocal laser scanning microscopy. CLSM was used to investigate the depth of the microchannels created by laser. After laser treatment of dermatomed human skin,

Fluoresoft® (0.35 %, 200 μL) was applied for 1 min on the treated sites. Excessive calcein was then removed using Kimwipes and alcohol swabs. The treated skin samples were placed on a glass slide without distortion or fixation artifacts and observed using a computerized Leica SP8 confocal laser microscope (Leica microsystems, Heerbrugg,

CH-9435 Switzerland) with 10X objective at an excitation wavelength of 496 nm.

Processing of the fluorescent images was done using Leica Application Suite-Advanced

Fluorescence software. In order to study the pattern of distribution of calcein in the channels and the depth of the created microchannels, X-Z sectioning was employed

(Nguyen & Banga, 2015).

Histological evaluation. The microchannels created by microneedles and ablative laser were also characterized by histological evaluation. The microporated human skin samples were stained using 1 % (w/v) methylene blue solution followed by cleaning with alcohol swabs and Kimwipes after 1 min. They were then placed flat in Tissue-Tek®,

optical coherence tomography compound medium (Sakura Finetek USA, Inc., Torrance,

CA, USA). The block was allowed to solidify by storing at −80 °C for 1 h and then sectioned using Microm HM 505 E (Southeast Scientific, Inc., GA, USA) to obtain 10

μm thick sections. For sectioning, the block was glued tightly onto the object holder in the cryotome chamber at −20 °C using an embedding medium. The samples were then cut, and the cryosections obtained were mounted on glass slides (Globe Scientific, Inc.,

NJ, USA) and observed under a Leica DM 750 microscope (Leica microsystems, Buffalo

Grove, IL, USA).

Results

Solubility studies

Solubility of PAL-353 in propylene glycol, oleic acid, 5% w/w oleic acid in propylene glycol, and PBS containing 25 mM sodium chloride was found to be 7.55 mg/mL, 1.28 mg/mL, 8.27 mg/mL, and 18.65 mg/mL, respectively.

In vitro permeation studies

Passive permeation of PAL-353. The permeation study of PAL-353 from its solution in propylene glycol (control) across dermatomed human skin after 24 h (1.03 ±

0.17 µg/cm2) indicated low passive permeability of the compound. Lag time of 22 h was observed (Figure 8).

Effects of oleic acid on the skin permeation of PAL-353. With the use of oleic acid (5% w/w) in propylene glycol as a chemical enhancer, a 37 fold enhancement in the skin permeation of PAL-353 (38.26 ± 5.56 µg/cm2) was observed as compared to the control group (p<0.05) and the lag time was reduced to 7.28 h as shown in Figure 8. The effect of 5% w/w oleic acid in propylene glycol on the chemical structure of stratum

corneum was investigated using ATR-FTIR and the observed spectra have been presented in Figure 9.

50.00 PASSIVE PG OLEIC ACID

40.00 * )

2 30.00

g/cm µ 20.00(

10.00

Cumulative amount of drug permeated of permeated drugamount Cumulative 0.00 0 5 10 15 20 25 Time (h)

Figure 8. Effect of oleic acid on permeation of PAL-353 through dermatomed human skin * signifies statistical difference (p<0.05) between the groups, unpaired student’s t test. PG: Propylene glycol

Figure 9. The FTIR spectra of human stratum corneum: (a). Untreated (control) (b). Treated with propylene glycol (c). Treated with 5% (w/w) oleic acid in propylene glycol.

The CH2 asymmetric and symmetric stretching vibrations were observed at the wavenumbers 2919.34 cm-1 and 2849.16 cm-1, respectively, in the untreated stratum corneum (control). These vibrations originate due to the lipids and proteins in the stratum corneum (Tanojo, Junginger, & Boddé, 1997). In the propylene glycol treated stratum corneum, these vibrations were observed at 2919.81 cm-1 and 2850.89 cm-1, respectively, which were very close to the values observed in the control stratum corneum. However, these vibrations shifted to higher frequencies of 2924.01 cm-1 and 2954.91 cm-1, respectively, in case of stratum corneum treated with 5% (w/w) oleic acid in propylene glycol. This shift depicted perturbation or fluidization of the lipid bilayers by oleic acid which forms separate domains that break up the multilamellar structure of stratum corneum and result in enhanced drug permeation. Also, the C=O stretch vibration peak corresponding to the free carboxylic group of oleic acid was observed at 1710 cm-1, indicating the presence of oleic acid in the stratum corneum even after 24 h (Tanojo et al.,

1997). Also, the enhancement in the drug permeation by the use of oleic acid observed in the present study was in concordance with the 28-fold and 56-fold increment in the flux of salicylic acid and 5-flurouracil, respectively, observed through human skin membrane as a result of permeation enhancing effect of oleic acid (Goodman & Barry, 1989;

Mitragotri, 2000).

Microneedle-mediated transdermal delivery of PAL-353. The effects of skin microporation with maltose microneedles on permeation of PAL-353 were also investigated. As shown in Figure 10, pre-treatment of skin with maltose microneedles significantly increased the drug permeation to 7.35 ± 4.87 µg/cm2 (p<0.05) and reduced the lag time to 2.72 h as compared to the passive propylene glycol control.

20.00 PASSIVE PG MICRONEEDLE TREATMENT

15.00

) 2

10.00

g/cm µ ( *

5.00 Cumulative Cumulative amount permeated drug of 0.00 0 2 4 6 8 10 12 14 16 18 20 22 24

Time (h)

Figure 10. Microneedle-mediated transdermal delivery of PAL-353 through dermatomed human skin * signifies statistical difference (p<0.05) between the groups, unpaired student’s t test.

The enhanced delivery of PAL-353 across the microneedle-treated skin was supported by the results of SEM, dye-binding, calcein imaging as well as histological studies that depicted successful formation of micron-sized channels in the stratum corneum with maltose microneedles (Figure 11).

Figure 11a and 11b show the SEM images of untreated and maltose microneedle- treated dermatomed human skin, respectively. The SEM image of untreated skin depicted normal surface morphology with intact stratum corneum and no pores or channels.

However, the SEM image of the microneedle-treated skin on the other hand, confirmed the penetration of microneedles in the skin by disrupting the stratum corneum. The pores or channels created by microneedles were evident as holes on the skin surface and a single microchannel has been magnified and shown in Figure 11b. The microchannels

measured about 36.37 ± 4.99 μm at their widest opening and had a surface area of 407.11

± 32.78 μm2.

In addition to SEM, dye-binding and calcein imaging studies were conducted to investigate the ability of the microneedles in terms of length, sharpness, and density to create micron-sized channels in the skin. In the dye-binding and calcein imaging studies, blue and green color stained microspots were observed due to the diffusion of methylene blue (Figure 11c) and calcein (Figure 11d), respectively through the hydrophilic microchannels created by microneedles in skin. Furthermore, the intensity of fluorescence due to calcein, in and around each pore was quantified using fluoropore software and used to calculate the pore permeability index value, which was found to be

27.3 ± 14.53 for the 80 pores created by the microneedle array. The pore permeability index value represents the total amount of calcein present in the microchannels in relative terms and is used for accurate as well as ratiometric comparisons of the flux that is enabled by each pore (Kolli & Banga, 2008). The software also generated a histogram which showed relatively uniform distribution of pores (Figure 11e).

Furthermore, histological sectioning of the microneedle-treated skin was performed to visualize the morphology of microchannels in skin. The vertical cryosection of untreated skin as shown in Figure 11f, depicted intact stratum corneum and was used as control for comparison with microneedle-treated skin histology sections. Figure 11g shows the skin sample treated with microneedles and stained with methylene blue dye.

The stratum corneum was found to be intact as well as unstained in the portions of the skin around the microchannels. The microchannels were observed as deep indentations originating with the disrupted stratum corneum at the base and were found to be stained

with methylene blue, confirming the hydrophilicity of the microchannels. Moreover, the vertical sections of the microneedle-treated skin indicated that the triangular shape of the microchannels was consistent with the pyramidal geometry of maltose microneedles

(Nguyen & Banga, 2015).

Transdermal delivery of PAL-353 through ablative laser treated skin. As shown in

Figure 12, permeation of PAL-353 through ablative laser porated skin was found to be

523.24 ± 86.79 µg/cm2 and was about 508 folds greater than passive propylene glycol control (p<0.05). Also, lag time of 1.86 h was observed which was less than that observed in case of passive (22 h) as well as microneedle-treated skin (2.72 h).

PASSIVE PG ABLATIVE LASER

IONTOPHORESIS PASSIVE PBS 3000.00

2500.00 )

2 2000.00

g/cm 1500.00 *

µ ( 1000.00

500.00 * Cumulative Cumulative permeated amount drug of 0.00 0 5 10 15 20 25 Time (h) Figure 12. Effect of iontophoresis and ablative laser treatment on transdermal permeation of PAL-353 across dermatomed human skin * signifies statistical difference (p<0.05) between the groups, unpaired student’s t test PG: Propylene glycol, PBS: 1X Phosphate buffered saline, pH 7.4

High permeation enhancement factor due to ablative laser treatment has also been reported in other studies. Lee at al reported 260.86-fold enhancement in the in vitro permeation of vitamin C by Er:YAG laser across mice skin at fluence of 2.8 J/cm2 (Woan

Ruoh Lee, Shen, Wang, Hu, & Fang, 2003). However, the fluence applied in the present study was much higher (41.5 J/cm2).

The formation of micropores in skin by P.L.E.A.S.E.® that eventually facilitated

PAL-353 delivery was confirmed by various characterization techniques that included

SEM, methylene blue staining, calcein imaging, CLSM, and histological evaluation.

Figure 13a and 13b show the SEM images of dermatomed human skin after exposure to the ablative laser treatment. As discussed earlier, the surface morphology of untreated skin (Figure 11a) was found to be intact with no evident disruption in the arrangement of corneocytes in the stratum corneum. However, the SEM images of the laser treated skin clearly showed well-defined cylindrical shaped zones of disrupted stratum corneum barrier (Figure 13a) as well as thermal coagulation around the created pore. A single micropore created by the laser has been magnified and shown in Figure 13b.

The micropores measured about 273.50 ± 7.07 μm at their widest opening and were found to have a surface area of 41634.09 ± 1145.687 μm2.

Furthermore, the results of enhanced permeation of PAL-353 by ablative laser were corroborated by the observations of dye-binding and calcein imaging studies. These staining techniques confirmed the creation of aqueous micropores by laser that facilitated diffusion of hydrophilic dyes such as methylene blue and calcein, as shown by blue and green color stained spots in Figure 13c and 13d, respectively.

Moreover, calcein imaging studies were also used to calculate the pore permeability index of the 118 pores created by laser and was found to be 30.5 ± 14.75.

Also, the histogram generated by the fluoropore software showed a relatively narrow distribution of the pore permeability index values centered about 30.5 and was indicative of uniformly created pores (Figure 13e).

Figure 13f shows the vertical histological section of laser porated dermatomed human skin that was stained with methylene blue. Treatment of skin with laser at fluence

100

of 41.5 J/cm2, that was targeted to reach the dermis was found to successfully ablate the stratum corneum as well as the lower epidermal layers as evident in Figure 13f. Also, as shown in Figure 13f, the pore created by the laser was observed to be stained with methylene blue which confirmed the aqueous nature of these pores. Finally, the depth of the channels created by laser treatment was confirmed with CLSM. Figure 13g shows the

CLSM image in the XY plane following P.L.E.A.S.E.® poration and subsequent permeation of calcein. It shows that laser treatment led to the formation of well-defined cylindrical shaped pores.

The pore depth was estimated by the z-stack that is defined as a sequence of the confocal images captured at the same horizontal position (x, y) and at different depths (z).

It was started from the surface of skin and was conducted with a step size of 10 μm to the point where the calcein signal was no more visible as shown in Figure 14. The z stack thus, showed that pretreatment of skin by laser created about 250 μm deep microchannels in skin.

101

Figure 14. Confocal microscopy z-stack of microchannels created by P.L.E.A.S.E.® laser in human dermatomed skin

Iontophoresis-mediated transdermal delivery of PAL-353. Application of anodal iontophoresis (at 0.5 mA/cm2 for 4 h) significantly increased the permeation of PAL-353

(2159.43 ± 301.14 µg/cm2) in comparison to its passive permeation from PBS solution

(3.94 ± 0.53 µg/cm2, p<0.05) as shown in Figure 12. Lag time of around 3.5 h was observed. Iontophoretic treatment increased the cumulative drug permeation by 548 fold as compared to passive permeation. Skin resistance after 4 h of current application was

102

found to have decreased by 87.69 ± 3.03 %, that was significantly greater than in the control group (18.47 ± 6.40%, p<0.05).

Discussion

Transdermal delivery of PAL-353 was investigated to assess the possibility of utilizing this route for the systemic delivery of the novel agent for the treatment of cocaine-use disorder. This route shall aid in delivering the drug at a slow and sustained rate and would be safe and patient-compliant as well. However, based on the results of this study, passive transdermal delivery of PAL-353 was observed to be low as well as showed high lag time (22 h). This may be attributed to the drug being highly hydrophilic as well as ionized (log D varies from -1.086 to -0.498 in the pH range of 3.5-7.4), due to which it had very slow permeation across the lipophilic stratum corneum (Scheuplein &

Blank, 1971). Many studies in literature report the use of chemical and physical enhancement techniques as means to enhance the skin permeation of poorly permeable drugs (Banga, 1998; W. R. Lee et al., 2006; Mitragotri, 2000; Nguyen & Banga, 2015;

Pathan & Setty, 2009; Saini et al., 2014; Y. Yang et al., 2011) and some of them were explored in this study, as mentioned earlier. Findings of the present study were very interesting and significant enhancement in the transdermal delivery of PAL-353 was observed with the use of oleic acid as chemical enhancer in the formulation as well as by physical enhancement techniques such as maltose microneedles, ablative laser, and anodal iontophoresis as compared to its passive permeation. Anodal iontophoresis as well as skin microporation by ablative laser enhanced the skin permeation of PAL-353 manifolds (548 and 508 times, respectively), and were thus, found to be the most efficient permeation enhancing strategies for the transdermal delivery of the novel agent.

103

Oleic acid is a very commonly used chemical penetration enhancer in topical and transdermal formulation and its mechanism of penetration enhancement has been well described in literature as fluidization or delipidization of the stratum corneum. It accumulates within the lipophilic domains of stratum corneum, fluidizes it, and creates permeable ‘pores’, that eventually result in reduced resistance for permeation of polar molecules. Structurally, oleic acid consists of 18 carbon long alkyl chain and a polar head group and has been reported to be more effective for hydrophilic permeants (H. Benson,

2005; Mitragotri, 2000; Pathan & Setty, 2009; Y. Yang et al., 2011). The mechanism of the permeation enhancing effect of oleic acid has been demonstrated by various techniques such as differential scanning calorimetry (Y. Yang et al., 2011), electron spin resonance studies, FTIR, Raman spectroscopy, and X-ray diffractometry (H. Benson,

2005). Results of ATR-FTIR studies in this study, confirmed the perturbation or fluidization of the lipid bilayers by oleic acid that eventually resulted in significant enhancement in drug permeation. These observations were in concordance with earlier studies (H. Benson, 2005; Golden, McKie, & Potts, 1987; Tanojo et al., 1997) and thus, confirmed the disruption of the stratum corneum barrier by oleic acid. Moreover, combination of oleic acid and propylene glycol has been shown to have a synergistic permeation enhancing effect. Propylene glycol, acting as a cosolvent enhances the concentration of both the enhancer as well as the drug in the stratum corneum. In addition to this, oleic acid, as explained earlier, due to its lipid fluidizing effect, enhances the free volume within the stratum corneum bilayers and thus, facilitates partitioning of both, the permeant and propylene glycol (H. Benson, 2005; Mitragotri, 2000). Thus, overall, the disruption of the integrity of stratum corneum by oleic acid as revealed by the ATR-FTIR

104

studies and synergistic penetration enhancing effect of oleic acid and propylene glycol, both indicate a mechanism for our observation of significant enhancement in permeation of PAL-353 by addition of oleic acid in propylene glycol.

Maltose microneedles successfully created channels for the movement of the hydrophilic drug, PAL-353, that enhanced its delivery into and through the skin as compared to the passive permeation, wherein the stratum corneum was intact and unlike with microneedle treatment, no microchannels were present. All the characterization techniques confirmed the formation of aqueous micropores in skin with the aid of maltose microneedles and the observations were consistent with our previous studies

(Nguyen & Banga, 2015; Puri et al., 2016). The hydrophilicity of the micropores was confirmed as methylene blue and calcein are hydrophilic dyes that are not taken up by intact skin due to the outer stratum corneum barrier layer which is lipophilic in nature.

However, the skin becomes porous after microneedle treatment due to disruption of stratum corneum layer at the microporated sites and thus, allows the hydrophilic dyes to diffuse through the created aqueous channels (Kolli & Banga, 2008). Also, the depth of the microchannels has been investigated in our earlier studies using CLSM and it has been observed that 500 μm long microneedles are able to create microchannels that are

150 – 160 μm deep (Kolli & Banga, 2008; Nguyen & Banga, 2015). Microneedles, however, provided significantly less enhancement in skin permeation of PAL-353 than oleic acid (p<0.05), but shortened the lag time more than oleic acid. These observations can be attributed to the time taken by oleic acid to diffuse from the formulation, reach the stratum corneum, fluidize the lipids and then enable drug permeation unlike microneedles where the microchannels are open and facilitate penetration of PAL-353 from early time

105

points (Yerramsetty, Rachakonda, Neely, Madihally, & Gasem, 2010). However, where on one hand, oleic acid acts on the entire stratum corneum and delipidizes it, microneedles, on the other hand, porate comparatively smaller surface area of stratum corneum corresponding to the number of microneedles in the array on skin. Therefore, overall permeation of PAL-353 was significantly higher with the use of oleic acid than microneedles due to greater surface area of the disrupted stratum corneum by the former that facilitated higher drug delivery.

Further, the results of enhanced drug permeation with the P.L.E.A.S.E.® system were consistent with the earlier studies that have demonstrated enhanced delivery of diclofenac, lidocaine, and prednisone across human and porcine skin (Bachhav et al.,

2010; Bachhav, Heinrich, & Kalia, 2011; Yu et al., 2010). Ablative laser works by the mechanism of partial ablation of the stratum corneum that decreases the intrinsic barrier property of the skin and thus, enhances skin permeation of drugs that are otherwise impermeable, such as PAL-353 in this study. As described earlier, the P.L.E.A.S.E.® device consists of Er:YAG laser beam (2940 nm) that is strongly absorbed by water in the superficial layers of irradiated tissue, stratum corneum, epidermis, and dermis and its vaporization results in formation of pores and channels in skin (Bachhav, Heinrich, &

Kalia, 2013; Li et al., 2013). Moreover, the concentrated laser beam is distributed into microbeams, that fractionally ablate the skin with less damage and also the short pulse duration assures localization of heat transfer to the skin surface and not the underlying viable tissues (Bachhav et al., 2013; Li et al., 2013). Fractional ablation results in removal of only 5-15% of the skin surface and the micropores created are surrounded by healthy tissue which aids in skin recovery (Bachhav et al., 2013). Overall, P.L.E.A.S.E.®

106

creates pores that provide passage for drugs that are impermeable or poorly permeable through intact skin such as PAL-353. This was confirmed by the results of the different characterization studies. The observations of the histological studies were consistent with those reported by Bachhav et al., where the P.L.E.A.S.E.® device when used at a fluence of 45.3 J/cm2 was found to reach the epidermal-dermal junction and created deeper pores up to the dermis (Bachhav et al., 2010). CLSM observations were also consistent with those reported in literature (Bachhav et al., 2010).

The permeation of PAL-353 through laser-ablated skin was found to be significantly higher than maltose microneedle-treated as well as oleic acid enhancer group. As shown by the various characterization techniques earlier, this can be attributed to the formation of more pores (118) as well as pores with wider diameters (273.50 ±

7.07 μm) and surface areas (41634.09 ± 1145.687 μm2) by laser in contrast to fewer (80) pores with smaller diameters (36.37 ± 4.99 μm) and surface areas (407.11 ± 32.78 μm2) created by the maltose microneedles, which would facilitate less drug delivery compared to laser-ablated skin. Also, the microchannels created by laser were about 250 µm deep as compared to those created by maltose microneedles (150-160 µm) that would also have resulted in significantly higher drug levels in the receptor with the use of laser as compared to microneedles. Furthermore, the drug permeation by laser treatment was found to be significantly greater than the oleic acid enhancer group as in the latter, the enhancer delipidizes only the stratum corneum and does not create any hydrophilic deep pores or channels as in case of laser, that provide a less resistive pathway for the movement of drugs, into and through skin.

107

Iontophoresis, a physical enhancement technique, that drives charged or neutral drugs, into and through skin by application of a low constant current, works on the principle of electrorepulsion and electro osmosis. It was observed to be the most efficient technique for enhancing the transdermal delivery of PAL-353. Hydrochloride salt of 3- fluoroamphetamine was used in this study and being polar and water soluble, it was considered as a good candidate for iontophoresis. Moreover, 3-fluoroamphetamine is basic in nature with a pka of 9.97 and at a pH of 4.0 (formulation pH), it would be positively charged and thus, anodal iontophoresis was used. Application of anodal iontophoresis resulted in the highest drug permeation amongst all the investigated physical and chemical enhancement techniques. Electroosmosis always occurs from anode to cathode and is also one of the mechanisms that contributes to the iontophoretic delivery of positively charged drug molecules. Permeation of PAL-353 was observed to increase linearly after termination of current that may be attributed to the change in the electrical properties of stratum corneum as evident by a significant drop in the skin resistance compared to the control group. Changes in the electrical properties of skin further indicated perturbation/disorganization of the stratum corneum barrier. This has been reported in earlier studies, where the effect of iontophoresis on the integrity of stratum corneum has been demonstrated with the help of FTIR, differential scanning calorimetry, trans-epidermal water loss, differential thermal analysis, freeze fracture electron microscopy, and X-RAY diffraction studies. Type and concentration of ions, increased hydration, dissipation of heat, and direct electric field effects during iontophoresis have been shown to result in disorganization of the stratum corneum

(Jadoul, Bouwstra, & Préat, 1999; Jadoul, Doucet, Durand, & Préat, 1996; M. R.

108

Prausnitz, 1996). Therefore, overall, the combined effects of iontophoresis, electroosmosis, and disorganization of stratum corneum due to current application, seem to have resulted in 548-fold enhancement in the permeation of PAL-353. Moreover, our results of enhanced transdermal drug delivery by anodal iontophoresis are consistent with those reported earlier in literature. Sachdeva et al. reported enhanced transdermal delivery of terbinafine hydrochloride across hairless rat skin by application of anodal iontophoresis with different current densities (0.2, 0.3 and 0.4 mA/cm2) applied for 1 h

(Sachdeva et al., 2010). Furthermore, Singh et al. showed enhanced permeation of venlafaxine hydrochloride across porcine ear skin using anodal iontophoresis by application of current of 0.5 mA/cm2. Moreover, electrorepulsion and electroosmosis were explained as the major mechanisms for the transdermal delivery of positively charged venlafaxine ions (G. Singh, Ghosh, Kaushalkumar, & Somsekhar, 2008).

Furthermore, combination of propylene glycol and anodal iontophoresis was reported to enhance zidovudine flux by about 400-folds across mice skin (Mitragotri, 2000). Anodal iontophoresis was thus, found to be the most effective strategy for enhancing the transdermal delivery of PAL-353 through dermatomed human skin.

Conclusion

PAL-353 can be delivered transdermally with the aid of chemical and physical enhancement techniques. Skin microporation by ablative laser as well as anodal iontophoresis, acting via different mechanisms, were found to be more effective than skin microporation by maltose microneedles or incorporation of oleic acid as chemical enhancer in the formulation, for enhancing the transdermal delivery of PAL-353 across dermatomed human skin. Iontophoresis resulted in the highest cumulative drug

109

permeation, whereas ablative laser showed minimum lag time, indicating faster onset of drug action. These preliminary results are very encouraging and indicate that transdermal route holds promise for delivery of PAL-353. These treatment modes need to be explored further in clinical studies to investigate transdermal delivery of therapeutically relevant doses of PAL-353 for cocaine-use disorder.

110

CHAPTER 54

IN VITRO MICRODIALYSIS AS A TOOL FOR QUANTITATIVE ANALYSIS OF DIFFUSION RATE OF DICLOFENAC SODIUM IN HUMAN SKIN

Abstract

The purpose of this study was to employ in vitro microdialysis as a tool for quantitative analysis of rate of diffusion of diclofenac sodium in vertical and lateral direction, in intact as well as microporated dermatomed human skin. In vitro permeation studies were performed using vertical Franz diffusion cells. Two linear probes were inserted into the dermis of untreated, poly(D,L-lactide-co-glycolide) microneedle-treated and ablative laser-treated human skin such that one was in the center of the diffusion area

(central probe), and the other probe was inserted parallel, at a distance of 8 mm from the central probe (lateral probe). The depth of the probes in the skin was determined using

DermaScan. Skin samples with the probes were then mounted on Franz cells, sandwiched between donor chamber containing diclofenac sodium solution (4 mg/mL, 2 mL) and receptor compartment containing phosphate buffered saline, pH 7.4 (PBS). Probes were perfused with PBS at 2 μL/min and the dialysate as well as receptor samples were collected at pre-determined intervals. After 24 h, drug levels in the skin were analyzed.

4 Nguyen, H. X., Puri, A., Bhattaccharjee, S. A., & Banga, A. K. (2018). Qualitative and quantitative analysis of lateral diffusion of drugs in human skin. International Journal of Pharmaceutics, 544(1), 62–74. http://doi.org/10.1016/j.ijpharm.2018.04.013

110 111

DermaScan images indicated insertion of probes into the dermis at an average depth of 346.67 ± 134.28 μm. The cumulative amount of diclofenac sodium permeated through laser-treated skin (659.71 ± 107.26 μg/cm2) was significantly higher than untreated (125.95 ± 23.75 μg/cm2, p<0.01) and microneedle-treated skin (153.90 ± 30.87

μg/cm2, p<0.01). Rate of drug diffusion in the central probe observed after 24 h in the microneedle-treated skin (11.79 ± 2.54 μg/h) was significantly higher than untreated

(p<0.01) as well as laser-treated skin (p<0.05). Rate of lateral diffusion in lateral probe after 24 h in untreated group (0.69 ± 0.05 μg/h) was determined to be significantly lower than microneedle-treated (1.67 ± 0.32 μg/h, p<0.05) and laser-treated skin (1.32 ± 0.07

μg/h, p<0.01). Also, significantly higher amount of drug was observed in the epidermis and dermis of the central as well as lateral diffusion area in the laser treated skin as compared to passive and microneedle treated skin (p<0.05).

Overall, in vitro microdialysis was demonstrated as a novel and valuable tool that can be employed for quantitative investigation of rate of vertical and lateral diffusion of drugs in skin.

Introduction

Skin is one of the most readily accessible organs of human body that is used as a route for local and systemic delivery of therapeutic agents. However, as discussed earlier, drugs have to traverse the outermost, dead and horny barrier layer of skin, the stratum corneum, in order to enter the body. Thus, transdermal or topical drugs penetrate through this barrier via intercellular, transcellular or transappendageal pathway (Gee et al., 2012;

Johnson et al., 1996; Nayak et al., 2015).

112

Migration of drug molecules through skin was initially interpreted as vertical diffusion through stratum corneum, viable epidermis, dermis, and into systemic circulation (Nayak et al., 2015). However, it has been reported that drugs have a tendency to diffuse both vertically and horizontally in skin depending on their physicochemical properties and concentration gradient as shown in Figure 15 (Nayak et al., 2015).

Figure 15. Schematic representation of vertical and lateral diffusion of drug molecules across stratum corneum Not drawn to scale; adapted from (Gee et al., 2012).

The process of lateral drug diffusion occurs in conjunction with the process of vertical drug penetration into the skin (Jacobi et al., 2004; Johnson et al., 1996, 1997;

Schicksnus & Muller-Goymann, 2004). It may occur on the surface of stratum corneum, along the lipid bilayers within the stratum corneum or from one bilayer into a parallel bilayer in the multilayered assembly of stratum corneum, followed by diffusion through the adjacent lipophilic channels between the corneocytes into the deeper skin layers (Gee

113

et al., 2012). Lateral diffusion depends on the drug’s solubility in the moisture layer on the skin surface, its molecular weight, as well as its lipophilicity (Gee et al., 2012;

Johnson et al., 1996; Lampe et al., 1983). Hydrophobic and hydrophilic molecules have been reported to diffuse faster in the lateral than vertical direction in the stratum corneum because of the continuous nature of the intercellular pathway (Johnson et al., 1997;

Schwindt, Wilhelm, & Maibach, 1998). Lateral diffusion is more pronounced for drugs with molecular weight up to 300 g/mol (Simonsen et al., 2002) as was shown in a study by Gee et al., where caffeine (194.5 g/mol) and ibuprofen (206.3 g/mol) exhibited greater lateral diffusion than hydrocortisone (362.5 g/mol) (Gee et al., 2012).

Furthermore, due to the lipophilic nature of stratum corneum, actives with a high partition coefficient are more likely to undergo lateral diffusion (Simonsen et al., 2002;

Weigmann et al., 1999). Overall, the rate of lateral diffusion is reported to be considerably faster than that of transbilayer transport when compared over equivalent distances (Almeida, Vaz, & Thompson, 1992; Lieb & Stein, 1986).

However, majority of in vitro skin permeation studies using vertical Franz diffusion cells, evaluate the ability of substances to cross the stratum corneum (vertical diffusion), giving little or no consideration to lateral diffusion (Nayak et al., 2015). It can also be speculated that lateral diffusion might be a contributing factor for drug loss and low mass balance in in vitro permeation studies (Gee et al., 2012). Thus, overall, the extent of lateral diffusion of drugs in skin may vary depending on the physicochemical properties of the drug but it is an important phenomenon which must be extensively studied and taken into account while evaluating the transdermal permeation in vitro.

114

Very few techniques have, however, been employed to investigate lateral diffusion of drugs in skin till date. Fluorescence-based techniques such as fluorescence recovery after photobleaching, raster image correlation spectroscopy, and confocal laser scanning microscopy (CLSM) have been used to study lateral diffusion of substances in both stratum corneum lipid models as well as in isolated stratum corneum (Bai et al.,

2014; Mizutani et al., 2009; Park et al., 2010; van Smeden et al., 2011). Vibrational spectroscopic methods such as Fourier transform infrared microscopy (FTIR) and confocal Raman microscopy have been used to investigate the lateral diffusion of drugs such as dithranol in artificial membrane constructs (Gotter et al., 2010) and urea in human stratum corneum (Hartmann et al., 2004). Also, studies using skin biopsy have been performed to investigate horizontal permeation of lidocaine in vivo in female mixed-breed agricultural swine (Zhang et al., 2012). Furthermore, skin biopsies at varied distances from the application area were collected to evaluate and compare the lateral diffusion of butyl salicylate and salicylic acid (Simonsen et al., 2002). Tape stripping protocols have also been used for monitoring lateral diffusion of drugs in stratum corneum (Gee et al., 2012). Concentric tape stripping method was used to investigate both lateral spreading as well as penetration of drugs with different physicochemical properties such as caffeine, ibuprofen, and hydrocortisone in vivo in humans (Gee et al.,

2012). Furthermore, Schicksnus and Muller-Goymann used custom-made steel punchers of different diameters to study the extent of lateral diffusion of ibuprofen in human skin and artificial skin constructs (Schicksnus & Muller-Goymann, 2004).

Chemical enhancers such as oleic acid, DMSO, azones, etc. and physical enhancement techniques such as microneedles, iontophoresis, laser microporation,

115

electroporation, sonophoresis, thermal, and radiofrequency treatment are commonly investigated to improve the transdermal permeation flux of drugs across skin (Puri et al.,

2016; Tas et al., 2017; Y. Yang et al., 2011). The effect of these techniques on vertical diffusion of drugs has been reported in literature. However, effect of the same on lateral drug diffusion has not been studied extensively before. Schicksnus and Muller-Goymann reported significant enhancement in the lateral diffusion of ibuprofen with the use of urea and DMSO as penetration enhancers (Schicksnus & Muller-Goymann, 2004). Bai et al. reported enhancement in the lateral diffusion of Alexafluor 555-bovine serum albumin conjugate with the use of combination of microneedles and iontophoresis (Bai et al.,

2014).

The effect of physical enhancement techniques on the lateral diffusion of drugs in skin needs further exploration. This study had two specific aims. The first aim was to simultaneously investigate vertical and lateral diffusion of diclofenac sodium (model drug) using quantitative analytical method of in vitro microdialysis in intact dermatomed human skin. The second aim was to evaluate the effect of two physical enhancement techniques used for skin microporation, namely, microneedles and ablative laser, on vertical and lateral diffusion of diclofenac sodium, using the same in vitro microdialysis set up. Poly lactide glycolic acid (PLGA) microneedle array (100 needles, 450 μm long) were used for skin microporation. P.L.E.A.S.E.® technology as described before, was used for skin microporation in this study as well (Li et al., 2013; Nguyen & Banga,

2018b).

The methods that have been used for studying lateral diffusion of drugs in skin, as mentioned above (fluorescence microscopy and FTIR), are mostly qualitative in nature

116

and are applicable for drug analysis in the stratum corneum. Other techniques, such as tape stripping (for stratum corneum) and skin biopsy (for deeper skin layers) are quantitative but can be performed only towards the end of the in vitro or in vivo permeation studies. The real-time vertical or lateral diffusion profile of drugs that are intended for transdermal delivery has thus, not been investigated before. Microdialysis, an analytical technique that is based on diffusion of the molecules through a semi- permeable membrane, has been widely used for studying dermal pharmacokinetics of various drug compounds in vivo. However, it has not been used for drug analysis in an in vitro set up before. Thus, the technique of microdialysis was used in vitro for the first time for the quantitative analysis of vertical and lateral diffusion of diclofenac sodium in dermatomed human skin in this study. Diclofenac sodium, due to its small molecular weight (318.1 g/mol), was selected as a model drug. Also, its lateral diffusion has not been reported in literature before.

Materials

Diclofenac sodium was purchased from Sigma Aldrich (St. Louis, MO, USA).

Phosphate buffered saline (PBS) (10X, pH 7.4 ± 0.1) was obtained from Fisher Scientific

(Fisher BioReagent, NJ, USA). Methanol and ethanol were obtained from Pharmco-aaper

(Brookfield, CT, USA). Linear microdialysis probe (LM-10, 10 mm membrane, 30 KDa molecular weight cut off, 320 µm outer membrane diameter) was purchased from BASi®

(West Lafayette, IN, USA). Dermatomed human skin was obtained from New York Fire

Fighter (NY, USA) and stored at -80 °C for no more than two months.

117

Methods

Skin treatment with microneedles and ablative laser

Dermatomed human skin was removed from the deep freezer, and thawed in 1X phosphate buffered saline, pH 7.4 (composition: 10 mM phosphate ions, 137 mM sodium chloride, and 2.7 mM potassium chloride: PBS) at 37 °C for 1 min. It was then dried by

Kimwipes, cut into 2x2 cm2 pieces, and treated with microneedles and ablative laser for further studies. Method of microneedle insertion into skin has been described in our previous studies. PLGA microneedles were inserted into the skin tissue for 2 min with a moderate pressure using a thumb (Cheung, Han, & Das, 2014; Nguyen & Banga, 2015,

2017). In this study, PLGA microneedles were pressed manually on skin, using thumb pressure to create 100 microchannels (pore area: 1868 ± 409 µm2, pore-to-pore distance:

0.5 ± 0.0 mm). In the laser group, skin pieces were treated with fractional ablative laser

(P.L.E.A.S.E®) (Nguyen & Banga, 2018a, 2018b) using the following specifications: 3 pulses/pore, pulse duration of 175 µs, treatment rate of 200 Hz with the energy of 34.1

J/cm2 (1.2 W), array size of 8 mm and laser density of 10% to generate 120 microchannels in skin (pore area: 43351 ± 3380 µm2, pore-to-pore distance: 0.7 ± 0.2 mm). The treated skin was then characterized and used for quantitative analysis of drug diffusion in vitro.

Skin integrity and thickness assessment

Human dermatomed skin was used for in vitro permeation studies and was prepared as explained above. Further, resistance of skin was measured in order to ensure selection of skin pieces with good initial barrier integrity for the permeation study. Silver- silver chloride electrodes attached to an arbitrary waveform generator (Agilent 33220A,

118

20 MHz Function) and 34410A 6 ½ digital multimeter (Agilent Technologies, CA, USA) were used for measurement. Skin pieces were mounted on vertical Franz diffusion cells and allowed to equilibrate for 30 min after addition of 300 µL and 5 mL of PBS in the donor and receptor, respectively. After equilibration, silver wire and silver chloride electrode were placed in the receptor and donor, respectively. Load resistor (RL) was attached in series with skin, and the drop in voltage across the entire circuit (VO) and skin

(VS), as displayed on the multimeter was recorded. Skin resistance (RS) was calculated using the following equation (Murthy et al., 2003; Puri et al., 2016):

RS=VS RL/ (VO - VS)

Where, VO and RL were 100 mV and 100 kΩ, respectively.

Thickness of dermatomed human skin pieces, selected for the permeation study was measured using a thickness gauge (Cedarhurst, NY, USA).

Probe insertion and depth measurement

Linear microdialysis probes were inserted in the dermis of dermatomed human skin using a 25G introducer needle. For preliminary investigation of the depth of the inserted probe, epidermis layer of the skin was removed and the images of the dermis were taken using ProScope HR Digital USB Microscope (Bodelin Technologies, OR,

USA). Furthermore, the exact depth of the probe in the skin was measured using

DermaScan® C equipped with DermaScan C V3 1.6.5.1 software (Cortex Technology,

Denmark) (Vemulapalli, Yang, Friden, & Banga, 2008). Microdialysis probe-embedded skin was covered with a thin layer of gel (Dane-Gel (CE), Rohde Produits, Denmark) and the DermaScan probe was placed perpendicular to the skin. The inserted probe was then

119

scanned using the automatic mode and its average depth was calculated using DermaScan

C software (n=3).

Probe recovery studies

These studies were conducted to calculate the probe recovery equation before initiating the permeation studies. Solutions of different concentrations (0.5, 5, 10, 50,

100, and 500 μg/mL) of diclofenac sodium in ethanol:PBS (1:3) were prepared. The semi-permeable membrane region of the linear microdialysis probes was placed in contact with the drug solutions of different concentrations and the probe was continuously perfused with PBS at flow rate of 2 μL/min using CMA 102 microdialysis pump AB (CMA Microdialysis, Stockholm, Sweden). The dialysates were analyzed using HPLC till the steady state was achieved and the graph of actual vs recovered drug concentration was hence, plotted. The equation obtained was then used to correct the in vitro permeation data.

In vitro permeation studies

In vitro permeation of diclofenac sodium through human dermatomed cadaver skin was investigated using vertical Franz diffusion cells and microdialysis system. After measuring skin resistance and thickness as explained earlier, two linear microdialysis probes were inserted into the dermis of untreated (passive control), PLGA microneedle- treated, and ablative laser-treated human skin using 25 G introducer needle (n=3). It was ensured that the two probes were parallel to and at a distance of 8 mm from each other.

The ends of the probes were attached to the skin using super glue (Loctite Super Glue

Easy Squeeze Gel, Loctite®, Westlake, OH, USA) so as to immobilize the semi- permeable membrane part of the probe in the diffusion area. The skin was then covered

120

using container seal TM sealing tape (Laboratory Essentials, Bio-world, OH, USA), such that the exposed diffusion area was 0.64 cm2 and it included all the micropores, alongwith one of the probes (central probe) being in the center of the microporated diffusion area and the probe parallel to it (lateral probe) was under the adhesive tape as shown in Figure 16a.

Figure 16. (a). Linear microdialysis probes in dermatomed human skin (b). Skin with probes mounted on Franz diffusion cell (c). Schematic representation of the in vitro microdialysis set up

121

Skin samples with the probes were then mounted on the Franz cells, sandwiched between donor chamber containing diclofenac sodium solution prepared in ethanol: PBS

(1:3) (4 mg/mL, 2 mL) and receptor compartment containing 5 mL of PBS. Wider donor tops were used so as to ensure that the lateral probe was not clamped as shown in Figure

16b. However, at the same time, effective diffusion area was kept as 0.64 cm2 using the adhesive tape design, as described above. The receptor solution was maintained at 37 °C using a recirculating water bath system so as to keep the temperature of skin around 32°

C. Probes were perfused with PBS at 2 μL/min and the dialysate as well as receptor samples (0.3 mL) were collected at pre-determined intervals and analyzed using HPLC.

Receptor solution was replaced with fresh buffer after each sampling. The rate of vertical and lateral diffusion was defined as the amount of diclofenac sodium that diffused from skin tissues vertically into the central probe and horizontally into the lateral probe, respectively per unit of time (µg/h). Schematic of the microdialysis set up has been shown in Figure 16c.

Investigation of the barrier and adhesive property of the tape

To investigate the barrier property of the sealing tape to drug permeation, in vitro permeation of diclofenac sodium through the adhesive tape (in place of skin) sandwiched between the receptor and donor chamber was also investigated. Donor and receptor solutions (maintained at 37 °C) were same as that used in the in vitro permeation studies and the diffusion area was also 0.64 cm2. Receptor solution (300 µL) was withdrawn at pre-determined time points and was replaced with equal volume of fresh buffer solution.

Analysis of the samples was performed using HPLC.

122

Besides the barrier property, it was essential to ensure that the tape had good adhesion to skin so as to prevent sideways leakage of the drug from the effective diffusion area to the lateral probe. Calcein imaging was performed to investigate the same. Skin, without the probes, but with the tape, was mounted on the Franz cells and 2 mL of calcein solution (0.35% w/v) was applied in the donor chamber. After 24 h, calcein solution was wiped off from skin using Kim wipes and alcohol swabs and images of the diffusion area were taken using the fluorescent digital camera. Further, the adhesive tape was removed and images of the back side of the tape were taken using the fluorescent camera. All the images were viewed using the Fluoropore image analysis software.

Drug distribution in skin

After 24 h of the in vitro permeation studies, drug levels in the skin were also analyzed. The unabsorbed donor solution was removed using a pipette and the skin surface (including the tape) was wiped off with the Kimwipes. The skin was then cleaned with two cotton buds soaked in receptor solution followed by two dry cotton buds. The tapes were carefully removed from the skin. Epidermis and dermis of the central diffusion area (Central Epi and Central Der) and the lateral region (Side Epi and Side

Der) were separated, minced, and placed individually in vials containing 2 mL methanol as extraction solvent. The vials were kept overnight on a shaker (New Brunswick

Scientific Co. Inc., Edison NJ, USA) at 150 rpm. The samples were then filtered through

0.22 μm syringe filters (Celltreat Scientific Products, Shirley, MA, USA) and analyzed using HPLC. Probe recovery equation was applied while calculating the drug amount in the dialysates.

123

Quantitative analysis

Reverse phase HPLC method based on UV detection was used for quantitative analysis of diclofenac sodium. Waters Alliance 2695 separation module (Milford, MA,

USA) and 2996 photodiode array detector was used. Isocratic elution was performed on

Phenomenex Luna 5µ C8 (2) , 250 x 4.6 mm, 100 A° (Phenomenex, CA, USA) column at a flow rate of 1.2 mL/min and column temperature of 40 °C after injecting 20 µL of sample. The chromatographic conditions were: methanol (phase A) and 10 mM sodium dihydrogen phosphate buffer (phase B) in the ratio of 66:34. The run time was 15 min and the retention time of the drug was around 9.8 min. Drug standards were prepared in

PBS and detected at wavelength of 275 nm. Linearity was observed in the concentration range of 0.1 - 50 µg/mL (R2 = 0.9999).

Data analysis

Microsoft Excel was used for all the data analysis. Statistical analysis was performed using Student’s t test at p values of 0.01 and 0.05 to conclude any significant difference.

Results

Skin integrity and thickness assessment

Electrical properties of skin have been demonstrated to be an indicator of the physical state of stratum corneum and hence, electrical conductivity is commonly used as a means to assess the barrier integrity of epidermis as well as full-thickness skin

(Lawrence, 1997). Electrical resistance of the skin samples used in the study, ranged from

7 to 10 kΩ. Also, the thickness of dermatomed human skin was found to be 790 ± 70 μm.

124

The thickness of the skin tissue was adequate to facilitate the insertion of the probe into dermis.

Probe insertion and depth measurement

Layers of skin were separated to visually study the depth of the probe. As shown in Figure 17a, microdialysis probe was inserted inside the skin tissue. When the epidermis layer (stratum corneum and viable epidermis) was removed, the probe was observed to be located in the dermis layer (Figure 17b). Further, ultrasound generated from DermaScan was then used to obtain vertical image of probe in skin and confirm the depth of the same. The actual depth of the probe in human skin, as measured using

DermaScan, was found to be 350 ± 130 μm (n=3) (Figure 17c,d). Taking into consideration the thickness of the epidermis (around 150 μm), the depth of the probe validated the location of the probe in the dermis.

125

a. b.

c. d.

Figure 17. Microscopic images of probe in (a). Dermatomed human cadaver skin (b). Dermis layer of human skin; DermaScan images of dermatomed human skin (c) without probe, (d) with probe.

Probe recovery studies

Depending on the physicochemical properties of the drug molecules and molecular weight cut off of the semi-permeable membrane of the microdialysis probe, recoveries of different drugs using these probes may differ. In this study, good correlation between the recovered drug concentrations (dialysates) and actual drug concentrations

126

(solutions in which the probe was placed), was observed as shown by the following equation:

y=0.2487 x-1.2698 (R2=0.997) (Figure 18)

Where, y represents the drug concentration recovered in the probe (µg/mL) and x represents the actual drug concentration in the solution (µg/mL). This equation was applied to calculate the actual drug levels in the skin tissue from the drug concentrations recovered in the probe. Recovery studies were essential for an accurate and reliable estimation of drug levels in the skin.

140 120 y = 0.2487x - 1.2698 100 R² = 0.997 80

60 (µg/mL) 40 20

RECOVERED CONCENTRATION CONCENTRATION RECOVERED 0 0 200 400 600 ACTUAL CONCENTRATION (µg/mL)

Figure 18. Recovery of diclofenac sodium from linear microdialysis probe

In vitro permeation studies

The rate of vertical diffusion of diclofenac sodium was significantly higher in microneedle-treated skin (11.8 ± 2.5 µg/h) than laser-treated skin (6.0 ± 2.3 µg/h, p=0.04) as well as untreated skin (0.9 ± 0.6 µg/h, p=0.01). Also, the vertical diffusion of diclofenac sodium into the central probe was faster in laser group than in the untreated group (p<0.05) (Figure 19a).

127

16.00 LASER MN PASSIVE SD

± 14.00 12.00

10.00 8.00 * 6.00 # 4.00 * 2.00 Rate of drug diffusion (µg/h) (µg/h) drugof diffusion Rate 0.00 0 4 8 12 16 20 24 Time (h)

2.50

± PASSIVE MN LASER 2.00

1.50

SD 1.00 # *

0.50

Rate of drug diffusion (µg/h) (µg/h) drugof diffusion Rate 0.00 0 4 8 12 16 20 24 Time (h) Figure 19. Rate of diffusion in (a). vertical direction into the central probe and (b). horizontal direction into the lateral probe Untreated skin (PASSIVE), PLGA microneedles-treated skin (MN), Ablative laser- treated skin (LASER) (*, # indicated statistical difference between groups, p<0.05, p<0.01, respectively)

The horizontal diffusion of drug in lateral probe was significantly lower in untreated group (0.7 ± 0.1 µg/h) than laser-treated (1.3 ± 0.1 µg/h, p=0.001) and microneedle-treated group (1.7 ± 0.3 µg/h, p=0.01). There was no marked difference in

128

the lateral diffusion of drug between microneedles and laser groups (p=0.27) (Figure

19b).

a. 800 PASSIVE LASER MN 700 SD #

± 600 ) ) 2 500 400 300 200

amount (µg/cm amount 100 Avgerage drugcumulative Avgerage 0 0 4 8 12 16 20 24 Time (h) b. MN LASER PASSIVE 200.00 180.00 * * 160.00 *

SD 140.00 ±

) ) 120.00 2 100.00

80.00 * (µg/cm 60.00 # 40.00 20.00 Average amount drugAverage cumulative 0.00 Side Epi Side Der Central Epi Central Der Skin groups

Figure 20. Drug levels in (a). receptor fluid and (b) skin layers Untreated skin (PASSIVE), PLGA microneedle-treated skin (MN), Ablative laser-treated skin (LASER). Drug amount in Side Epi (epidermis of the tape-covered skin), Side-Der (dermis of the tape-covered skin), Central Epi (epidermis of the formulation-exposed skin), and Central Der (dermis of the formulation-exposed skin). (*, # indicated statistical difference between groups, p<0.05, p<0.01, respectively)

The rate of drug diffusion into both vertical and lateral probes increased with time in all three groups. The cumulative amount of drug delivered into the receptor

129

chamber at 24 h was significantly higher in laser group (659.7 ± 107.3 µg/cm2) than in microneedles (153.9 ± 30.9 µg/cm2, p=0.004) as well as untreated group (125.9 ± 23.8

µg/cm2, p=0.003). There was no marked difference in the amount of drug in the receptor between untreated and microneedle-treated groups (p=0.28) (Figure 20a).

However, drug permeation was observed to increase with time in all the three groups.

Investigation of the barrier and adhesive property of the adhesive tape.

a B

Figure 21. Fluorescent images of (a). with-tape skin (b). backside of tape

As shown in Figure 21a, calcein was not observed to diffuse through the black tape. Sufficient adhesion of tape with the skin prevented the dye from diffusing under the tape (Figure 21b). This design effectively blocked the leakage of the drug solution under the tape to the area of the lateral probe. Calcein was used in this study due to its hydrophilicity and high molecular weight that prevents it from diffusing passively through skin. Furthermore, as per the results for the permeation of diclofenac sodium through the adhesive tapes, no drug was observed in the receptor at the end of 24 h,

130

which confirmed the impermeability of drug across the adhesive tape used in this study design.

Skin distribution studies

In all the skin areas (Side Epi, Side Der, Central Epi, and Central Der), drug levels in the laser group were significantly higher than microneedles as well as untreated group (p<0.05) (Figure 20b). The amount of drug in the epidermis of the tape-covered skin in untreated group (9.9 ± 4.6 µg/cm2) was markedly lower than that in microneedles

(21.3 ± 1.4 µg/cm2, p<0.05) as well as laser group (39.6 ± 8.2 µg/cm2, p<0.01). There was no significant difference in the amount of drug in the dermis of the tape-covered skin as well as epidermis and dermis of the formulation-exposed skin between the untreated and microneedle-treated groups (p>0.05).

Discussion

Microdialysis is a commonly employed analytical technique for studying dermatopharmacokinetics of drugs in vivo. However, it has not been used previously for quantitative estimation of diffusion of drugs in skin in an in vitro setting. The observations and results of this investigation were very interesting. The rate of diffusion of diclofenac sodium in skin as well as its amount in the receptor solution was found to increase with time. This may be explained due to the infinite dose of the drug applied in the donor chamber (2 mL diclofenac sodium solution, 4 mg/mL), that maintained the drug concentration gradient between the surrounding tissue and the probe as well as receptor, and eventually resulted in its increased diffusion and permeation.

The difference in delivery and diffusion rate of diclofenac sodium between untreated and microneedle and laser-treated groups can be explained due to the difference

131

in the barrier function of stratum corneum in the different groups. In the untreated group, stratum corneum was intact and therefore, restricted the penetration of drug into the epidermis and hence, in the probe in the dermis layer. However, in case of the other groups, ablative laser as well as PLGA microneedles disrupted the skin barrier by creating microchannels in the stratum corneum and epidermis, that resulted in diffusion of diclofenac sodium into the dermis layer and subsequently, into the receptor. The depth of microchannels created by PLGA microneedles and ablative laser (approximately 110

µm shown in previous studies) depicted that the microneedles as well as laser successfully porated the epidermis and reached the superficial layer of the dermis

(Nguyen & Banga, 2018a). This, therefore resulted in enhanced vertical as well as lateral diffusion of diclofenac sodium into the probes in the dermis layer in these groups as compared to the untreated skin. These observations were similar to those reported by

Nayak et al., where microneedle-treated skin showed faster diffusion of lidocaine, both vertically as well as horizontally, as compared to untreated skin (Nayak et al., 2015).

Furthermore, the author showed the effects of pressure applied for the insertion of microneedles on the horizontal diffusion profile of the drug in skin (Nayak et al., 2015).

The lower rate of drug diffusion in the central and lateral probe in the laser-treated skin as compared to the microneedle-treated skin may be explained due to differences in characteristics and dimensions of microchannels in the two groups. The pore-to-pore distance in microneedles group (0.5 ± 0.0 mm) was shorter than that in laser group (0.7 ±

0.2 mm, n=10) (Nguyen & Banga, 2018a). This meant that there were more number of pores closer to the probe in microneedle than laser group. Also, the laser-created channels were distributed in larger area, and lesser number of pores were close to the probe.

132

Furthermore, the carbonization layer or coagulation zone around the pores created by laser treatment was found to limit drug diffusion, both vertically and horizontally, in this group (Nguyen & Banga, 2018a; Taudorf et al., 2014, 2015, 2016). On the other hand,

PLGA microneedles perforated the skin tissue and created interstitial fluid-filled microchannels without any carbonization. PLGA microneedles left a minimal trace of the polymer in the microchannels to minimize their interference in drug diffusion (Nguyen &

Banga, 2018a).

Microdialysis is a technique employed for real-time monitoring of drug concentrations in different body tissues such as muscle, brain, etc. In this study, the linear microdialysis probe in skin simulated the blood vessels in the dermis layer and was used to investigate vertical and lateral drug diffusion in skin tissue.

Passive permeation of diclofenac sodium across human cadaver skin and into the receptor solution was consistent with its physicochemical properties: small molecular weight and moderate lipophilicity. The difference in the drug permeability between microneedle and laser-treated groups can be explained due to the dimensions of the microchannels. Laser-created pores were significantly bigger than those formed by microneedles (p<0.05) (Nguyen & Banga, 2018a). Thus, the total volume of the pores created by laser was considerably higher than those created by microneedles, and hence, the amount of drug permeation was greater in the former. The results of drug amount in receptor and skin in this study may be underestimated as compared to the other conventionally designed in vitro permeation studies, since most of the drug that diffused into the skin was recovered from the probes. Drug permeation in the receptor across the microneedle-treated skin was not found to be significantly different from the passive skin

133

(p>0.05). This may be attributed to the high amount of drug recovered from the central and lateral probes in microneedle-treated group than passive as well as laser-treated group.

Further, the results of skin distribution studies can be ascribed to the larger pores as well as greater area of microporation in case of laser-treated skin. The insignificant difference in the drug levels in skin between microneedle-treated and untreated skin might be due to the ability of the drug to enter the skin passively as well as high drug uptake into the probe in the microneedle group.

In this study, we have described an in vitro set-up with the linear microdialysis probe inserted in the dermis layer, which simulates blood circulation in the in vivo scenario and clinical settings. This technique allows to monitor and assess the drug level and rate of drug diffusion in the dermis layer in real time. From a clinical point of view, evaluation of lateral drug diffusion in skin is important as it enables the investigators to manipulate drug delivery by controlling the applied dose and thus, evaluating the treatment efficacy.

Moreover, this study also provides an understanding of the pattern of drug delivery, into and across skin. While conducting in vitro studies, mostly vertical drug diffusion (drug amount in receptor) is analyzed and lateral diffusion is not taken into consideration. The proposed in vitro set up described in this paper, enables determination of vertical as well as lateral drug diffusion at the same time.

Conclusion

This study investigated the vertical and lateral diffusion of drug molecules in dermatomed cadaver human skin. In in vitro permeation studies using microdialysis, a significantly higher rate of both vertical and lateral diffusion of diclofenac sodium was

134

obtained in microneedle and laser-treated skin than untreated group. However, laser treatment resulted in higher amount of drug delivered into and across the skin than microneedle-porated and untreated group. The present study demonstrated a novel, effective, and reliable technique of in vitro microdialysis for simultaneous evaluation of rate of lateral and vertical drug diffusion in skin tissue.

135

CHAPTER 6

FEASIBILITY INVESTIGATION FOR TRANSDERMAL DELIVERY OF ELVITEGRAVIR FORMULATIONS

Abstract

The aim of this study was to test the feasibility of transdermal delivery of elvitegravir (EVG), a potent HIV-1 integrase inhibitor for development of a transdermal patch of the same for HIV prophylaxis. The targeted skin permeation flux was about 25

μg/cm2/h. In vitro drug permeation studies were performed using vertical Franz diffusion cells. Passive permeation of EVG through dermatomed human skin and human epidermis was investigated. Effect of 5% w/w oleic acid, 25% v/v ethanol, 40% w/w dimethyl sulfoxide, 10% w/w lauric acid, and combination of 20% w/w dimethyl sulfoxide, 10% w/w oleic acid, and 5% w/w lauric acid in propylene glycol on the permeation of EVG through human epidermis was evaluated. Phosphate buffered saline (pH 7.4) containing

10% v/v propylene carbonate and/or polyethylene glycol 400 was used as the receptor.

Sampling of the receptor was performed at pre-determined time points for 168 h and analysis was done using HPLC.

Permeation of EVG through dermatomed human skin and human epidermis from its free solution in propylene glycol was found to be 5.51 ± 3.48 μg/cm2 and 21.14 ± 3.23

μg/cm2 (control), respectively after 168 h. Oleic acid and dimethyl sulfoxide significantly enhanced the permeation of EVG to 235.97 ± 49.06 μg/cm2 and 700.01 ± 107.03 μg/cm2, respectively, through human epidermis in comparison to the control group (p<0.05). Lag

135 136

time of 19 h and 3.8 h was observed with the use of oleic acid and dimethyl sulfoxide, respectively. Maximum average transdermal permeation flux of about 5.3 ± 1.58

μg/cm2/h was achieved using dimethyl sulfoxide. Ethanol and lauric acid effectively reduced the lag time to 1.03 h and 1.22 h, respectively, but did not show any enhancement in the permeation of EVG as compared to the control. Combination of different enhancers delivered 348.71 ± 96.50 μg/cm2 of EVG in the receptor after 168 h, which was significantly higher than the control group (p<0.05). Lag time of 1.08 h was observed with this group. Overall, even with the use of different enhancers or their combination, the targeted permeation flux was not achieved. EVG was thus, not considered further for patch development.

Introduction

EVG, a potent HIV-1 integrase inhibitor, is used in combination with other antiretrovirals, as a first line therapy for HIV infection (Y. Gong et al., 2017; Ramanathan et al., 2011). It is effective against wild as well as drug resistant HIV strains (Ramanathan et al., 2011). In addition to the therapeutic effects, EVG has also been investigated for its efficacy as part of a multiple-drug regimen (emtricitabine, cobicistat, tenofovir disoproxil fumarate) for post-exposure prophylaxis (Mayer et al., 2017). Pharmacokinetic studies in macaques demonstrated EVG as one of the promising HIV prophylactic agents (Massud et al., 2014).

Single oral dose of EVG (combination with emtricitabine, cobicistat, tenofovir disoproxil fumarate: available as STRIBILD®) has been reported to be efficacious for

HIV treatment (Desimmie, Schrijvers, & Debyser, 2012; Ramanathan et al., 2011). The absolute oral bioavailability of EVG has not been determined in humans. However in

137

animals, it has been reported to have moderate oral bioavailability (~30%), probably due to extensive first pass metabolism (Ramanathan et al., 2011). EVG also displays non- linear pharmacokinetics that indicates its solubility-limited absorption (Deeks, 2014).

Transdermal drug delivery systems have the potential to bypass the hepatic first- pass effect, resulting in improvement of EVG’s bioavailability. In addition to enhancing the bioavailability, if an intermediate use transdermal dosage form (3-7 day patch) be made available for EVG for HIV prophylaxis, it would be more patient compliant. This may be attributed to the various advantages of transdermal delivery systems over the short-acting tablets and topical gels and long-acting intravaginal rings, injectables, and implants, currently being investigated for different drugs for HIV prevention (Puri,

Sivaraman, et al., 2017).

The major challenge in developing transdermal drug delivery systems is to ensure permeation of drugs across stratum corneum, the outermost dead barrier layer of skin.

However, as discussed in chapter 2, different enhancement technologies can be applied for improving the permeation kinetics of drugs across skin. These include the use of chemical penetration enhancers, microneedles, iontophoresis, laser ablation, microdermabrasion, and sonophoresis. (Banga, 2011; Puri, Sivaraman, et al., 2017).

EVG is a lipophilic (log P = 4.67) drug with molecular weight: 447.8 g/mol, pKa:

6.5, melting point: 162.5 °C, and water solubility of <0.3 μg/mL. It is a dihydroquinoline- type carboxylic acid derivative and its chemical structure has been shown in Figure 22

(Puri, Sivaraman, et al., 2017; Ramanathan et al., 2011).

138

Figure 22. Chemical structure of EVG

Melting point and molecular weight of EVG seem suitable for passive permeation across skin. However, due to high log P and poor aqueous solubility, transdermal delivery of EVG was anticipated to be challenging without using any of the enhancement technologies. As most of the physical enhancement strategies (microneedles, iontophoresis, laser) are applicable and preferred for hydrophilic drugs, we employed chemical penetration enhancers, individually or in combination, for investigating transdermal permeation of EVG across human epidermis. Chemical enhancers improve drug permeation by reversible modification of the stratum corneum barrier resistance

(Banga, 2011; Puri, Sivaraman, et al., 2017).

Thus, the aim of this study was to investigate the formulation feasibility of transdermal delivery of EVG by evaluating the effect of various chemical penetration enhancers on its skin permeation. Effect of oleic acid, DMSO, lauric acid, ethanol, and combination of oleic acid, DMSO, and lauric acid on permeation of EVG across human skin was evaluated in vitro.

139

Materials

EVG was provided by CONRAD (Contraception Research and Development,

Arlington, VA, USA). Propylene glycol and oleic acid were purchased from Ekichem

(Joliet, IL, USA) and Croda Inc. (Edison, NJ, USA), respectively. DMSO was procured from Gaylord Chemical Company, L.L.C (Tuscaloosa, AL, USA). Phosphate buffered saline, 10X- pH 7.4 (PBS) was obtained from Fisher Scientific (Pittsburgh, PA, USA).

Polyethylene glycol 400 (PEG-400), lauric acid, and propylene carbonate were purchased from Sigma–Aldrich (St. Louis, MO, USA). Gentamycin sulfate was purchased from

ACROS (Morris Plains, NJ, USA). Methanol, ethanol, acetonitrile, and trifluoroacetic acid were purchased from Pharmco-aaper (Brookfield, CT, USA). Dermatomed human cadaver skin was obtained from New York Fire Fighters (New York, NY, USA).

Methods

Solubility studies

Solubility of EVG in propylene glycol; DMSO, oleic acid, lauric acid and propylene glycol (20:10:5:65); 1X PBS and PEG 400 (1:1); 1X PBS, PEG 400, and propylene carbonate (4:5:1) was determined. Composition of 1X PBS was 10 mM phosphate ions, 137 mM sodium chloride, and 2.7 mM potassium chloride and its pH was

7.4. An excess amount of drug was added to 1 mL of each solvent and was allowed to shake at room temperature at 150 rpm for 24 h. Samples were then centrifuged at 13,400 rpm for 5 min. The supernatants were diluted with methanol, filtered through 0.45 μm syringe filters (Celltreat Scientific Products, Shirley, MA, USA) and analyzed using the

HPLC method described below under the section of quantitative analysis.

140

Skin preparation

Dermatomed human skin (leg sections) was stored at -80 °C until used. For the permeation studies, it was thawed in 1X PBS at 37 °C and, then cut into pieces of suitable sizes or used for separation of epidermis that was further employed for permeation studies.

Separation of epidermis

For each study, epidermis was freshly isolated from dermatomed human skin by heat-separation method. Skin was immersed in 1X PBS for 2 min at 60 °C. Epidermis was then peeled off with the help of forceps and spatula and thereafter, cut into pieces of suitable size for mounting on the Franz cells (Kassis & Søndergaard, 1982).

Skin integrity and thickness assessment

Resistance of all the skin pieces (dermatomed and epidermis) was measured in order to ensure their barrier integrity before using them for the in vitro permeation studies

(Puri, Murnane, et al., 2017). This was performed using silver-silver chloride electrodes connected to a digital multimeter: 34410A 6 ½ (Agilent Technologies, CA, USA) and arbitrary waveform generator (Agilent 33220A, 20 MHz Function). Skin pieces were then mounted on the Franz cells and allowed to equilibrate for 15 min. PBS (1X, 300 µL) was then added in the donor as well as receptor (5 mL), respectively. After equilibration, silver electrodes were placed in the receptor and silver chloride in the donor compartment, respectively. Load resistor (RL) was connected in series with skin, and drop in the voltage across the skin and entire circuit (VS) was displayed on the multimeter

(VO). Following equation was used to measure the skin resistance (RS):

Rs = Vs RL/ (Vo - Vs)

141

Where, VO and RL were 100 mV and 100 kΩ, respectively. Skin samples with more than 10 kΩ electrical resistance, were selected for the permeation studies. The thickness of the selected skin pieces was measured using a thickness gauge (Cedarhurst,

NY, USA). The thickness of dermatomed skin pieces ranged between 500-700 μm and that of epidermis from 50-150 μm.

In vitro permeation study paradigm

Delivery of EVG through dermatomed human skin and epidermis was investigated by performing in vitro permeation studies for 7 days using jacketed vertical

Franz diffusion cells (PermeGear, Inc., Hellertown, PA, USA) with an effective diffusion area of 0.64 cm2. The receptor compartment was maintained at 37 °C using a recirculating water bath system to maintain skin at 32 °C. Cells were washed and filled with 5 mL of receptor solution (as specified below) to maintain sink conditions. The skin pieces were then mounted on the Franz cells with the stratum corneum side facing upwards. Drug solutions were added in the donor chamber and receptor solution (300 µL) was sampled at pre-determined time points with replacement with equal volume of fresh buffer solution. EVG amount in the samples was quantified using HPLC. Results were reported as average ± SE (n = 6) for each test group.

The details of the specifications of the procedure followed to investigate the passive permeation of EVG as well as to evaluate the effect of different chemical enhancers on its permeation through human skin have been provided in the following sections.

142

Passive permeation of EVG. Passive permeation of EVG across both dermatomed human cadaver skin as well as epidermis, was evaluated. In this study, 500

μL of EVG solution in propylene glycol (3.24 mg/mL, equivalent to 90% saturation solubility) was added in the donor. PBS (1X): PEG 400 (1:1) containing gentamycin sulfate (80 mg/L) was used as the receptor solution. Gentamycin was added as an antibacterial as the in vitro permeation studies were performed for 7 days. The receptor

(300 μL) was withdrawn at 2, 4, 6, 8, 22, 24, 36, 48, 60, 72, 84, 96, 108, 120, 132, 144,

156, 168 h and replaced with equivalent volume of fresh receptor solution.

Effect of different chemical enhancers on the skin permeation of EVG. The effect of different chemical enhancers on transdermal delivery of EVG across human epidermis was investigated in this study. Drug vehicles comprised of 5% w/w oleic acid, 25% v/v ethanol, 40% w/w DMSO, and 10% w/w lauric acid, prepared individually in propylene glycol. The donor solution consisted of 500 µL of EVG in each of the vehicles. The drug concentration, sampling time points, and replacement strategy was similar to the passive permeation study.

Effect of combination of enhancers on the skin permeation of EVG. The effect of combination of 10% w/w oleic acid, 5% w/w lauric acid, and 20% w/w DMSO in propylene glycol (drug vehicle) on transdermal delivery of EVG across human epidermis was investigated in this study. Donor solution comprised of EVG solution in the drug vehicle (200 μL, concentration of 31.29 mg/mL- equivalent to 90% saturation solubility in the vehicle). PBS (1X): PEG 400: propylene carbonate (4:5:1) containing gentamycin sulfate (80 mg/L) was used as the receptor solution to maintain sink conditions. Sampling time points and replacement strategy were similar to the passive permeation study.

143

Calculation of permeation flux and lag time. Permeation flux was calculated as the slope of linear portion of the permeation profiles (cumulative drug permeated/cm2 plotted against the time). The x-intercept of the extrapolated linear portion of the same profile indicated the lag time (Puri, Murnane, et al., 2017).

Skin extraction

After 168 h, donor formulations were removed with 3 dry cotton swabs first, followed by 2 cotton swabs soaked in receptor solution. The epidermis from each cell was then minced individually and placed in 2 mL methanol in 6 well plate. The plate was placed on a shaker (New Brunswick Scientific Co. Inc., Edison NJ, USA) for overnight shaking at 100 rpm. The samples were filtered through 0.45 μm syringe filters (Cell treat

Scientific Products, Shirley, MA, USA) and analyzed using HPLC.

Quantitative analysis

A UV detection based RP-HPLC method was used for quantitative analysis of

EVG. Waters Alliance 2695 separation module (Milford, MA, USA) coupled with 2996

PDA detector was employed. Isocratic elution was done on Phenomenex Luna 5μ C8 (2)

100A, 250*4.6 mm (Phenomenex, CA, USA) at 1.0 mL/min flow rate and column temperature of 30 °C after injecting 20 μL of sample. The chromatographic conditions were: acetonitrile (phase A) and 0.1% v/v trifluoroacetic acid in DI water (phase B) in the ratio of 70:30. The run time was 10 min and the retention time of EVG was around 6.2 min. Drug standards were prepared in 1X PBS: PEG 400 (1:1) and detected at wavelength of 258 nm. The precision limit of detection and quantification were 0.002

μg/mL and 0.006 μg/mL, respectively. Linearity was observed in the concentration range of 0.1–50 μg/mL (R2 = 0.9999).

144

Data analysis

All the data was analyzed using Microsoft Excel. Unpaired Student’s t-test was applied for statistical analysis and p value of less than 0.05 was considered for concluding significant difference between the test groups. All the results have been presented as mean ± SE.

Results and Discussion

Owing to the lipophilic properties of EVG (log P: 4.67 and poor aqueous solubility: <0.3 μg/mL), it was expected to have low passive skin permeation and achieving the targeted flux of 25 μg/cm2 /h was anticipated to not be possible without using different chemical and physical enhancement strategies. However, as the ultimate goal was to investigate the feasibility of formulating a transdermal patch of EVG and since most of the physical enhancement strategies are employed for hydrophilic drugs, it was considered better to investigate the effect of various chemical enhancers on its permeation across skin. Chemicals that are commonly incorporated as skin penetration enhancers in topical and transdermal products such as oleic acid, ethanol, lauric acid, and

DMSO were explored for their effect on permeation of EVG.

Solubility studies

Table 3 shows the solubility of EVG determined in different solvent systems. The saturation solubility in propylene glycol as well as combination of enhancers in propylene glycol was investigated so as to be able to use 90 % saturated drug concentration as donor for the in vitro studies (providing maximum thermodynamic activity). Also, as EVG was reported to have poor aqueous solubility, solubility in mixture of 1X PBS and PEG 400/

PEG 400 + propylene carbonate was determined in order to investigate the satisfaction of

145

sink conditions to employ these solvents as receptor solution for the in vitro permeation studies.

Table 3. Solubility of EVG in different solvents

SOLVENTS SOLUBILITY (mg/mL) Propylene glycol 3.6

DMSO, oleic acid, lauric acid and propylene glycol (20:10:5:65) 34.77

1X PBS and PEG 400 (1:1) 0.21 1X PBS, PEG 400, and propylene carbonate (4:5:1) 3.38

In vitro permeation studies

Passive permeation of EVG. As shown in Figure 23, permeation of EVG through dermatomed human skin and human epidermis from its solution in propylene glycol was found to be 5.51 ± 3.48 μg/cm2 and 21.14 ± 3.23 μg/cm2 (control), respectively after 168 h. In addition to 4-fold enhancement in drug permeation, lag time was reduced from 116 h to 9.5 h with the use of human epidermis as compared to dermatomed skin.

These results were in agreement with those reported in the literature previously.

Amalia et al. investigated the amount of nafarelin delivered across dermatomed human skin (300-500 μm in thickness) and epidermis after application of iontophoresis in an in vitro set up. The cumulative amount (nmol/cm2) of drug that was delivered after 24 h was

3.97 ± 0.1 for dermatomed full thickness skin and 28 ± 13.3 for epidermis (Rodríguez

Bayón & Guy, 1996). The significant difference between drug delivery across epidermis and dermatomed skin was attributed to the kinetic effect associated with longer diffusional path length in case of the thicker skin piece (Rodríguez Bayón & Guy, 1996).

146

Thus, the same reasoning may support the observations made in our study as well. Also, due to lesser diffusional pathlength for the drug in case of epidermis, in addition to observance of more drug in the receptor, it was detected in the receptor earlier (9.5 h) as compared to dermatomed skin (116 h).

30.00

25.00 EPIDERMIS DERMATOMED SKIN *

20.00

± ) )

2 15.00 /cm

10.00

μg (

5.00

Average cumulative amount of EVG amount cumulative Average 0.00 0 50 100 150 200 Time (h)

Figure 23. Passive permeation profile of EVG

In case of transdermal delivery, where the drug is expected to reach the systemic circulation, human epidermis has been stated as a better model than full-thickness and dermatomed human skin for in vitro studies. This has been explained as in actual in vivo conditions, drugs are required to successfully cross the epidermis in order to reach the blood circulation as the latter lies just below the epidermal-dermal junction. The use of full-thickness or dermatomed skin in an in vitro set up may actually underestimate the amount of drug that can be delivered in an in vivo scenario (Banga, 2011). Therefore, as

147

EVG was intended to be delivered transdermally, human epidermis was selected for the future studies.

Effect of chemical enhancers on skin permeation of EVG. The amount of EVG in receptor and skin was analyzed.

Amount of EVG in receptor. Oleic acid (5% w/w) significantly enhanced the permeation of EVG to 235.97 ± 49.06 μg/cm2 through human epidermis in comparison to the control group (p<0.05) as shown in Figure 24. However, the lag time for drug permeation was observed to be about 19 h. Incorporation of 40% w/w DMSO into the donor formulation, further enhanced the permeation of EVG to 700.01 ± 107.03 μg/cm2 as shown in Figure 24 and reduced the lag time to 3.8 h as compared to the control group

(p<0.05). The average permeation flux observed with passive, oleic acid, and DMSO group was 0.12 ± 0.02, 1.76 ± 0.53, and 5.3 ± 1.58 μg/cm2/h, respectively. However, targeted flux of 25 μg/cm2/h for HIV prophylaxis could not be achieved with oleic acid and DMSO.

Further, ethanol (6.74 ± 1.84 μg/cm2) and lauric acid (26.53 ± 4.16 μg/cm2) did not show any enhancement in the permeation of EVG as compared to the control after

168 h, but effectively reduced the lag time to 1.03 h and 1.22 h as can be seen from the day 1 permeation profile (Figure 24, 25). EVG solution containing lauric acid was observed to gel after 24 h. The comparison of lag times observed in case of different enhancers has been depicted in Figure 26.

148

Figure 24. Effect of chemical enhancers on EVG permeation through human epidermis * denotes significant difference (p<0.05) as compared to other groups, Student’s t test

12 PROPYLENE GLYCOL LAURIC ACID ETHANOL

SE 10

) ) 2

g/cm μ

( 6

2 Average cumulative amount of EVG of amount cumulative Average

0 0 5 10 15 20 25 Time (h)

Figure 25. Effect of ethanol and lauric acid on EVG permeation through human epidermis during day 1

149

20 19.32

9.5

10 Time Time (h)

5 3.79

1.03 1.22 0 PASSIVE PG OA DMSO ETHANOL LAURIC ACID Test groups

Figure 26. Lag times for EVG permeation through human epidermis in different test groups

Thus, overall, oleic acid and DMSO were successful in enhancing the permeation of EVG across human epidermis. Oleic acid is a very commonly employed permeation enhancer in topical and transdermal formulations and its mechanism of permeation enhancement has already been well described in literature as delipidization or fluidization of the stratum corneum. It accumulates in the lipophilic domains of stratum corneum, fluidizes it, and creates permeable pores that facilitate drug permeation. The mechanism of the permeation enhancing effect of oleic acid has been demonstrated by differential scanning calorimetry (Y. Yang et al., 2011), electron spin resonance studies, FTIR,

Raman spectroscopy, and X-ray diffractometry (H. Benson, 2005). Moreover, oleic acid and propylene glycol work synergistically and enhance the skin permeation of drugs.

Propylene glycol acts as a cosolvent and aids in enhancing the concentration of both the drug as well as chemical enhancer in stratum corneum (Puri, Murnane, et al., 2017). Also,

150

the observation of high lag time of 19 h with oleic acid may be attributed to the time taken by oleic acid (due to its high molecular weight) to diffuse from the formulation to reach the stratum corneum, fluidize the lipids and then enable drug permeation

(Yerramsetty et al., 2010).

DMSO has also been reported to enhance the transdermal permeation of drugs.

Mechanistically, it acts by inducing conformational changes in keratin, interacting with the alkyl chains of stratum corneum lipids, displacing water bound with keratin, and extracting lipids from skin (Lane, 2013). Thus, overall due to these effects on the barrier layer of skin, DMSO probably facilitated increased delivery of EVG across the epidermis layer.

Ethanol and lauric acid are also commonly employed skin permeation enhancers.

However, they were not able to significantly enhance the delivery of EVG across human epidermis. But, they considerably reduced the lag time to about 1 h as compared to 9.5 h in passive and 19 h in oleic acid group. Therefore, EVG permeated faster in presence of ethanol and lauric acid as compared to other test groups. In the former, this can be explained due to the evaporative loss of ethanol from the formulation that results in increase in the drug concentration beyond saturation, providing a supersaturated state with a greater driving force for permeation. Furthermore, ethanol itself rapidly permeates into skin (steady state flux ~ 1 mg/cm2/h) and thus, can carry the permeants along with it

(Berner et al., 1989). Therefore, higher amount of drug permeation was observed in presence of ethanol than the passive group only in the initial hours of permeation. As the permeation study was performed for 7 days and ethanol would have evaporated in the initial hours itself, no overall enhancement in drug delivery was observed.

151

Lauric acid has previously been investigated as a transdermal permeation enhancer, even though its mechanism of action is not clearly understood. Lauric acid in propylene glycol (10% w/w) has previously been demonstrated to work as an effective combination for enhancing the delivery of highly lipophilic drugs across mice skin

(Funke et al., 2002). However, in our study, solution of EVG in lauric acid-propylene glycol mixture was found to gel after day 1 and some crystals were also visible in the formulation. This may have happened due to incompatibility between EVG and lauric acid. However, during day 1, before gelling and crystallization was evident, EVG was found to permeate across skin at a faster rate as compared to passive and oleic acid group.

This was understood from the lag time of 1 h observed for lauric acid group.

Figure 27. Amount of EVG retained in human epidermis after 7 days * denotes significant difference (p<0.05) as compared to other groups, Student’s t test

152

EVG amount in skin. As shown in Figure 27, oleic acid (148.66 ± 26.91

µg/cm2) and passive group (219.56 ± 23.46 µg/cm2) showed highest drug retention in epidermis than other groups (p<0.05). In case of passive group, propylene glycol alone facilitated penetration of EVG across stratum corneum, but most of the drug amount was retained in the epidermis and did not reach the receptor. However, oleic acid was effective in enhancing permeation of EVG into as well as across epidermis. This may be attributed to the permeating enhancing effect of oleic acid as well as its lipophilic nature that may have resulted in its binding with skin and thus, drug retention as well. However,

DMSO did not facilitate in retaining high EVG amount in the skin (40.00 ± 6.18 µg/cm2), but significantly enhanced drug permeation across epidermis (700.01 ± 107.03 μg/cm2).

With ethanol and lauric acid, significantly low amount of EVG was observed in the epidermis (15.97 ± 2.67 µg/cm2 and 53.60 ± 11.52 µg/cm2, respectively) as compared to propylene glycol and oleic acid groups (p<0.05). Thus, in these two cases, most of the drug remained unabsorbed (in the donor itself). In case of passive as well as ethanol and lauric acid, more drug was obtained in the skin than receptor. This may be attributed to the lipophilic properties of EVG (log P: 4.67) that may have resulted in its binding to lipids in stratum corneum and thus, more retention than permeation across epidermis.

Effect of combination of chemical enhancers on skin permeation of EVG.

Combination of 20% w/w DMSO, 10% w/w oleic acid, and 5% w/w lauric acid in propylene glycol was selected based on the observations with the studies conducted with individual enhancers. As DMSO and oleic acid were found to be effective in the abovementioned studies, they were selected for the combination study, but as 40%

DMSO was too high from the perspective of including in a transdermal patch, therefore

153

we used 20% w/w. However, oleic acid was enhanced from 5 to 10% w/w to explore the effect of increasing its concentration. Also, as lauric acid was successful in reducing the lag time for drug permeation, it was included in the combination study. However, due to crystallization issues with 10% w/w lauric acid, its concentration was reduced to 5% w/w in this study. Also, as our aim was to assess the feasibility of achieving transdermal permeation flux of EVG of 25 μg/cm2/h, drug concentration in the donor was increased and corresponded to the saturation solubility of EVG in the combination of enhancers.

1200.00

PROPYLENE GLYCOL OLEIC ACID 1000.00 DMSO COMBINATION

800.00

± ) )

2 600.00

g/cm μ ( 400.00

200.00 * Average cumulative Average cumulative amount EVGof 0.00 * 0 50 100 150 Time (h)

Figure 28. Effect of combination of enhancers on EVG permeation through human epidermis * denotes significant difference (p<0.05) as compared to other groups, Student’s t test

As depicted in Figure 28, combination of 20% w/w DMSO, 10% w/w oleic acid, and 5% w/w lauric acid in propylene glycol delivered 348.71 ± 96.50 μg/cm2 of EVG in the receptor after 168 h, which was significantly higher than the control group (p<0.05)

154

but not as compared to DMSO and oleic acid group (p>0.05). Permeation flux of 2.18 ±

0.49 μg/cm2 and lag time of 1.08 h was observed with the combination of penetration enhancers. Also, the formulation was observed to gel after 24 h, that slowed down the permeation of EVG and thus, significant enhancement in delivery of EVG (that was expected) was not observed.

Thus, overall, none of the enhancers evaluated provided the targeted permeation flux for EVG, even after being added at the maximum possible concentrations. Maximum average permeation flux of about 5.3 ± 1.58 μg/cm2/h was achieved using DMSO, but that was also 5 fold lesser than the target. Therefore, due to high dose requirements of

EVG for HIV prophylaxis and its poor permeation profile across human epidermis in the solution studies, it was not considered as a good candidate for patch formulation and further studies were not performed.

Conclusion

EVG was observed to permeate passively through human dermatomed skin as well as epidermis. However, the drug permeation flux without using chemical enhancers was low and showed high lag time for permeation. Oleic acid (5% w/w) and DMSO

(40% w/w) were found to be most effective in significantly enhancing the amount of

EVG permeation through human epidermis. Ethanol and lauric acid did not improve the transdermal permeation of EVG, but shortened the lag time considerably as compared to the control and other enhancers. However, as the targeted permeation flux for HIV prophylaxis could not be achieved for EVG with the use of different enhancers, it was not considered as a suitable drug candidate for transdermal patch formulation.

155

CHAPTER 7

DEVELOPMENT OF A TRANSDERMAL PATCH OF TENOFOVIR ALAFENAMIDE FOR HIV PROPHYLAXIS

Abstract

The aim of this study was to develop a seven-day transdermal patch of tenofovir alafenamide (TAF), an effective nucleotide reverse transcriptase inhibitor that suppresses

HIV-1, and investigate its permeation through human epidermis. The targeted transdermal permeation flux of TAF for HIV prophylaxis was 7 μg/cm2/h. Patches with drug concentrations (2-15% w/w), 5% w/w oleic acid and coat weight of 200 gsm were prepared in DURO-TAK® 387-2516. Polyvinyl pyrrolidone was included as a crystallization inhibitor in patches with high TAF concentrations (10 and 15% w/w).

Effect of chemical enhancers such as oleyl alcohol (5% w/w), triacetin (5% w/w), and octisalate (8.5% w/w), each of them in combination with 5% w/w propylene glycol, on the permeation of TAF from the 2% drug patches of higher coat weight (400 gsm) was investigated. In vitro permeation studies of TAF patches through human epidermis were performed using Franz cells.

Cumulative amount of TAF that permeated from the 2% drug patch containing only oleic acid in the adhesive, was found to be 19.2 ± 2.92 μg/cm2 after 7 days (control).

However, 2% TAF patches containing oleic acid and combination of propylene glycol and oleyl alcohol, propylene glycol and triacetin, propylene glycol and octisalate, significantly enhanced the permeation to 96.40 ± 14.83 μg/cm2, 87.15 ± 8.60 μg/cm2, and

155 156

89.56 ± 17.28 μg/cm2, respectively (p<0.05). There was no significant difference in the drug permeation between the groups with combination of enhancers (p>0.05). However, the highest permeation flux observed with the use of combination of enhancers was 0.60

± 0.10 µg/cm2/h. Overall, increasing coat weight of the patches and addition of different permeation enhancers resulted in 5 fold increase in the permeation of TAF through human epidermis. All the 2% TAF patches were found to be stable at room temperature,

40 °C, and -20 °C for six months. Thus, a stable transdermal patch formulation of TAF was formulated, but achieving the targeted permeation flux with the same was not feasible.

Introduction

Tenofovir (TFV) represents the cornerstone of prophylactic and therapeutic approaches to control HIV infection (De Clercq, 2017). TFV disoproxil fumarate is currently being replaced by another prodrug, tenofovir alafenamide fumarate (TAF), which shows better affinity for lymphoid tissue, yielding higher levels of TFV- diphosphate (active metabolite) and lower levels of plasma TFV, and displaying a better renal and bone safety profile (Snopková, Havlíčková, & Husa, 2016). For HIV treatment, it is typically administered in combination with other antiretrovirals (ARVs) in daily single tablet regimens (STRs). Although STRs represent a significant advantage over the need to take multiple tablets 3 times a day, which characterized the beginnings of highly

ARV therapy, it still remains a high burden to the patient leading to poor adherence, lower efficacy, and increased risk of viral resistance. Sustained release of ARVs, which can prevent these problems, is therefore needed. Development of sustained release delivery technologies in this space would be of high clinical relevance. Currently, there

157

are no delivery systems designed for clinically using TAF, specifically for HIV prophylaxis.

A variety of delivery systems such as tablets (oral/vaginal), gels (rectal/vaginal), vaginal films, ointments, rings, diaphragm based devices and electrospun fibers, oral/parenteral liposomes and nanoparticles, injectables, and subdermal implants for prophylactic anti-HIV agents have been explored (Ball et al., 2012; Dou et al., 2006;

Gunawardana et al., 2015; Vedha Hari et al., 2013). Drug delivery systems such as oral tablets and coitally-dependent vaginal gels require daily application, are typically short- acting, and have been associated with poor user adherence. Long-acting coitally- independent systems such as implants and injectables do not present these problems, but are riskier if safety or idiosyncratic issues arise (Gunawardana et al., 2015; Q. A. Karim et al., 2010; S. S. A. Karim et al., 2011; Puri, Sivaraman, et al., 2017). An intermediate duration drug delivery system, such as transdermal patch, exerting its pharmacological effect for a few days to a week, would thus be beneficial to the users, providing a more sophisticated and convenient option.

Transdermal drug delivery offers an attractive alternative to oral and other systemic routes of delivery, allowing the drug substances to reach the systemic circulation across the skin barrier. Transdermal drug delivery systems can be beneficial to prevent the hepatic first pass effect of drugs, eliminate peak and valley plasma drug concentrations which are usually associated with the oral and injectable drug delivery, avoid the degradation of drugs in GI tract, and in general, be non-invasive and convenient to administer (Banga, 2011). ARV drug substances may potentially be delivered through the transdermal route in a consistent and sustained manner (e.g. zero order release

158

kinetics), which would provide several advantages over the other delivery systems explored so far (Ham & Buckheit, 2015). Being semi-discreet and easy to apply, without clinical intervention, transdermal patches are more patient compliant than the injectable delivery systems currently under investigation. Additionally, transdermal patches offer ease of discontinuation making them safer to use. The aim of this study was to develop and characterize a transdermal patch for delivery of TAF and investigate its permeation through human epidermis.

There are generally two types of passive transdermal patches, namely, matrix

(drug-in-adhesive) and reservoir-based patch. Matrix-based transdermal patch formulation, the system of choice in this study, consists of a release liner, an adhesive matrix and an impermeable backing membrane. In addition to ease of use and manufacturability as well as acceptable cost of goods, one of the significant advantages of matrix type transdermal patches compared to reservoir patches is the absence of dose dumping (Alexander et al., 2012). The matrix transdermal patches are usually prepared using organic solvent based pressure sensitive adhesives (PSAs) such as acrylate copolymer, silicone, polyisobutylene (PIB), either alone or in combination with each other (Lobo, Sachdeva, & Goswami, 2016). Drug substance can be either dissolved or dispersed in the adhesive matrix, resulting in clear/transparent or suspension patches, respectively.

From transdermal patch development perspective, drugs with low molecular weight (<500 Da), suitable melting point (<250 °C) and moderate log P (1-3) are ideal for passive permeation through skin. With a molecular weight of 476.47 g/mol, melting point of 279 °C and moderate lipophilicity (logP = 1.8) as calculated using chemicalize

159

software (MarvinSketch: version 6.2.2, ChemAxon, Hungary, Europe), TAF, the newest prodrug of TFV in the market, is a promising ARV candidate for development of transdermal patch formulation.

The aim of this study was to develop seven-day transdermal patches for sustained release of TAF. To our knowledge, transdermal patches comprising of ARV agents for

HIV therapy or prophylaxis have not yet been developed. Thus, this is the first study to report such a system. In the present study, transparent acrylate based patches of various compositions were prepared and evaluated for in vitro permeation across human epidermis for 7 days. Considering achieving release of a dose of 8 mg/day from a 50 cm2 transdermal patch, targeted permeation flux of TAF for HIV prophylaxis was about 7

µg/cm2/h. Different chemical enhancers, plasticizers, as well as crystallization inhibiting agents were included in the patch formulations and all the patches were assessed for stability for 6 months at ambient and high temperature conditions.

Materials

TAF was purchased from Pharmacodia Co., LTD (Beijing, China). Backing membranes (ScotchpakTM 9733-2.05 mil polyester film) and release liner (ScotchpakTM

1022- 3 mil fluoropolymer coated polyester film) were gifted by 3M (St. Paul, MN,

USA). Acrylate PSA (DURO-TAK® 87-2516) was obtained as gift sample from Henkel

Corporation (Dusseldorf, Germany). Oleyl alcohol, PEG 400, and polyvinyl pyrrolidone

(PVP 360) were obtained from Sigma–Aldrich (St. Louis, MO, USA). Propylene glycol and oleic acid were purchased from Ekichem (Joliet, IL, USA) and Croda Inc. (Edison,

NJ, USA), respectively. Triacetin, octisalate, gentamycin sulfate, tetra hydrogen furan

(THF), sodium dihydrogen phosphate, sodium hydroxide, 10X PBS, pH 7.4, were

160

obtained from Fisher Scientific (Pittsburgh, PA, USA). Kollidon® VA64, Kollidon® 90F, and Kollidon® 30LP were purchased from BASF (Florham Park, NJ, USA). Methanol, ethanol, acetonitrile, and trifluoroacetic acid were purchased from Pharmco-aaper

(Brookfield, CT, USA). Dermatomed human cadaver skin was obtained from New York

Fire Fighters (New York, NY, USA).

Methods

Slide crystallization studies

TAF was dissolved in methanol (10 mg/mL) and a drop of this solution was placed on a glass slide. Methanol was allowed to evaporate at room temperature under a fume hood. TAF crystals obtained were then observed under Leica DM 750 optical microscope (Leica Microsystems Inc., Buffalo Grove, IL, USA). The DFC-295 camera, attached to the microscope was used to capture images at magnifications of 10x or 20x

(as specified). Similar procedure was used to determine the saturation solubility of TAF in different adhesives (DURO-TAK® 87-2516, DURO-TAK® 87-6908, and BIO-

PSA 7-4301), solution of 5% (w/w) oleic acid in DURO-TAK® 87-2516 as well as in the additives: PVP 360, Kollidon® VA64, Kollidon® 90F, and Kollidon® 30LP. For solubility studies in adhesives, blends of different concentrations of TAF in adhesives, ranging from 1-12% w/w (dry weight) were prepared and a drop of the transparent blends (with completely dissolved drug) was individually placed on glass slides and kept in a flameproof oven at 100 °C for 30 min for evaporation of the solvents in the adhesives.

Similarly, different blends with TAF concentration ranging from 3-5% (w/w) in a mixture of 5% (w/w) oleic acid in DURO-TAK® 87-2516 (dry weight basis) were

161

prepared and a drop of each of the solutions was placed on the glass slides and exposed to the same drying conditions as mentioned above.

In order to determine the solubility of TAF in different additives, the two components were weighed in the following ratios (% w/w): 1:9, 2:8, 3:7, 4:6, 5:5, 6:4,

7:3, 8:2, 9:1, and dissolved in methanol. A drop of each of these solutions was mounted individually on glass slides and observed for crystals after allowing evaporation of methanol at room temperature. The appearance of crystals in all the test samples was observed for a week before concluding the saturation solubility of the drug in different components. The latter was indicated by the highest concentration at which no crystals were observed (Sachdeva, Bai, Kydonieus, & Banga, 2013). The percentages of solid content that were used to calculate the wet weight of adhesives were 41.5% for DURO-

TAK® 87-2516, 38% for DURO-TAK® 87-6908, and 60% for BIO-PSA 7-4301.

Preparation of drug in adhesive patches

Pre-determined amounts of TAF, adhesives, and additives (oleic acid, oleyl alcohol, propylene glycol, octisalate, triacetin as specified in Table 4) were weighed into glass jars with airtight lids to minimize loss of organic solvents. For the patch formulations AP4 to AP11, excipients such as PVP 360, Kollidon® VA64, Kollidon®

90F, and Kollidon® 30LP were dissolved separately in methanol and the solutions were then added to the mixture of adhesive and other excipients. The blends were kept overnight on the rotary mixer (Preiser Scientific Inc., St. Albans, WV, USA). The compositions of the different acrylate based drug patches have been shown in Table 4.

Solid content % of 41.5 was used for the calculation of the wet weight of DURO-TAK®

87-2516 for the formulations. The homogenous mixtures were then cast on the release

162

liner (fluoropolymer coated side of ScotchpakTM 1022) using a Gardner film casting knife

(BYK-AG-4300 series, Columbia, MD, USA). The target coat weight of different patches varied from 100-400 gsm (grams per square meter). The cast sheets were dried in a flameproof oven at 95 °C for 40 min. After drying, they were laminated using

ScotchpakTM 9733 (backing membrane) with the help of a roller, ensuring that no air pockets were formed. The 200 gsm and 400 gsm patches were made by preparing one

100 gsm patch and then stacking 1 and 3, 100 gsm cast films (coated on release liner), respectively on that one by one, after removing the liner of the previous laminate.

Microscopy of acrylate-based patches

The prepared acrylate patches were punched and stored at 40 °C, room temperature, and -20 °C and observed for occurrence of drug crystallization under an optical microscope (Leica DM 750) at predetermined time points for up to 6 months

(Banerjee, Chattopadhyay, Ghosh, Bhattacharya, et al., 2014). In case of patches that showed appearance of crystals in the first week of storage at room temperature or 40 °C, microscopic observations after storage at -20 °C were not performed.

Coat weight and drug content of the patches

Coat weight of the prepared patches was determined by punching and weighing

163

then analyzed using HPLC to quantitate the amount of TAF in the same. Stability of the drug after exposure to high temperature conditions used for processing, was indicated by comparing the experimental and theoretical drug content values. Results have been reported as average ± SD.

In vitro skin permeation studies

In vitro drug permeation studies from transdermal patches through human epidermis were performed for 7 days using in vitro vertical Franz diffusion cells

(PermeGear, Inc., Hellertown, PA, USA), providing an effective diffusion area of 0.64 cm2 (n=6).

Separation of epidermis. For each permeation study, human epidermis was freshly isolated from dermatomed human skin by heat-separation method. Skin was immersed in

1X PBS (10 mM phosphate ions, 137 mM sodium chloride, and 2.7 mM potassium chloride, pH 7.4) for 2 min at 60° C. Epidermis was then carefully peeled off manually, with the help of forceps and spatula and thereafter, cut into pieces of suitable size for mounting on the Franz cells (Kassis & Søndergaard, 1982).

Epidermal Integrity and Thickness Assessment. Resistance of human epidermis was measured in order to select the pieces with acceptable initial barrier integrity for the in vitro permeation study (Puri, Murnane, et al., 2017). This was carried out with the use of silver-silver chloride electrodes attached to a digital multimeter: 34410A 6 ½ (Agilent

Technologies, CA, USA) as well as an arbitrary waveform generator (Agilent 33220A,

20 MHz Function). Epidermis was mounted on the Franz diffusion cells and left for equilibration for 15 min. Phosphate buffer, pH 6.0 was then added in the donor (300 µL)

164

Table 4. Composition of TAF containing acrylate-based transdermal patches

PATCH FORMULATION CODES

PROPERTIES AP1 AP2 AP3 AP4 AP5 AP6 AP7 AP8 AP9 AP10 AP11 AP12 AP13 AP14 AP15 Coat weight (gsm) 200 200 200 100 100 100 100 100 100 100 100 400 400 400 400

Components % Dry weight (% w/w)

TAF 2 3 4 5 10 15 10 10 10 7.5 7.5 2 2 2 2

Oleic acid 5 5 5 5 5 5 5 5 5 5 5 5 5 5 -

DURO-TAK® 87 2516 93 92 91 87.86 80.71 73.57 83.89 83.89 83.89 77.50 79.29 83 83 79.5 83

PVP 360 - - - 2.14 4.29 6.43 - - - 5 - - - - -

Kollidon® VA64 ------1.11 ------

Kollidon® 90 F ------1.11 ------

Kollidon® 30 LP ------1.11 - 3.21 - - - -

Oleyl alcohol ------5 - - 10

Propylene glycol ------5 5 5 5 5 5

Triacetin ------5 - -

Octisalate ------8.5 -

165

and receptor (5 mL), respectively. Following equilibration, silver and silver chloride electrodes were placed in the receptor and donor compartment, respectively. Load resistor (RL) attachment was in series with skin, and the voltage drop across the entire circuit and skin (VS) was displayed on the multimeter (VO). Following formula was used to measure the skin resistance (RS):

Rs = Vs RL/ (Vo - Vs)

Where, VO and RL were 100 mV and 100 kΩ, respectively. Epidermis pieces with electrical resistance of more than 10 kΩ were selected for the permeation study. The thickness of the selected epidermis pieces was measured using a thickness gauge

(Cedarhurst, NY, USA).

Selection of the receptor solution. Stability of TAF in different solvent systems

(1X PBS; phosphate buffer, pH 6.0; 1X PBS: PEG 400- 1:1; PEG 400) at 37° C was assessed in order to select a suitable receptor solution for the 7 day permeation studies.

For this study, different TAF concentrations were prepared (0.5, 5, 50 µg/mL) in the above specified solvents and kept at 37 ° C. The solutions were analyzed for drug content at 0, and after 24 h using HPLC. Solubility of TAF in the selected receptor media

(phosphate buffer, pH 6.0) was determined to ensure maintenance of sink conditions. For this purpose, excess drug was added to 1.0 mL of the buffer and it was allowed to shake for 24 h at room temperature. Thereafter, the solution was filtered using 0.22 µm syringe filters (Cell treat Scientific Products, Shirley, MA, USA) and analyzed using HPLC after suitable dilution.

In vitro skin permeation set up. Permeation of TAF from different transdermal patch formulations (donor) mentioned in Table 6, through human epidermis was

166

investigated using the Franz diffusion cell set up. Taking into consideration the stability of the drug as well as for maintenance of sink conditions, phosphate buffer, pH 6.0 having gentamycin sulfate (80 mg/L) as an anti-microbial agent for the 7 day permeation study (Sachdeva et al., 2013), was selected as the receptor compartment. The temperature of the receptor phase was maintained at 37 °C and it was constantly stirred at 600 rpm.

The freshly isolated epidermis pieces were clamped between the donor and receptor compartments.

After measuring the skin resistance using the procedure described above, phosphate buffer from the donor was pipetted out, epidermis was removed, placed flat on a glass plate, and dried with the help of Kimwipes. Release liner of the patches (of size enough to cover the diffusion area) were removed and the latter were applied carefully on the dried epidermis such that the adhesive layer of the patch was adhered to the stratum corneum side of the skin. Glass rod was rolled on the patches to ensure their adherence to the skin. The epidermis pieces with the applied patches were then mounted between the donor and receptor compartment and the entire set up was secured in place using a clamp.

Receptor (0.3 mL) was sampled at 0, 2, 4, 6, 8, 24, 48, 72, 96, 120, 144, and 168 h. It was replaced with fresh receptor solution after 2, 4, 6, 8 h and entire receptor (5 mL) was removed and replaced with fresh buffer at 24, 48, 72, 96, 120, 144 h. All the samples were analyzed for drug content using HPLC. As human skin from different donors was employed for different permeation studies, the in vitro permeation data was normalized using the flux obtained by repeating permeation of a previously evaluated patch, within every new permeation study.

167

Quantitative analysis

A UV detection based reverse phase HPLC method was used for quantitative analysis of TAF. Waters Alliance 2695 separation module (Milford, MA, USA) attached to a 2996 photodiode array detector was used. Isocratic elution was performed on

Phenomenex Luna 5µ C8 (2) 100A°, 250*4.6 mm (Phenomenex, CA, USA) at a flow rate of 1.0 mL/min and column temperature of 35 °C, after injecting 50 µL of sample. Mobile phase consisted of acetonitrile (phase A) and 0.1% v/v trifluoroacetic acid in DI water

(phase B) in the ratio of 30:70. The run time was 12 min and the retention time of TAF was around 7.4 min. Drug standards were prepared in the receptor solution and detected at wavelength of 262 nm. The precision limit of detection and quantification were 0.01

μg/mL and 0.03 μg/mL, respectively and linearity was observed in the concentration range of 0.1 - 50 µg/mL (R2 = 0.9999).

Data analysis

Microsoft Excel was used for analyzing data. Single factor analysis of variance

(ANOVA) and Student’s t-test were used for statistical analysis and p value of less than

0.05 was considered for concluding significant difference between the test groups.

Results and Discussion

Slide crystallization studies

Adhesiveness is a fundamental property of transdermal patches that is essentially required to ensure complete contact between the entire surface area of the patch and skin during the wear period, for efficient delivery of drugs. PSAs deform upon application of slight pressure, provide intimate contact with surfaces by establishing inter-atomic and molecular forces at the interface, and are thus, used for the preparation of transdermal

168

patches (Banerjee, Chattopadhyay, Ghosh, Datta, et al., 2014). A good adhesive is one that doesn’t leave any residue upon removal, is easy to use, stable to environmental changes, non-irritant and non-sensitive to skin, compatible with other formulation components, allows sufficient drug solubility, and possesses the necessary adhesive properties such as tack, shear, and skin adhesion (Pfennig, 1995; Sachdeva et al., 2013).

Acrylic, silicone, and PIB-based adhesives are most commonly used in the design of transdermal patches (Lobo et al., 2016). Therefore, all three types of adhesives (acrylate:

DURO-TAK® 87-2516, silicone: BIO-PSA 7-4301 and PIB: DURO-TAK® 87-6908) were explored for formulation development of transdermal patches of the antiretroviral drug.

Determination of saturation solubility of TAF in different adhesives was required to determine the maximum amount of dissolvable drug that could be incorporated in the adhesives. Slide crystallization is a preliminary and relatively fast method that is employed as an alternate to preparing complete patches for estimating the solubility of drugs in adhesives (Sachdeva et al., 2013). Images of pure TAF, as observed under the microscope at a magnification of 10x have been shown in Figure 29a. TAF at a concentration of 1% w/w (dry weight) was not soluble in silicone (BIO-PSA 7-4301) as well as PIB (DURO-TAK® 87-6908) adhesive, even before evaporation of solvents.

Therefore, due to poor solubility in these adhesives, further slide crystallization studies were not performed with them. However, drug blends at a concentration of 1-12% w/w

(dry weight) in DURO-TAK® 87-2516 were successfully prepared and drops of these were allowed to dry on glass slides and images of dried drug-adhesive blends as observed under the microscope, are shown in Figure 29 (b-f). TAF at concentrations of 3-12% w/w

169

was observed to crystallize immediately after drying (Figure 29b-d). However, 2% w/w drug blend crystallized after 3 days (Figure 29e) and 1% w/w blend did not crystallize even after 7 days at room temperature (Figure 29f). Therefore, the saturation solubility of the TAF in DURO-TAK® 87-2516 was observed to be between 1-2% w/w (dry weight).

Moreover, addition of 5% w/w oleic acid to DURO-TAK® 87-2516 adhesive enhanced the its solubility to about 4% w/w.

Figure 29. Images of slide crystallization studies, under optical microscope (a). Pure TAF (at 10x) (b-d). TAF crystals in DURO-TAK® 87-2516 immediately after evaporation of organic solvent at 10x: (b). 3% w/w TAF (c). 5% w/w TAF (d). 12% w/w TAF (e). 2% w/w TAF in DURO-TAK® 87-2516, 3 days after evaporation of organic solvent (10x) (f). 1% w/w TAF in DURO-TAK® 87-2516, 7 days after evaporation of organic solvent (20x (g). 4% w/w TAF in 5% w/w oleic acid in DURO-TAK® 87-2516 (10x) (h). 5% w/w TAF in 5% w/w oleic acid in DURO-TAK® 87-2516 (20x) (i). TAF:PVP 360 (9:1) under 20x (j). TAF:PVP 360 (8:2) under 10x (k).TAF:PVP 360 (7:3) under 20x (arrows depicting drug crystals)

170

The blend containing 5% w/w TAF in adhesive-oleic acid blend showed crystals after 7 days, whereas the 4% w/w TAF solution did not, as shown in Figure 29h and 29g, respectively. Further, in order to select the composition ratios of different crystallization inhibitors and TAF for patch formulations with 5%, 10%, and 15% w/w drug (higher than its solubility in adhesive alone), different ratios of the two components were tested in slide crystallization studies. The mixtures of TAF and PVP 360 in the ratios of 9:1 and

8:2, showed appearance of crystals after 7 days, but the compositions with TAF:PVP ratios from 7:3 to 1:9, did not show any crystals (Figure 29 i, j, k). All the other crystallization inhibitors (Kollidon® VA64, Kollidon® 90F, and Kollidon® 30LP) with

TAF, each in the ratio of 1:9, did not show appearance of any crystals after 7 days.

Therefore, ratios of 7:3 (TAF:PVP 360) and 9:1 (TAF: Kollidon® VA64 or Kollidon®

90F or Kollidon® 30LP), comprising of the lowest amount of PVP with the potential to inhibit crystallization, were selected for the preparation of patches containing 5% w/w or higher concentrations of drug.

Formulation of patches

Slide crystallization is a faster means for estimating saturation solubility of drugs in adhesives. However, the phenomenon of crystallization may be affected by thickness of the actual patches as well as processing conditions and scale up procedures. Therefore, based on the observations of slide crystallization studies, different transparent drug patches were prepared in DURO-TAK® 87-2516 adhesive as shown in Table 4 and observed for crystallization for 6 months under the microscope. Solubility of TAF in 5% w/w oleic acid in DURO-TAK® 87-2516 blend (dry weight basis) as determined by slide crystallization studies was found to be about 4%. Therefore, 2%, 3%, and 4% TAF

171

patches (200 gsm) were prepared in this matrix (AP1, AP2, AP3 as specified in Table 4).

Due to solubility limitations of TAF in the adhesive, it was not possible to prepare patches with higher drug concentrations in the absence of crystallization inhibiting agents that act as solubilizers and anti-nucleants. Use of different grades of PVP as crystallization inhibitors has been well-reported in literature and some of them were hence, investigated in the current study (Jain & Banga, 2013; Sachdeva et al., 2013;

Weng, Quan, Liu, Zhao, & Fang, 2016). In order to prepare 100 gsm patches containing higher percentage of TAF (5%, 10%, and 15% w/w- AP4, AP5, AP6, respectively), PVP

360 was added as a solubilizer and crystallization inhibitor, with TAF and PVP in the ratio of 7:3, in addition to oleic acid. Patch ‘AP10’ was prepared to investigate the effect of enhancing the percentage of PVP 360 as well as addition of 5% w/w propylene glycol in the 7.5% TAF patch on drug crystallization (100 gsm). Drug and PVP were in the ratio of 6:4 (w/w) in this formulation. Formulations AP7 to AP9 included addition of other grades of PVP (Kollidon® VA64, Kollidon® 90F, and Kollidon® 30LP) as solubilizers and crystallization inhibitors with TAF (10% w/w), in the ratio of 1:9 (PVP: drug) in oleic acid and DURO-TAK® 87-2516 blend. Patch AP11 (100 gsm) was prepared to investigate the effect of enhancing the percentage of Kollidon® 30LP as well as addition of 5% w/w propylene glycol in the 7.5% drug patch on its crystallization. TAF and

Kollidon® 30LP were in the ratio of 7:3 (w/w) in this formulation. Further, patches AP12 to AP14 (400 gsm) containing 2% w/w TAF were prepared to evaluate the effect of combination of permeation enhancers (5% w/w oleic acid+ 5% w/w propylene glycol +

5% w/w oleyl alcohol, 5% w/w oleic acid+ 5% w/w propylene glycol + 5% w/w triacetin,

5% w/w oleic acid+ 5% w/w propylene glycol + 8.5% w/w octisalate, respectively) as

172

well as increasing coat weight on drug permeation and compare it with that of AP1.

Further, AP15 (400 gsm) patch was prepared by replacing 5% w/w oleic acid in AP12 with 5% oleyl alcohol, such that total oleyl alcohol content was 10% w/w in this formulation. The transdermal permeation enhancing effects of oleic acid (Mitragotri,

2000; Pathan & Setty, 2009; Saini et al., 2014), oleyl alcohol (Andega et al., 2001;

Santoyo et al., 1995), triacetin (Moghimipour, Rezaee, & Omidi, 2011; Nicoli et al.,

2006; Patel et al., 2012), and octisalate (Fraser et al., 2007; Nicolazzo, Morgan, Reed, &

Finnin, 2005) are well-known and have been reported in literature and thus, these chemical enhancers were selected for the formulation development of drug patches.

Microscopy and stability assessment of patches

During the storage period, changes in temperature can alter the thermodynamic activity of the transdermal patch formulation, and result in precipitation or crystallization or changes in the crystal habit of the active ingredient. Therefore, during stability testing, patches are exposed to stress and real world storage conditions that are representative of the product’s proposed marketing conditions (European Meedicines Agency, 2014). The stability of the acrylate-based drug patches was assessed in terms of occurrence of crystallization over a period of time at different temperature conditions. The microscopic observations of the same after 6 months have been shown in Table 5. Representative images of TAF crystals in transdermal patches have been shown in Figure 30.

173

Figure 30. TAF crystals in acrylate-based transdermal patches (at 10x) (a). Patch AP5 (crystals at 40 °C after 2 weeks) (b). Patch AP9 (crystals at room temperature in 12 days) (c). Patch AP8 (crystals at room temperature in first week)

Table 5. Microscopy observations of acrylate-based TAF containing transdermal patches for 6 months

TEMPERATURE PATCH -20 ° C ROOM TEMPERATURE 40 °C CODES AP1 NO NO NO AP2 NO Crystals after three weeks Crystals after three weeks AP3 NO Crystals after two weeks Crystals after two weeks AP4 NO Crystals after three weeks Crystals after three weeks AP5 NO Crystals after two weeks Crystals after two weeks Crystals after 2 AP6 Crystals after 9 days Crystals after one week months AP7 Not observed Crystals in first week Crystals in first week AP8 Not observed Crystals in first week Crystals in first week Smaller crystals than AP7 and AP8 Crystals in first week (2-3 AP9 Not observed after 12 days crystals)

174

TABLE 5 (Continued)

AP10 NO Crystals after three weeks Crystals after two weeks AP11 NO Crystals after three weeks Crystals after two weeks AP12 NO NO NO AP13 NO NO NO AP14 NO NO NO AP15 NO NO NO

Patches with 2% w/w TAF were found to be stable and did not show crystals at all the tested temperature conditions, even after 6 months, as the drug concentrations were within the saturation solubility limits in these matrices. However, in order to incorporate drugs in amounts higher than its saturation solubility in the adhesive alone, crystallization inhibitors were included in the patches. Mechanistically, these additives stabilize the systems against crystallization, either due to enhancing the solubility of the drug or by aiding adsorption of drug crystals in the adhesive matrix. PVP has previously been demonstrated as one of the most effective additives in inhibiting drug crystallization in patch formulations acting by the formation of amorphous co-precipitates with the drugs

(K. Gong et al., 2005; Jain & Banga, 2013; Ohm, 2000; Sekikawa et al., 1978). Also, it has been previously shown that PVP inhibits crystallization of drugs by getting adsorbed on the growing crystal surfaces (Jain & Banga, 2013; Ma, Taw, & Chiang, 1996).

Therefore, different grades of PVP were incorporated in patches containing higher concentrations of drug. However, patches with 15% w/w TAF (AP6) and PVP 360, crystallized in less than ten days at room temperature and 40 °C, depicting the inability of

PVP 360 to inhibit crystallization at high drug levels. At lower drug concentrations (5% and 10% w/w drug, AP4 and AP5, respectively), addition of PVP 360 in the same ratio

175

with drug (3:7, w/w) as in AP6, slowed down the process of crystallization, where it was observed after three and two weeks, respectively, at room temperature as well as 40 °C.

Further, increasing PVP:drug ratio to 4:6 (w/w) and adding propylene glycol as additional solubilizer in patch AP10 formulation did not delay the process of crystallization further and crystals were still observed in less than a month as well at room temperature and 40 °C. Interestingly, these patches did not show presence of any crystals even after 6 months at -20 °C. Further, other crystallization inhibitors (Kollidon®

VA64, Kollidon® 90F) were found to be ineffective, as crystals were visible during the first week after storage of all these patches at room temperature and 40 °C. However, the crystals observed in patch AP9 (containing Kollidon® 30LP) were smaller than AP7 and

AP8 and were observed after about two weeks at room temperature and after first week at

40 °C. Therefore, patch AP11 with reduced TAF concentration (7.5% w/w drug) and higher amount of Kollidon® 30LP than AP9 as well as with addition of propylene glycol was formulated and it showed slower appearance of crystals (observed after three weeks at room temperature). However, the inability of PVP to prevent appearance of drug crystals in the transdermal patch formulation for longer time duration, may be attributed to the difficulty in the molecular interactions between PVP and growing drug crystals because of the higher viscosity of the patch formulation as compared to a single drop of blend on the slides (Jain & Banga, 2013). Overall, though most of the studies showed

PVP as an effective drug crystallization inhibitor, there are a few findings regarding ineffectiveness of some grades of PVP. For example, Weng et al. reported that PVP K30 was found to be ineffective in inhibiting crystallization of risperidone in drug-in-acrylate adhesive transdermal patches. However, in the same study, oleic acid was found to be an

176

effective crystallization inhibitor (Weng et al., 2016). In this study, all the compositions of 2% w/w TAF transdermal patches were found to be stable at all the temperature conditions for more than 6 months, which may be attributed to drug solubility in the adhesive matrix being lower than the saturation level as well as the crystallization inhibiting effect conferred by oleic acid. Additionally, AP2, AP3, AP4, AP5, AP10,

AP11 transdermal patch formulations with higher TAF concentrations were stable only at

-20 °C.

Coat weight and drug content

Coat weight and drug content of the patches evaluated for the in vitro permeation studies were measured and have been reported in Table 6. These measured properties are influenced by various factors such as percentage of non-volatile components in the adhesive blend, scale set-up, surface level of the casting knife, textural properties of release liner and backing membrane, and are reflective of the coating efficiency of the patches. Results showed that all the patches had uniformity in both coat weight and drug assay.

Table 6. Coat weight and drug content of acrylate-based clear patches

PATCH COAT WEIGHT (mg/cm2), DRUG CONTENT (mg/0.64 cm2), FORMULATION CODE mean ± SD mean ± SD Theoretical Experimental Theoretical Experimental AP1 20 18.20 ± 1.88 0.256 0.25 ± 0.02 AP5 10 7.10 ± 0.14 0.640 0.53 ± 0.05 AP12 40 37.37 ± 6.94 0.512 0.478 ± 0.09 AP13 40 40.80 ± 1.47 0.512 0.522 ± 0.02 AP14 40 44.17 ± 4.96 0.512 0.565 ± 0.06 AP15 40 40.43 ± 2.35 0.512 0.643 ± 0.106

177

Also, the values of theoretical and experimental TAF content were close for all the patches that confirmed its stability after exposure to the temperature conditions used for the processing of respective patches.

In vitro permeation studies

Epidermal integrity and thickness assessment. Integrity of human epidermis samples must be evaluated prior to permeation studies, as the procedures employed for procurement of human skin such as surgical removal, dermatoming, storage as well as technical procedure for separation of epidermis from dermatomed skin, can damage the skin and thus, ultimately influence drug permeation. The physical integrity of stratum corneum has been reported to be indicated by its electrical properties. Therefore, measurement of electrical conductivity is used as a means of assessing the barrier integrity for full thickness skin as well as epidermal membranes (Cai, Söderkvist,

Engqvist, & Bredenberg, 2012; Lawrence, 1997). Hence, for selection of epidermal pieces with optimum barrier integrity, electrical resistance of skin was measured and considered as the selection criteria. Epidermal pieces with the resistance above 10 kΩ, were selected for the study. Further, the thickness of human epidermis used for the permeation studies ranged from 50-150 µm.

Selection of receptor solution and in vitro drug permeation studies. The percentage degradation of TAF in different receptor solutions at 37 °C has been shown in

Table 7. It was evident that TAF degrades in 1X PBS, with or without PEG 400.

However, lowering the pH to 6.0 and using non-aqueous media such as pure PEG 400, reduced drug degradation. However, due to limitations such as high viscosity as well as low drug sensitivity in HPLC, PEG 400 was not selected as the receptor media.

178

Phosphate buffer, pH 6.0 was thus, selected as the receptor solution and it was completely replaced every 24 h during the 7-day skin permeation studies. Degradation of

TAF at both acidic and basic pH has been previously reported in the literature. Also, it has been found to be stable at pH of around 5-6.8. In addition, solubility of TAF in phosphate buffer, pH 6.0 was found to be about 4.95 mg/mL. Experimental conditions ensured maintenance of sink conditions for our in vitro permeation studies.

Table 7. TAF stability in different receptor solutions

TAF concentration % Drug degradation in 24 h at 37 °C (µg/mL) PBS, pH Phosphate buffer, PEG 400: PBS, pH PEG 400 7.4 pH 6.0 7.4 (1:1) 0.5 22.28 3.65 41.85 Not detected due to low LOD 5 28.29 1.61 32.40 0.00 50 30.54 2.27 33.84 0.85

In vitro permeation studies. The different patches that were evaluated in seven day in vitro permeation studies have been described in Table 6. The amount of TAF that permeated from patch AP1 (simplest matrix comprising of 2% w/w drug dissolved in mixture of oleic acid in acrylate adhesive) after 7 days was found to be 19.20 ± 2.92 µg/cm2.

The average permeation flux over a duration of 7 days was calculated to be 0.12 ± 0.01

µg/cm2/h. To evaluate the effect of TAF concentration in the patch on its skin permeation, patch AP5 with 10% w/w, drug that was found to be stable for about two weeks at room temperature, was selected for in vitro permeation study. As shown in Figure 31, increasing

TAF concentration to 10% w/w (AP5), significantly enhanced its permeation to 137.76 ±

179

10.42 µg/cm2 after 7 days and flux rate was calculated to be 0.88 ± 0.07 µg/cm2/h as compared to that of AP1 (p<0.05).

Figure 31. Permeation profiles of TAF from acrylate-based patches across human epidermis *represents statistical significant difference, Student’s t test (p<0.05)

However, since AP5 as well as other acrylate adhesive-based patches with TAF concentration of more than 2% (w/w), crystallized at room temperature over a period of time, transdermal patches with 2% w/w TAF and consisting of combination of chemical penetration enhancers as well as higher coat weight, were prepared and evaluated for skin permeation testing (AP12 to AP15).

As shown in Figure 32a, the cumulative amount of TAF permeation after 7 days in the patches AP12 (96.40 ± 14.83 µg/cm2), AP13 (87.15 ± 8.60 µg/cm2), and AP14

(89.56 ± 17.28 µg/cm2) was found to be significantly higher than that of AP1 (19.20 ±

2.92 µg/cm2, p<0.05). Also, the permeation flux observed from the three aforementioned

180

groups was 0.60 ± 0.10 µg/cm2/h, 0.54 ± 0.05 µg/cm2/h, and 0.56 ± 0.12 µg/cm2/h, respectively, which was about five-fold greater than that observed from AP1 (0.12 ±

0.01 µg/cm2/h). Thus, increasing the patch coat weight to 400 gsm from 200 gsm along with incorporation of combination of chemical enhancers (oleic acid+ propylene glycol+ oleyl alcohol, oleic acid+ propylene glycol+ triacetin, or oleic acid+ propylene glycol + octisalate) resulted in significant enhancement in drug permeation. However, no significant difference in drug permeation between AP12, AP13, and AP14 was observed.

As shown in Figure 32b, the amount of TAF permeation over 6 h was significantly higher in patch AP12 (6.15 ± 0.41 µg/cm2) than AP13 (3.38 ± 1.40 µg/cm2) as well as AP14 (3.17 ± 1.30 µg/cm2, p<0.05). Patch AP15 was formulated by replacing

5% w/w oleic acid in AP12 with oleyl alcohol, thus, having total of 10% w/w oleyl alcohol and a coat weight of 400 gsm. However, as shown in Figure 33, there was no significant difference between the amount of TAF permeation after 7 days between the two groups (AP12: 96.40 ± 14.83 µg/cm2 and AP15: 122.69±28.34 µg/cm2, p>0.05).

181

182

Figure 33. Permeation profile of acrylate-based TAF patches comprising of oleyl alcohol and comparison with the control patch through human epidermis * represents statistical significant difference, ANOVA one way test (p<0.05)

The effects of various chemicals used in the patches such as oleic acid

(Mitragotri, 2000; Pathan & Setty, 2009; Saini et al., 2014), oleyl alcohol (Andega et al.,

2001; Santoyo et al., 1995), triacetin (Moghimipour et al., 2011; Nicoli et al., 2006; Patel et al., 2012), and octisalate (Fraser et al., 2007; Nicolazzo et al., 2005) on skin permeation enhancement are well-known. Synergistic permeation enhancing effect of combination of propylene glycol and oleic acid has also been reported earlier. In our study, propylene glycol is expected to act as a cosolvent and thus, enhance the concentrations of the drug substance as well as the enhancer in the stratum corneum.

Oleic acid, on the other hand, has the potential to delipidize the stratum corneum and therefore, facilitate the partitioning of both, the drug molecules and propylene glycol into skin (H. Benson, 2005; Chatterjee, Li, & Koda, 1997; Mitragotri, 2000). Also, the

183

synergistic effect of propylene glycol and oleyl alcohol was found to markedly increase the skin permeation of tenoxicam across hairless mouse skin (Chatterjee et al., 1997).

Furthermore, a combination use of propylene glycol and oleyl alcohol, under occlusive conditions has been reported to result in considerable enhancement in the permeation of testosterone across full thickness neonatal porcine skin. The mechanism of penetration enhancement by octisalate is not fully understood. However, due to the structural similarity as well as similar observations in differential scanning calorimetry and attenuated total reflectance spectroscopy, in terms of the effect of delipidization on stratum corneum, octisalate is assumed to work mechanistically in the similar way to azone (Nicolazzo et al., 2005). Additionally, octisalate has been incorporated as a penetration enhancer in sunscreens, cosmetic products, as well as in metered spray transdermal systems (Fraser et al., 2007). Further, triacetin has been incorporated as a penetration enhancer in the commercially available Oxytrol® patch for overactive bladder

(Nicoli et al., 2006; Quan & Deshpanday, 2016). Patel et al also reported enhancement in diclofenac permeation through human cadaver skin using 10% triacetin (Patel et al.,

2012).

Overall, even after addition of a combination of different enhancers as well as after increasing the coat weight of the patch, the desired transdermal permeation flux of 7

µg/cm2/h for HIV prophylaxis could not be achieved for TAF with the acrylate-based clear patches.

Conclusion

Transdermal patches containing 2% w/w TAF, oleic acid, propylene glycol, and oleyl alcohol or triacetin or octisalate in DURO-TAK® 387-2516 were found to be stable

184

at room temperature for six months. Addition of a combination of chemical enhancers and increasing the patch coat weight showed significant enhancement in the drug permeation as compared to the control. The present study thus, demonstrated feasibility of transdermal delivery of a potential HIV prophylactic agent. However, the required transdermal permeation flux of 7 µg/cm2/h could not be achieved with the clear acrylate- based drug patches. Therefore, different patch design strategies were investigated to achieve the prophylactically relevant dose of TAF for HIV and have been described in detail in the following chapter.

185

CHAPTER 8

DEVELOPMENT OF SUSPENSION-BASED TRANSDERMAL PATCH OF TENOFOVIR ALAFENAMIDE

Abstract

Achieving the targeted transdermal permeation flux of TAF for HIV prophylaxis was not feasible with the formulation of stable, clear acrylate-based transdermal patches as described in chapter 7, due to limitations of drug solubility in adhesive formulation as well as issues of drug crystallization. Thus, the aim of this study was to demonstrate the feasibility of developing a suspension-type patch of TAF for transdermal delivery of its prophylactically relevant dose for HIV (7 µg/cm2/h).

Suspension patches containing different concentrations of TAF (5-25% w/w) were prepared using silicone adhesive (BIO-PSA 7-4301). 3M™ Scotchpak™1022 and silicone coated poly-ethylene terephthalate films were used as the release liner and backing membrane, respectively. In vitro permeation studies of the prepared patches through human epidermis were conducted using vertical Franz diffusion cells for 7 days.

The optimized patch was evaluated for stability, drug release, drug content, coat weight, thickness, tack properties, peel adhesion, and skin irritation potential as well.

The cumulative amount of TAF that permeated after 7 days was found to be

191.51 ± 23.26, 695.09 ± 68.85, 1028.69 ± 78.63, 1158.86 ± 57.96, and 1266.19 ± 162.04

µg/cm2 from patches containing 5, 10, 15, 20, and 25% w/w drug, respectively.

185 186

Significant increase in permeation was observed with the increase in TAF concentration from 5 to 10% w/w as well as from 10 to 15% w/w (p<0.05), but not between patches containing 15, 20, and 25% w/w drug (p>0.05). Therefore, 15% drug concentration was found to be optimum in order to achieve the targeted permeation flux. The optimized patch demonstrated uniformity in drug release profile as well as other characteristics such as drug content, coat weight, and thickness, peel adhesion, and tack properties after 3 months of storage at ambient and higher temperature conditions. The average peel adhesion and absolute positive force (depicting the tack properties) were determined to be

202.78 ± 43.07 g and 1116.88  167.79 g/cm2, respectively. The final silicone-based suspension patch was found to be non-irritant to human skin.

A stable silicone adhesive-based suspension-type transdermal patch for delivery of prophylactically relevant dose of TAF, suitable for weekly drug dosing was thus, successfully developed.

Introduction

The requirement of an effective drug delivery system to be used for an intermediate duration such as transdermal patches, for the purpose of HIV prophylaxis, has been elaborated in chapter 2 and 7. In order to develop the same, efforts were made to prepare the conventional, transparent patches containing drug in the dissolved form as described in chapter 7. However, due to low solubility of TAF in the acrylate adhesive and its high tendency to crystallize out when dissolved, it was challenging to deliver the prophylactic dose by the conventional patch design. Many crystallization inhibitors and enhancers were incorporated in the patches, but our investigation revealed that it was feasible to achieve only 0.6 μg/cm2/h flux and we could load only 2% TAF in the same,

187

with no issues of drug crystallization. However, the latter was not sufficient to deliver the required dose of the drug across skin. Therefore, suspension-type patch design was explored as it would enable much higher drug loading than the clear patches. Adhesives, such as silicone and poly isobutylene that showed significantly less solubility of TAF, could be employed for this patch design formulation. Moreover, low solubility of drug in the adhesives would reduce the chances of drug crystallization as well. Also, the excess amount of drug would provide a constant thermodynamic activity/driving force for its permeation. Further, acrylate patches have more potential of causing skin irritation than silicone or poly isobutylene ones. Due to all these potential advantages, suspension patches of TAF were formulated and investigated for its permeation across human epidermis.

Feasibility of developing a silicone-based suspension patch for fentanyl has been demonstrated in literature (Kenneth Miller, Govil, & Bhatia, 2008). The patent also claims the silicone-based suspension patch to provide improved adhesion, better transdermal release and delivery of fentanyl as compared to the solution-based clear matrix patches (Kenneth Miller et al., 2008). Thus, suspension patch of TAF investigated in this study was formulated on similar lines, but with utilization of different procedure as well as excipients (plasticizer and enhancers).

In this study, silicone and PIB based suspension patches of various compositions were prepared and evaluated for in vitro permeation of TAF across human epidermis for

7 days. Considering delivery of a dose of 8 mg/day from a 50 cm2 transdermal patch, targeted permeation flux was about 7 µg/cm2/h. Different chemical enhancers and plasticizing agent was included in the patch formulations. Patches were evaluated for in

188

vitro release, drug content, stability, skin irritation, and physical characteristics such as tack and peel adhesion.

Materials

TAF was purchased from Pharmacodia Co., LTD (Beijing, China). Backing membranes (ScotchpakTM 9733-2.05 mil polyester film; CoTran™ 9702, 9706, 9728- ethylene vinyl acetate copolymer (EVAC); CoTran™ 9718, 9720 – polyethylene film) and release liner (ScotchpakTM 1022- 3 mil fluoropolymer coated polyester film) were gifted by 3M (St. Paul, MN, USA). Silicone coated poly-ethylene terephthalate (PET) films (1 mil): 48101- 4400B/000, 44916 –7300 AM/000, and 40987- U4162/000 were procured from Loparex (Cary, NC, USA). PIB adhesive (DURO-TAK® 87-6908) was obtained as gift sample from Henkel Corporation (Dusseldorf, Germany). Silicone adhesive (BIO-PSA 7-4301) was also provided as gift sample by Dow Corning

Corporation (Washington, DC, USA). Mineral oil, oleyl alcohol, and polyvinyl pyrrolidone (PVP 360) were obtained from Sigma–Aldrich (St. Louis, MO, USA). Oleic acid was purchased from Ekichem (Joliet, IL, USA) and Croda Inc. (Edison, NJ, USA), respectively. Gentamycin sulfate, tetra hydrogen furan (THF), sodium dihydrogen phosphate, sodium hydroxide, and 10X phosphate buffered saline, pH 7.4 were obtained from Fisher Scientific (Pittsburgh, PA, USA). Methanol, ethanol, acetonitrile, and trifluoroacetic acid were purchased from Pharmco-aaper (Brookfield, CT, USA).

Dermatomed human cadaver skin was obtained from New York Fire Fighters (New

York, NY, USA).

189

Methods

Formulation of suspension patches

Suspension patches were prepared using silicone (BIO-PSA 7-4301) and PIB

(DURO-TAK® 87-6908) adhesive. TAF was first levigated with mineral oil and other additives (oleic acid and oleyl alcohol) were then added to it and mixed using

BenchmixerTM (Benchmark Scientific Inc., Edison, NJ, USA). Heptane was then added to the blend and the drug suspension was homogenized using a high-speed homogenizer

(OmniTHQ, Omni International, NW, GA, USA). Heptane was then allowed to evaporate at 90° C for 5 min and the adhesive was added to the remaining mixture. The final blend was allowed to mix overnight, then homogenized for 1 min, and cast on the release liner. It was dried in a flameproof oven and the cast film was then laminated with the backing membrane. The compositions, homogenization conditions, release liners, backing membranes, and drying conditions employed for the formulation of different

TAF containing suspension patches have been elaborated in Table 8.

Visual observations of the suspension patches

The suspension patches with varying composition and material components were prepared and observed for the following visual changes for about two weeks: phase separation, contraction/shrinkage of the film, residue on release liner after peeling, ease of peeling off the patches applied on human skin as well as any residue on skin after removal of the patches.

190

Table 8. Composition and manufacturing parameters of the silicone-based TAF containing suspension transdermal patches

PATCH FORMULATION CODES PROPER SP1 SP2 SP3 SP4 SP5 SP6 SP7 SP8 SP9 SP SP SP SP SP14 TIES 10 11 12 13 Coat weight 35 35 20 20 100 250 250 250 250 250 300 300 300 50 (~gsm) 0 0 0 0

Release Fluoro- Fluoro- liner polymer polymer coated coated side of Uncoated side of ScotchpakTM 1022 side of Scotchp Scotchp akTM akTM 1022 1022 Backing Silic Silic Silic membrane one one one Scotchp Scotchp CoTran™ Scotchp CoTran™ CoTran™ CoTran™ coate coate coate akTM akTM 9702, Silicone coated PET akTM 9720 9718 9718 d d d 9733 9733 9706, film 9733 (poly- (poly- (polyethyl PET PET PET (polyest (poly- 9728 (48101) (polyest ethylene) ethylene) ene) film film film er) ester) (EVAC) er) (481 (449 (409 01) 16) 87) Homogeni zation speed (rpm) and 5000, 3 30000, 20 32000, 30 5000, 3 time (min) 32000, 5 before addition of adhesive Homogeni zation speed 5000, 1 30000, 1 32000, 1 15000, 1 5000, 1 (rpm) and time (min) after

191 TABLE 8 (Continued) addition of adhesive

78 °C Air drying for 5 min followed by drying at 78 °C for 15 min 78 °C Drying for 10 Air drying, 15 min for 10 conditions min min Components (% Dry weight, w/w)

TAF 15 15 15 15 15 15 15 15 15 5 10 20 25 15 Oleic acid 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Oleyl - 10 10 10 10 10 10 10 10 10 10 10 10 - alcohol Mineral 14 14 14 14 14 14 14 14 14 14 14 14 14 14 oil BIO-PSA 66 56 56 56 56 56 56 56 56 66 61 51 46 - 7-4301 DURO------66 TAK® 87- 6908

192

Effect of homogenization on particle size of TAF

Particle size of TAF, before and after homogenization was determined using optical microscopy. For the former, pure TAF was dispersed in heptane and a drop of the same was placed on a glass slide and observed under optical microscope. For the size of drug particles after homogenization, a drop of the suspension, before addition of the adhesive, was mounted on glass slide and observed under the optical microscope. Size of particles (n=50) was measured using ImageJ 1.41o software (National Institutes of

Health, USA) and has been reported as average ± SD.

Coat weight and drug content of the patches

Coat weight of the prepared patches was determined by punching and weighing

4.91 cm2 or 1 cm2 laminates (from different areas of the patch, n=3) using analytical balance (Mettler Toledo, Columbus, OH, USA) and subtracting the weight of equal sized respective release liner and backing membrane from the same. For determination of drug content in the patches, the punched patches were placed in 10 mL of THF after removal of the release liner and allowed to shake overnight. The solutions were then diluted ten times with methanol and centrifuged at 13,400 rpm for 10 min. The supernatants were then analyzed using HPLC to quantitate the amount of TAF in the same. Stability of TAF after exposure to high temperature conditions used for processing, was indicated by comparing the experimental and theoretical drug content values. Results have been reported as average ± SD.

In vitro skin permeation studies

In vitro drug permeation studies from transdermal patches through human epidermis were performed for 7 days using in vitro vertical Franz diffusion cells

193

(PermeGear, Inc., Hellertown, PA, USA), providing an effective diffusion area of 0.64 cm2 (n=6).

Separation of epidermis. For each permeation study, human epidermis was freshly isolated from dermatomed human skin by heat-separation method. Skin was immersed in

1X PBS (10 mM phosphate ions, 137 mM sodium chloride, and 2.7 mM potassium chloride pH 7.4) for 2 min at 60° C. Epidermis was then carefully peeled off manually, with the help of forceps and spatula and thereafter, cut into pieces of suitable size for mounting on the Franz cells (Kassis & Søndergaard, 1982).

Technologies, CA, USA) as well as an arbitrary waveform generator (Agilent 33220A,

20 MHz Function). Epidermis was mounted on the Franz diffusion cells and left for equilibration for 15 min. Phosphate buffer, pH 6.0 was then added in the donor (300 µL) and receptor (5 mL), respectively. Following equilibration, silver and silver chloride electrodes were placed in the receptor and donor compartment, respectively. Load resistor (RL) attachment was in series with skin, and the voltage drop across the entire circuit and skin (VS) was displayed on the multimeter (VO). Following formula was used to measure the skin resistance (RS):

Rs = Vs RL/ (Vo - Vs)

Where, VO and RL were 100 mV and 100 kΩ, respectively. Epidermis pieces with electrical resistance of more than 10 kΩ were selected for the permeation study. The

194

thickness of the selected epidermis pieces was measured using a thickness gauge

(Cedarhurst, NY, USA). .

In vitro skin permeation set up. Permeation of TAF from different transdermal patch formulations (donor) mentioned in Table 10, through human epidermis was investigated using the Franz diffusion cell set up. Taking into consideration drug stability as well as for maintenance of sink conditions, phosphate buffer, pH 6.0 having gentamycin sulfate (80 mg/L) as an anti-microbial agent for the 7 day permeation study

(Sachdeva et al., 2013), was selected as the receptor compartment (as described in chapter 7). The temperature of the receptor phase was maintained at 37 °C and it was constantly stirred at 600 rpm. The freshly isolated epidermis pieces were clamped between the donor and receptor compartments. After measuring the skin resistance using the procedure described above, phosphate buffer from the donor was pipetted out, epidermis was removed, placed flat on a glass plate, and dried with the help of Kimwipes.

Release liner of the patches (of size enough to cover the diffusion area) were removed and the latter were applied carefully on the dried epidermis such that the adhesive layer of the patch was adhered to the stratum corneum side of the skin. Glass rod was rolled on the patches to ensure their adherence to the skin. The epidermis pieces with the applied patches were then mounted between the donor and receptor compartment and the entire set up was secured in place using a clamp. Receptor (0.3 mL) was sampled at 0, 2, 4, 6, 8 h with replacement with fresh receptor solution and entire receptor (5 mL) was removed and replaced with fresh buffer at 24, 48, 72, 96, 120, 144 h. All the samples were analyzed for TAF content using HPLC. As human skin from different donors was employed for the different permeation studies, the in vitro permeation data was

195

normalized using the flux obtained by repeating permeation of a previously evaluated patch, within every new permeation study.

In vitro drug release studies

Release profile of TAF from the optimized patch (SP7) was evaluated using

Paddle over Disk, USP V dissolution bath apparatus (Sotax AT7 Smart, Sotax AG,

Switzerland) (Cai et al., 2012; Lakhani, Rishabh, & Bafna, 2015). This study was performed during week 1 and after 1.5 and 3 months of patch preparation and storage at room temperature and 40 °C. The patches (0.712 cm2) were applied on the teflon mesh and placed on the glass discs, such that the release surface was facing upwards (n=3). The distance between the paddle and surface of the disk was about 25 mm. The disks were placed at the bottom of the vessels containing 500 mL of phosphate buffer (pH 6) as the receptor media. Temperature of the media was set at 32± 0.5°C and the paddle speed was

100 rpm. The samples (1 mL) were drawn at 0, 2, 4, 6, 8 h with replacement with fresh media and the entire receptor (500 mL) was removed and replaced with fresh buffer at

24, 48, 72, 96, 120, 144 h. All the samples were analyzed for TAF content using HPLC.

Coat thickness, coat weight, and drug content of optimized patch

Coat thickness of the 1 cm2 laminates of SP7 (from different areas of the patch) was determined by measuring the patch thickness using a thickness gauge (Cedarhurst,

NY, USA), and subtracting the thickness of the release liner and backing membrane from the former. Coat weight and drug content of the optimized patch was determined using the procedure described above under the heading “Coat weight and drug content of the patches”. These parameters were determined at day 1 and after 1.5 and 3 months of storage of patch at room temperature and 40 °C (n=4-6).

196

Quantitative analysis

A UV detection based reverse phase HPLC method was used for quantitative analysis of the drug as described in chapter 7. Waters Alliance 2695 separation module

(Milford, MA, USA) attached to a 2996 photodiode array detector was used. Isocratic elution was performed on Phenomenex Luna 5 µ C8 (2) 100A, 250*4.6 mm

(Phenomenex, CA, USA) at a flow rate of 1.0 mL/min and column temperature of 35 °C, after injecting 50 µL of sample. The mobile phase consisted of acetonitrile (phase A) and

0.1% v/v trifluoroacetic acid in DI water (phase B) in the ratio of 30:70. The run time was 12 min and the retention time of the drug was around 7.4 min. Drug standards were prepared in the receptor solution and detected at wavelength of 262 nm. The precision limit of detection and quantification were 0.01 μg/mL and 0.03 μg/mL, respectively and linearity was observed in the concentration range of 0.1 - 50 µg/mL (R2 = 0.9999).

Physical characterizations of optimized patch

Peel adhesion. The 180° peel adhesion tester (ChemInstruments, Fairfield, OH, USA) was used to evaluate the peel adhesion force (required to peel away the transdermal patch from dermatomed human cadaver skin) (Wokovich, Prodduturi, Doub, Hussain, & Buhse,

2006). Prior to running the test patches, the instrument was calibrated for experimental parameters such as tension, speed, and peel length with a load cell weighing 50 g.

Transdermal patches with the length and width of 3.0 inches and 1.0 inch, respectively, were cut. One end of the adhesive coated backing membrane was adhered to the human skin affixed on the stainless steel plate and the other end was attached to load cell grip. The average force required to peel off the adhesive film was measured and recorded during the

197

first week after patch preparation at room temperature and after 3 months of its storage at room temperature and 40 °C (n=3) as well. Results have been shown as average ± SD.

Tack properties. A Texture analyzer (Texture Technologies Corp, Marietta, GA,

USA) consisting of a 7 mm probe was used to determine the adhesion efficiency of the patch (Z. Lu & Fassihi, 2015). Prior to running the test samples, the instrument was calibrated and parameters such as target force, approach speed, return speed as well as distance and hold time were optimized. A transdermal patch (width and length of about

1.0 inch x 1.0 inch, respectively) was cut and adhered on to the sample holder after removal of release liner. As the test run was initiated, the probe was allowed to touch the adhesive film with a target force and hold time of 50 g and 10 s, respectively. This resulted in creation of a bond between the probe surface and transdermal patch. Further, as the probe was pulled off, it resulted in debonding between the two surfaces.

Parameters such as work of adhesion, positive area, and separation distance were hence, recorded during the first week after patch preparation at room temperature and after 3 months of storage at RT and 40 °C (n=3). Results have been shown as average ± SD.

Evaluation of skin irritation potential of optimized patch

An in vitro EpiDermTM skin irritation test (EPI-200-SIT) with a 3D in vitro reconstructed epidermis (RhE) model (MatTek Corporation 200 Homer Ave, Ashland,

MA 01721) was employed to assess the cell viability of skin using methyl thiazolyl tetrazolium viability assay (Kandárová, Hayden, Klausner, Kubilus, & Sheasgreen, 2009) after application of patch SP7. Three replicates of the patch were tested and compared to a negative control (Dulbecco’s phosphate buffered saline) and positive control (5% aqueous sodium dodecyl sulfate solution). After removal of the release liners, patches

198

were adhered to the tissues and kept in an incubator for 1 h at 37 °C. Next, the patches were removed from the surface of tissue inserts, the inserts were washed using 1X PBS and transferred to a fresh assay medium for 24 h incubation. The media for the inserts was exchanged and incubated again for 18 h. This was followed by transferring the inserts into yellow methyl thiazolyl tetrazolium solution and incubation for 3 h. During the 3 h incubation, mitochondrial metabolism was expected to occur and was detected by the formation of a purple-blue formazan salt. The plate with the inserts were filled with isopropyl alcohol (2 mL) and kept on a shaker at 120 rpm for 2 -3 h. The aliquots were then transferred to a 96 well enzyme linked immunosorbent assay (ELISA) plate and the optical density of the extracted formazan salt was measured at 560 nm using a Synergy high throughput plate reader (BioTek Instruments, Inc, Winooski, VT, USA).

Data analysis

Microsoft Excel was used for analyzing data. Single factor analysis of variance

(ANOVA) and Student’s t-test were used for statistical analysis and p value of less than

0.05 was considered for concluding significant difference between the test groups.

Results and Discussion

Formulation of suspension patches

Due to low solubility of TAF in silicone and PIB adhesives, it was suspended in the blends of these adhesives with different additives as elaborated in Table 8. Mineral oil was added as a plasticizer that increases diffusivity of the drug by decreasing the resistance offered by the patch matrix (Sachdeva et al., 2013). In addition to plasticizing effect, mineral oil has also been reported as a transdermal permeation enhancer (Karande

& Mitragotri, 2009). Oleic acid and oleyl alcohol are also well-known chemical

199

penetration enhancers (Andega et al., 2001; Mitragotri, 2000; Pathan & Setty, 2009; Saini et al., 2014; Santoyo et al., 1995) that were included in the suspension patches. A number of different silicone-based drug containing patches were prepared in order to select the optimal type of the release liner and backing membrane. Homogenization speed and time were also optimized to reduce the particle size of TAF.

Patch SP1 (~100 gsm) was prepared using the fluoropolymer coated side of

ScotchpakTM 1022 and ScotchpakTM 9733 (polyester backing). For patch formulations,

SP2-SP13, higher coat weight (~200-350 gsm) was applied and oleyl alcohol was included as an additional enhancer. For the preparation of SP2 patch, same release liner and backing membrane as SP1 were initially selected. However, while formulating patch formulation with oleyl alcohol, it was found that when the formulation was coated on the fluoropolymer coated side of the liner, it could not come off the release liner, depicting the affinity of the formulation binding to the fluoropolymer coated release liner was more than polyester backing membrane. Therefore, the formulation was cast on the uncoated side of the release liner (polyester only) and then laminated using the polyester

(ScotchpakTM 9733) backing film. This was an interesting experimental observation as silicone based transdermal patches are usually cast on the fluoropolymer coated side of release liner in order to have easy release/transfer of the formulation from the liner to backing. Further, when the patch formulation SP2 was peeled off, it was observed that the adhesive layer formed a film on the skin transiently, depicting less affinity of the formulation for polyester backing material.

Therefore, other materials such as EVAC based (Corona-treated - CoTran™

9702, 9706, 9728), and polyethylene based (CoTran™ 9720 and CoTran™ 9718) were

200

tried for optimization of the backing membrane when the uncoated side of ScotchpakTM

1022 was used as the release liner for the patch formulations, SP3-SP6. The SP3 patch could not be prepared successfully as the formulation film did not come off from the liner onto the EVAC backing membranes. Formulation SP5 and SP6, using CoTran™ 9718 as the backing membrane, showed comparatively better peeling characteristic than SP4 patch, prepared using CoTran™ 9720. Higher speed and duration of homogenization was employed for formulation SP6 as compared to that SP5 in an attempt to investigate the effect of homogenization with different parameters on the drug particle size and skin permeation. However, the issue of the residual film of the formulation on skin was not completely resolved in SP5 and SP6, indicating the requirement of a more compatible backing membrane material to facilitate the peeling off process of the patch formulation.

In addition, upon long term storage (after a month), patches SP5 and SP6 showed contraction of the film and the appearance of streaks.

As was evident from patches SP2-SP6, the silicone-based suspension patch formulation did not have sufficient affinity to the backing membrane materials including polyester and polyethylene. Therefore, silicone coated PET films (48101- 4400B/000,

44916 –7300 AM/000, and or 40987- U4162/000) were explored as the backing membrane for SP7, SP8, and SP9, respectively. Patch SP7 (containing 15% w/w TAF) showed the best efficiency in terms of least resistance and minimal formulation residue while peeling the patch formulation off from the skin. Hence, patches SP10-13 were prepared using the same material components and processing parameters as SP7, except that different TAF concentrations (5, 10, 20, 25% w/w) were included in these patches.

For preparation of PIB-based TAF suspension patch formulation (SP14), ScotchpakTM

201

1022 (Fluoropolymer coated side) and ScotchpakTM 9733 were used as the release liner and the backing membrane, respectively.

Visual observation of patches

The details of the visual observations of the different suspension patches have been summarized in Table 9. Patches SP1, SP7, SP10-13, and SP14 were observed to be acceptable in terms of the properties specified in the table. The visual observations depicted that addition of oleyl alcohol in the silicone suspension formulation, rendered it more lipophilic and thus provided more affinity or binding with skin. Using a silicone- coated film as a backing membrane resolved the issue of the residual film after peeling off. The PIB patch formulation (SP14) did not have any issues during the peeling off process as compared to the silicone patch formulations.

Table 9. Visual observations of suspension patches

FORMULATION PROPERTIES CODE Phase Contraction/ Residue Ease of Residue on the separation shrinkage on peeling glove after patch of films release patches removal liner off the after skin peeling SP1 No No No Yes No SP2 No No No Yes Yes SP4 No No No Yes Yes Not when applied SP5 No Yes (after No Yes afresh, but if applied a month) after storing for few days Not when applied SP6 No No Yes afresh, but if applied

202 TABLE 9 (Continued)

Yes (after after storing for few a month) days SP7 No No No Yes No SP8 No No No No No SP9 No No No No No SP10 No No No Yes No SP11 No No No Yes No SP12 No No No Yes No SP13 No No No Yes No SP14 No No No Yes No

Effect of homogenization on the particle size of TAF

The effect of homogenization parameters on drug’s particle size was also investigated in order to eventually evaluate the effect of the particle size of API in the adhesive matrix on the skin permeation. Homogenization is a process of micronization or reducing particle size of dispersions by application of high sheer, pressure, turbulence, as well as acceleration and impact (Dhankhar, 2014). Decrease in particle size would aid in enhancing the solubility of the drug in the adhesive formulation that may affect its permeation rate (Chu, Lee, Jeong, & Park, 2012). Effect of homogenization speed and duration on particle size of API has previously been reported and thus, similar parameters were selected to reduce the particle size of TAF in this study (Anarjan et al., 2015; N.

Sharma, Madan, & Lin, 2016). As reported, homogenization speed is an indicator of the amount of energy applied to the system that is determined by the velocity of the rotating mixing heads. Further, the mechanical impingement of the particles against the wall, due to high acceleration of the fluid and shear stress in the gap between the rotor and stator, leads to the reduction in the particle size of the drug substance (Anarjan et al., 2015).

203

The average particle size of pure TAF was observed to be 53.39 ± 16.15 µm under the optical microscope (Figure 34a). After homogenization at 30,000 rpm for 20 min, particle size was reduced to be 11.51 ± 2.89 µm (p<0.05) (Figure 34b). Further increasing the speed to 32,000 rpm and the homogenization time to 30 min, significantly reduced the particle size to 6.0 ± 1.8 µm as compared to that of homogenization at 30,000 rpm for 20 min (p<0.05) (Figure 34c). The observation of the reduction in the particle size of TAF after the increase in the speed and time of homogenization was in concordance with those reported previously (Anarjan et al., 2015). As shown in Figure

34d, some of the particles were smaller and not visible under the microscope and thus, could not be measured using the software. However, it was found that homogenization at

32,000 rpm over 5 min resulted in a similar reduction in drug particle size (5.28 ± 2.71

µm, Figure 34d) to that of homogenization at 32,000 rpm for 30 min. Collectively, the speed of homogenization had relatively more impact on the reduction in particle size as compared to the duration of the homogenization. Thus, 32,000 rpm and 5 min were the homogenization parameters employed for the formulation of SP7-SP13.

204

Figure 34. (a). Microscopic images of TAF particles suspended in heptane TAF particles in suspension formulation, before addition of adhesive (at 20x) after homogenization at: (b). 30,000 rpm for 20 min (c). 32,000 rpm for 30 min (d). 32,000 rpm for 5 min

Coat weight and drug content

Coat weight and drug content of the patches evaluated for the in vitro permeation studies were measured and have been reported in Table 10. As discussed in previous chapter, these measured properties are influenced by various factors such as percentage of non-volatile components in the adhesive blend, scale set-up, surface level of the casting knife, textural properties of release liner and backing membrane, and are reflective of the coating efficiency of the patches. Results showed that all the patches had uniformity in both coat weight and drug assay. Also, the values of theoretical and

205

experimental TAF content were close for all the patches that confirmed the stability of the drug after exposure to the temperature conditions used for the processing of respective patches.

Table 10. Coat weight and drug content of silicone-based suspension patches

PATCH COAT WEIGHT (mg/cm2), DRUG CONTENT (mg/0.64 cm2), FORMULATION CODE mean ± SD mean ± SD Theoretical Experimental Theoretical Experimental SP1 10 12.30 ± 3.15 0.96 1.35 ± 0.28 SP5 25 24.23 ± 2.88 2.40 2.52 ± 0.19 SP6 25 20.99 ± 5.80 2.40 2.17 ± 0.55 SP7 30 29.60 ± 1.94 2.88 2.94 ± 0.65 SP10 35 34.15 ± 3.75 1.12 1.11 ± 0.06 SP11 35 36.05 ± 1.29 2.24 2.53 ± 0.11 SP12 20 16.43 ± 0.39 2.56 2.31 ± 0.16 SP13 20 18.29 ± 1.82 3.20 3.02 ± 0.14 SP14 5 4.54 ± 0.55 0.48 0.80 ± 0.06

In vitro permeation studies

Epidermal integrity and thickness assessment. Epidermal pieces with the resistance above 10 kΩ, were selected for the study. Further, the thickness of human epidermis used for the permeation studies ranged from 50-150 µm.

Determination of permeation flux of TAF. PIB based suspension patch (SP14) showed permeation of 90.40 ± 10.18 µg/cm2 after 7 days through human epidermis.

Permeation flux rate of 0.60 ± 0.07 µg/cm2/h and a lag time of about 19 h was observed for TAF with this patch. Results showed that the flux rate from SP14 was significantly higher than AP1 (p<0.05), but similar to that of AP12, AP13, AP14, and AP15 (p>0.05).

206

Further, silicone patch (SP1) with the same composition as PIB (SP14) was prepared.

The cumulative amount of TAF that permeated from SP1 across human epidermis in 7 days was observed to be 432.21 ± 41.25 µg/cm2 and was significantly higher than all the other patches evaluated previously (p<0.05), including the acrylate-based clear patches.

The skin permeation profiles of PIB (SP14) and silicone (SP1) based patch have been shown in Figure 35.

Figure 35. Comparison of permeation profile of TAF from silicone and PIB based patches through human epidermis * represents statistical significant difference, Student’s t test (p<0.05)

The resultant permeation flux from SP1 for 7 days was found to be 3.39 ± 0.03

µg/cm2/h and the lag time was observed to be only 1.2 h. Further, with the addition of

10% w/w oleyl alcohol as well as increased coat weight of ~250 gsm in patch SP5, permeation flux of TAF was observed to increase significantly to 5.97 ± 0.47 µg/cm2/h

(p<0.05). The total amount of TAF observed in the receptor after 7 days was 958.07 ±

82.81 µg/cm2 and the lag time was about 1 h.

207

Briefly, suspension blend for patch SP5 was homogenized at 30,000 rpm for 20 min, while homogenization at 32,000 rpm for 30 min was performed for the preparation of SP6, to investigate the effect of homogenization parameters on the skin permeation.

However, the amount of drug permeation after 7 days was not found to be significantly different between the two patch formulations (SP5: 958.07 ± 82.81 µg/cm2 and SP6:

1054.38 ± 214.40 µg/cm2). It was found that permeation flux rate of TAF from patch SP6 was 6.60 ± 1.48 µg/cm2/h and lag time of about 1 h was observed. Furthermore, the patch

SP7 blend was prepared by homogenization at 32,000 rpm for only 5 min and it did not show any significant difference in the amount of drug permeation as well as the lag time as compared to those of SP5 and SP6 (p>0.05) as shown in Figure 36. The permeation flux of TAF observed from patch SP7 was 7.24 ± 0.47 μg/cm2/h. Also, the use of different backing materials in patches SP5/6, and SP7, did not influence the drug permeation, indicating no major interactions between the drug and the backing membranes.

Figure 36. Comparison of permeation profile of TAF from silicone-based suspension patches through human epidermis

208

Further, silicone-based suspension patches, SP10-13, were prepared to investigate the effect of different TAF concentrations on its permeation profile across human epidermis. As shown in Figure 37, the cumulative amount of TAF that permeated after 7 days was found to be 191.51 ± 23.26, 695.09 ± 68.85, 1028.69 ± 78.63, 1158.86 ± 57.96,

1266.19 ± 162.04 µg/cm2 from SP10, SP11, SP7, SP12, and SP13, respectively.

Figure 37. Effect of TAF concentration in the silicone-based suspension patches on its permeation through human epidermis * represents significant difference between 5 and 10% patch ** represents significant difference as compared to 5 and 10% patch, ANOVA one-way test (p<0.05)

Significant increase in permeation was observed with the increase in TAF concentration from 5 (SP10) to 10% w/w (SP11) as well as from 10 (SP11) to 15% w/w

(SP7), (p<0.05). However, there was no statistically significant difference between the amount of drug that permeated from patches containing 15 (SP7), 20 (SP12), and 25% w/w (SP13) TAF (p>0.05). Thus, increasing the drug concentration from 5 to 15% w/w resulted in higher amount of undissolved drug as a reservoir in the latter that probably

209

provided an additional and constant driving force, eventually aiding in considerably enhancing the drug permeation. However, increasing the drug concentration from 15 to

25% w/w did not result in further enhancement in permeation. Therefore, 15% drug concentration (SP7) was found to be optimum in order to achieve the targeted permeation flux.

The aim of this study was to demonstrate proof of concept of feasibility of developing a transdermal patch of TAF and achieve a clinically relevant flux of about 7

µg/cm2/h across human epidermis. From in vivo perspective, as blood vessels are present in the dermis layer, once the drug crosses the stratum corneum and viable epidermis, it reaches the systemic circulation. Hence, human epidermis was used in this study for the in vitro permeation testing. Owing to the desirable physicochemical properties (molecular weight ~450 g/mol, log P~ 2.0), TAF showed permeation from the simplest transdermal patch formulation (AP1), across the human epidermis, depicting its potential to permeate passively through human skin as described in chapter 7. However, for the acrylate-based drug containing patch formulations, it was only possible to achieve a flux rate of 0.6

µg/cm2/h, with a combination use of different chemical penetration enhancers at a low drug concentration (2% w/w). However, silicone-based suspension patch formulations containing 15-25% w/w TAF showed higher permeation flux of about 7 µg/cm2/h as well as a lower lag time (about 1 h) that may be attributed to the release characteristics of the drug from the silicone adhesive matrix, and a higher drug loading in a suspension form which can provide constant concentration gradient/driving force for its release and continuous permeation through epidermis. In addition, all the excipients included in the suspension patch: mineral oil, oleic acid, and oleyl alcohol have been previously reported

210

for transdermal permeation enhancing effects, which may contribute to higher flux. Thus, overall the patch configuration designed in this study (15% drug suspended in silicone adhesive) showed 12 times higher flux (7.23 ± 0.58 µg/cm2/h), where most of the drug was in the insoluble form) as compared to a transdermal clear patch (0.60 ± 0.10

µg/cm2/h), that can be possibly designed with no issues of crystallization (2% drug) as described in chapter 7.

Thus, in our study it was possible to achieve a relatively high flux rate of 7

µg/cm2/h, from silicone-based optimized suspension patch “SP7”. It was characterized for drug release, coat weight and thickness, and drug content after storage at room temperature and 40 °C for 3 months. It was characterized for physical properties such as tack and peel adhesion as well.

In vitro drug release studies

The 7 day in vitro release profiles of TAF from the optimized patch “SP7” have been presented in Figure 38. The average percentage of TAF released after 7 days was observed to be 64.27 ± 5.47, in the study performed during week 1 after preparation of the patch. Release profile of TAF from the same batch of suspension patch was also studied after 1.5 and 3 months of storage at room temperature and 40 °C and the results have been shown in Table 11. No significant difference in the total percentage of drug released from the patches exposed to different test conditions was observed (p>0.05).

This depicted uniformity in the release profile and percentage of drug released after exposure to higher temperature for 3 months, further indicating stability of the drug in the suspension type patches.

211

Figure 38. In vitro release profiles of TAF from the optimized suspension patch

Table 11. Characterization parameters of optimized suspension patch

AFTER 1.5 AFTER 1.5 AFTER 3 AFTER 3 WEEK 1 PARAMETERS MONTHS MONTHS MONTHS MONTHS (RT) (RT) (40 ͦºC) (RT) (40 ͦºC) % TAF released after 7 64.27 ± 69.71 ± 66.82 ± 97.43 ± 88.86 ± days ± SD 5.47 5.63 12.83 19.03 7.88 Coat weight (mg/cm2) ± 29.60 ± 30.38 ± 32.73 ± 29.39 ± 33.55 ± SD 1.94 0.95 3.06 4.57 3.14 182.66 ± 186.67 ± 188.00 ± 186.00 ± 189.33 ± Coat thickness (µm) ± SD 1.15 3.06 3.46 4.00 3.06 Drug content (mg/cm2) ± 4.60 ± 4.95 ± 4.92 ± 5.25 ± 5.36 ± SD 1.01 0.31 0.39 0.32 0.43 RT= Room temperature

Coat thickness, coat weight, and drug content of optimized patch

The average patch thickness, coat weight as well as drug content of patch SP7, measured after exposure to different temperature conditions for different durations has

212

been presented in Table 11. No significant difference in the measured parameters was observed after exposure of the patches to different test conditions (p>0.05), signifying stability and integrity of the suspension based patches over time. Also, as depicted by the results of drug content analysis, no degradation of TAF was observed after 1.5 and 3 months of patch storage at RT as well as 40°C, further indicating its stability in the silicone based suspension patches.

Physical characterization of the optimized patch

Peel adhesion. An ideal transdermal patch should peel off after application skin, without causing delamination. Peel adhesion is not only affected by the intrinsic adhesiveness of the PSA but involves the stretching and the bending of the patch matrix, and the backing layer prior to the separation as well. The force required to peel the patch should be consistent for different batches and the value of peel adhesion obtained in the test varies the width of the test material (Cilurzo, Gennari, & Minghetti, 2012). No significant difference (p>0.05) in the average force required to peel the patch from human dermatomed skin was observed when tested during week 1 (202.78 ± 43.07 g) and 3 months after patch storage at RT (149.35 ± 30.76 g) and 40°C (171.17 ± 13.56 g). No delamination was observed for the tested transdermal patches.

Tack properties. The adhesion efficiency of a transdermal patch can be tested by evaluating its tack, which is a measure of the force of debonding on application of a light pressure for a short time. A probe tack test was employed in this study, where the force required to separate a probe from the adhesive surface of a transdermal patch was measured. Tack was then expressed as the maximum value of the force required to break the bond between the probe and transdermal patch after a brief period of contact (Cilurzo

213

et al., 2012). The average absolute positive force (adhesive or sticky property), average positive area (work of adhesion) and average separation distance (degree of legging), recorded for three replicates during the first week after patch preparation, was found to be

1116.88  167.79 g/cm2, 34.30  10.98 g.s, and 1.27  0.50 mm, respectively. Storage at room temperature (1012.03  220.59 g/cm2, 24.77  8.50 g.s, and 0.8 ± 0.1 mm, respectively) and 40°C (983.72  201.56 g/cm2, 23.33 ± 6.20 g.s, and 1.0  0.17 mm, respectively) for 3 months did not impact the tack properties of the optimized patch formulation (p>0.05).

Evaluation of skin irritation potential of optimized patch

The principle of irritation assay using the in vitro skin model is based on the premise that irritant chemicals can penetrate the stratum corneum by diffusion and are cytotoxic to the cells in the underlying layers. In a transdermal patch, the drug by itself or in combination with other additives can be irritant to skin (Kandárová et al., 2009).

Hence, the entire patch was tested for its irritation potential, and not just the drug by itself. The tested transdermal patch (SP7) resulted in a mean relative cell viability of

104.64 ± 7.42 % which was comparable to the negative control (100.00 ± 6.09) and significantly higher than the positive control (14.27 ± 8.09). Hence, it can be concluded that the final optimized silicone suspension patch (SP7) is non-irritant to human skin.

Conclusion

A silicone adhesive-based suspension-type transdermal patch configuration, containing 15% w/w TAF designed in this study showed significantly higher permeation flux across the human epidermis as compared to the transdermal clear patch that could possibly be designed with no issues of crystallization (containing 2% w/w TAF). Also,

214

the developed patch was found to be suitable for weekly drug dosing for HIV prophylaxis. Unlike the transparent acrylate-based drug patches, suspension patches did not show any issues of drug crystallization. The optimized patch also demonstrated uniformity in drug release profile as well as other characteristics such as drug content, coat weight, thickness, peel adhesion, and tack properties after 3 months of storage at ambient and higher temperature conditions.

215

CHAPTER 9

SUMMARY AND CONCLUSIONS

The main objective of this work was to employ different skin permeation enhancement strategies to improve the topical or transdermal delivery of cosmetic, therapeutic, and prophylactic actives with poor passive permeation properties. These included already known agents such as EGCG, EVG, TAF as well as new molecular entities such as PAL-353. Physical enhancement technologies including ablative laser, iontophoresis, and/or microneedles were applied for hydrophilic compounds, PAL-353 and EGCG, and different chemical penetration enhancers were explored for improving the transdermal permeation of lipophilic and moderately lipophilic antiretroviral drugs, such as EVG and TAF, respectively. In addition to these studies, two interesting elements of this work included: (1) demonstration of application of in vitro microdialysis for simultaneous quantification of vertical and lateral diffusion of diclofenac sodium (model drug) in human skin and understand the effect of skin microporation via laser and microneedles on the same (2) Formulation of suspension-type transdermal patch of TAF, incorporating different chemical penetration enhancers.

EGCG, the physiologically, most active and abundant flavanol in green tea, possesses anti-inflammatory, anti-oxidant, chemopreventive and skin photoprotective properties. However, due to its low permeability across the stratum corneum as well as susceptibility to photodegradation, it is not used as a key ingredient in cosmetic products.

215 216

As microneedles find major application in cosmetology, are already in use in the cosmetic clinics, and are also available for self-application, they were applied in this study to investigate their effect on skin permeation of EGCG. A rheologically stable hydrogel of EGCG, comprising of 1.5% w/v Carbopol was formulated and L-glutathione was demonstrated to impart significant photostability to EGCG. Further, skin microporation via maltose microneedles was found to significantly enhance the delivery of EGCG to deeper skin layers, viable epidermis and dermis, that are the ideal sites for pharmacological activity of EGCG.

Feasibility of transdermal delivery of PAL-353, a novel phenethylamine substrate-based dopamine/norepinephrine releaser that has demonstrated promising efficacy in preclinical models for treatment of cocaine-use disorder or cocaine dependency, was investigated in this study. Passive permeation of the hydrophilic salt form of the drug was found to be very low. However, skin microporation by ablative laser as well as anodal iontophoresis were found to be the most effective techniques

(~500 fold enhancement), for enhancing the transdermal delivery of PAL-353 across dermatomed human skin. Iontophoresis resulted in the highest cumulative drug permeation, whereas ablative laser showed minimum lag time, indicating faster onset of drug action.

Further, in vitro microdialysis was demonstrated as a novel, effective, and reliable technique for investigating lateral drug diffusion in skin tissue, using diclofenac sodium as a model drug. Thus, a novel application of cutaneous microdialysis for simultaneous quantification of vertical and lateral diffusion of topical and transdermal drugs in real- time in an in vitro set up, was described in this study. Furthermore, using the

217

microdialysis based in vitro study design, physical enhancement techniques such as microneedles and ablative laser were found to enhance vertical and lateral diffusion of diclofenac sodium in dermatomed human skin.

Transdermal permeation of antiretroviral drugs, EVG and TAF across human epidermis was explored to evaluate the feasibility of developing a transdermal patch for both the drugs, for the purpose of HIV prophylaxis. The targeted permeation flux values for EVG and TAF across skin, were about 25 and 7 μg/cm2/h, respectively. However, even with the use of different chemical enhancers such as oleic acid, ethanol, lauric acid,

DMSO, individually or in combination, delivering the prophylactically relevant dose of

EVG was not found to be achievable, probably due its highly lipophilic properties. For

TAF, clear acrylate-based patches incorporating different chemical enhancers such as oleic acid, oleyl alcohol, propylene glycol, octisalate, triacetin were developed. However, due to limited drug solubility in the adhesive formulation and high tendency of drug crystallization, maximum flux of only about 0.6 μg/cm2/h was observed. Finally, a silicone adhesive-based suspension-type transdermal patch was designed for transdermal delivery of TAF. The targeted permeation flux for HIV prophylaxis was achieved with this patch configuration. Unlike the transparent acrylate-based TAF patches, suspension patches did not show any issues of drug crystallization. The optimized patches also demonstrated uniformity in drug release profile as well as other characteristics such as drug content, coat weight and thickness, peel adhesion, and tack properties after 3 months of storage at ambient and higher temperature conditions.

Overall, the investigated chemical and physical enhancement techniques for different actives, successfully improved their delivery from their respective formulations,

218

into or across skin. Also, in vitro cutaneous microdialysis can be concluded as a novel and reliable tool for attaining better understanding regarding vertical and lateral distribution of drugs in skin.

219

BIBLIOGRAPHY

A. Machekposhti, S., Soltani, M., Najafizadeh, P., Ebrahimi, S. A., & Chen, P. (2017). Biocompatible polymer microneedle for topical/dermal delivery of tranexamic acid. Journal of Controlled Release, 261, 87–92. http://doi.org/10.1016/j.jconrel.2017.06.016

Abla, M. J., & Banga, A. K. (2013). Quantification of skin penetration of antioxidants of varying lipophilicity. International Journal of Cosmetic Science, 35(1), 19–26. http://doi.org/10.1111/j.1468-2494.2012.00728.x

Alexander, A., Dwivedi, S., Ajazuddin, Giri, T. K., Saraf, S., Saraf, S., & Tripathi, D. K. (2012). Approaches for breaking the barriers of drug permeation through transdermal drug delivery. Journal of Controlled Release, 164(1), 26-40. http://doi.org/10.1016/j.jconrel.2012.09.017

Alexiades-Armenakas, M. R., Dover, J. S., & Arndt, K. A. (2008). The spectrum of laser skin resurfacing: Nonablative, fractional, and ablative laser resurfacing. Journal of the American Academy of Dermatology, 58(5), 719-737. http://doi.org/10.1016/j.jaad.2008.01.003

Alkilani, A. Z., McCrudden, M. T. C., & Donnelly, R. F. (2015). Transdermal drug delivery: Innovative pharmaceutical developments based on disruption of the barrier properties of the stratum corneum. Pharmaceutics, 7(4), 438-470. http://doi.org/10.3390/pharmaceutics7040438

Almeida, P. F., Vaz, W. L., & Thompson, T. E. (1992). Lateral diffusion in the liquid phases of dimyristoylphosphatidylcholine/cholesterol lipid bilayers: a free volume analysis. Biochemistry, 31(29), 6739–47. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1637810

Anarjan, N., Jafarizadeh-Malmiri, H., Nehdi, I. A., Sbihi, H. M., Al-Resayes, S. I., & Tan, C. P. (2015). Effects of homogenization process parameters on physicochemical properties of astaxanthin nanodispersions prepared using a solvent- diffusion technique. International Journal of Nanomedicine, 10, 1109–1118. http://doi.org/10.2147/IJN.S72835

Andega, S., Kanikkannan, N., & Singh, M. (2001). Comparison of the effect of fatty alcohols on the permeation of melatonin between porcine and human skin. Journal of Controlled Release, 77(1-2), 17–25. http://doi.org/10.1016/S0168-3659(01)004

219 220

Anderson, C., Andersson, T., & Wårdell, K. (1994). Changes in skin circulation after insertion of a microdialysis probe visualized by laser doppler perfusion imaging. Journal of Investigative Dermatology, 102(5), 807–811. http://doi.org/10.1111/1523-1747.ep12378630

Antimisiaris, S. G., & Mourtas, S. (2015). Recent advances on anti-HIV vaginal delivery systems development. Advanced Drug Delivery Reviews, 92, 123-145. http://doi.org/10.1016/j.addr.2015.03.015

Aslam, A., & Alster, T. S. (2014). Evolution of laser skin resurfacing: from scanning to fractional technology. Dermatologic Surgery : Official Publication for American Society for Dermatologic Surgery [et Al.], 40(11), 1163-1172. http://doi.org/10.1097/01.DSS.0000452648.22012.a0

Bachhav, Y. G., Heinrich, A., & Kalia, Y. N. (2011). Using laser microporation to improve transdermal delivery of diclofenac: Increasing bioavailability and the range of therapeutic applications. European Journal of Pharmaceutics and Biopharmaceutics, 78(3), 408–414. http://doi.org/10.1016/j.ejpb.2011.03.006

Bachhav, Y. G., Heinrich, A., & Kalia, Y. N. (2013). Controlled intra- and transdermal protein delivery using a minimally invasive Erbium:YAG fractional laser ablation technology. European Journal of Pharmaceutics and Biopharmaceutics, 84(2), 355– 364. http://doi.org/10.1016/j.ejpb.2012.11.018

Bachhav, Y. G., Summer, S., Heinrich, A., Bragagna, T., Böhler, C., & Kalia, Y. N. (2010). Effect of controlled laser microporation on drug transport kinetics into and across the skin. Journal of Controlled Release, 146(1), 31–36. http://doi.org/10.1016/j.jconrel.2010.05.025

Badran, M. M., Kuntsche, J., & Fahr, A. (2009). Skin penetration enhancement by a microneedle device (Dermaroller®) in vitro: Dependency on needle size and applied formulation. European Journal of Pharmaceutical Sciences, 36(4-5), 511–523. http://doi.org/10.1016/j.ejps.2008.12.008

Baeten, J. M., Donnell, D., Ndase, P., Mugo, N. R., Campbell, J. D., Wangisi, J., … Celum, C. (2012). Antiretroviral Prophylaxis for HIV Prevention in Heterosexual Men and Women. New England Journal of Medicine, 367(5), 399–410. http://doi.org/10.1056/NEJMoa1108524

Baeten, J. M., Palanee-Phillips, T., Brown, E. R., Schwartz, K., Soto-Torres, L. E., Govender, V., … Hillier, S. (2016). Use of a vaginal ring containing dapivirine for HIV-1 prevention in women. Obstetrical and Gynecological Survey, 71(8), 466– 468. http://doi.org/10.1097/01.ogx.0000489577.60775.40

221

Bai, Y., Sachdeva, V., Kim, H., Friden, P. M., & Banga, A. K. (2014). Transdermal delivery of proteins using a combination of iontophoresis and microporation. Therapeutic Delivery, 5(5), 525–536. http://doi.org/10.4155/tde.14.19

Balázs, B., Vizserálek, G., Berkó, S., Budai-Szucs, M., Kelemen, A., Sinkó, B., … Csányi, E. (2016). Investigation of the Efficacy of Transdermal Penetration Enhancers Through the Use of Human Skin and a Skin Mimic Artificial Membrane. Journal of Pharmaceutical Sciences, 105(3), 1134–1140. http://doi.org/10.1016/S0022-3549(15)00172-0

Ball, C., Krogstad, E., Chaowanachan, T., & Woodrow, K. A. (2012). Drug-Eluting Fibers for HIV-1 Inhibition and Contraception. PLoS ONE, 7(11), e49792. http://doi.org/10.1371/journal.pone.0049792

Banerjee, S., Chattopadhyay, P., Ghosh, A., Bhattacharya, S. S., Kundu, A., & Veer, V. (2014). Accelerated stability testing of a transdermal patch composed of eserine and pralidoxime chloride for prophylaxis against (±)-anatoxin A poisoning. Journal of Food and Drug Analysis, 22(2), 264–270. http://doi.org/10.1016/J.JFDA.2014.01.022

Banerjee, S., Chattopadhyay, P., Ghosh, A., Datta, P., & Veer, V. (2014). Aspect of adhesives in transdermal drug delivery systems. International Journal of Adhesion, and Adhesives, 50, 70-84. http://doi.org/10.1016/j.ijadhadh.2014.01.001

Banga, A. K. (1998). Electrically assisted transdermal and topical drug delivery. Taylor & Francis Group.

Banga, A. K. (2011). Transdermal and intradermal delivery of therapeutic agents : application of physical technologies. CRC Press.

Banks, M. L., Blough, B. E., Fennell, T. R., Snyder, R. W., & Negus, S. S. (2013). Effects of phendimetrazine treatment on cocaine vs food choice and extended-access cocaine consumption in rhesus monkeys. Neuropsychopharmacology, 38(13), 2698– 2707. http://doi.org/10.1038/npp.2013.180

Banks, M. L., Blough, B. E., & Negus, S. S. (2011). Effects of monoamine releasers with varying selectivity for releasing dopamine/norepinephrine versus serotonin on choice between cocaine and food in rhesus monkeys. Behavioural Pharmacology, 22(8), 824–36. http://doi.org/10.1097/FBP.0b013e32834d63ac

Barrett, C. W., Hadgraft, J. W., Caron, G. A., & Sarkany, I. (1965). The effect of particle size and vehicle on the percutaneous absorption of fluocinolone acetonide. British Journal of Dermatology, 77(11), 576–578. http://doi.org/10.1111/j.1365- 2133.1965.tb14578.x

222

Batchelder, R. J., Calder, R. J., Thomas, C. P., & Heard, C. M. (2004). In vitro transdermal delivery of the major catechins and caffeine from extract of Camellia sinensis. International Journal of Pharmaceutics, 283(1-2), 45–51. http://doi.org/10.1016/j.ijpharm.2004.06.007

Beauchamp, M., & Lands, L. C. (2005). Sweat-testing: A review of current technical requirements. Pediatric Pulmonology, 39(6), 507-511. http://doi.org/10.1002/ppul.20226

Benson, H. (2005). Transdermal Drug Delivery: Penetration Enhancement Techniques. Current Drug Delivery, 2(1), 23–33. http://doi.org/10.2174/1567201052772915

Benson, H. A. E., & Watkinson, A. C. (2011). Topical and transdermal drug delivery : principles and practice. Wiley.

Berner, B., Mazzenga, G. C., Otte, J. H., Steffens, R. J., Juang, R. ‐H, & Ebert, C. D. (1989). Ethanol: Water mutually enhanced transdermal therapeutic system II: Skin permeation of ethanol and nitroglycerin. Journal of Pharmaceutical Sciences, 78(5), 402–407. http://doi.org/10.1002/jps.2600780512

Bhatnagar, S., Dave, K., & Venuganti, V. V. K. (2017). Microneedles in the clinic. Journal of Controlled Release, 260, 164-182. http://doi.org/10.1016/j.jconrel.2017.05.029

Bianchi, A., Marchetti, N., & Scalia, S. (2011). Photodegradation of (−)-epigallocatechin- 3-gallate in topical cream formulations and its photostabilization. Journal of Pharmaceutical and Biomedical Analysis, 56(4), 692–697. http://doi.org/10.1016/j.jpba.2011.07.007

Boddé, H. E., van den Brink, I., Koerten, H. K., & de Haan, F. H. N. (1991). Visualization of in vitro percutaneous penetration of mercuric chloride; transport through intercellular space versus cellular uptake through desmosomes. Journal of Controlled Release, 15(3), 227–236. http://doi.org/10.1016/0168-3659(91)90114-S

Bonacucina, G., Cespi, M., Misici-Falzi, M., & Palmieri, G. F. (2008). Rheological evaluation of silicon/carbopol hydrophilic gel systems as a vehicle for delivery of water insoluble drugs. The AAPS Journal, 10(1), 84–91. http://doi.org/10.1208/s12248-008-9008-9

Bonacucina, G., Martelli, S., & Palmieri, G. F. (2004). Rheological, mucoadhesive and release properties of Carbopol gels in hydrophilic cosolvents. International Journal of Pharmaceutics, 282(1-2), 115–130. http://doi.org/10.1016/j.ijpharm.2004.06.012

Bouwstra, J. A., Honeywell-Nguyen, P. L., Gooris, G. S., & Ponec, M. (2003). Structure of the skin barrier and its modulation by vesicular formulations. Progress in Lipid Research, 42(1), 1-36. http://doi.org/10.1016/S0163-7827(02)00028-0

223

Bronaugh, R. L., & Maibach, H. I. (2005). Percutaneous Absorption. Drugs and the Pharmaceutical Sciences. Taylor & Francis.

Brown, M. B., Martin, G. P., Jones, S. A., & Akomeah, F. K. (2006). Dermal and Transdermal Drug Delivery Systems: Current and Future Prospects. Drug Delivery, 13(3), 175–187. http://doi.org/10.1080/10717540500455975

Cai, B., Söderkvist, K., Engqvist, H., & Bredenberg, S. (2012). A new drug release method in early development of transdermal drug delivery systems. Pain Research and Treatment, 2012, Article ID 953140. http://doi.org/10.1155/2012/953140

Center for Behavioral Health Statistics and Quality. (2015). Behavioral health trends in the United States: Results from the 2014 National Survey on Drug Use and Health. (HHS Pulication No. SMA 15-4927, NSDUH Series H-50., 64). Retrieved from http://www.samhsa.gov/

Chandrashekar, N. S., & Shobha Rani, R. H. (2008). Physicochemical and pharmacokinetic parameters in drug selection and loading for transdermal drug delivery. Indian Journal of Pharmaceutical Sciences, 70(1), 94–96. http://doi.org/10.4103/0250-474X.40340

Charoenputtakun, P., Li, S. K., & Ngawhirunpat, T. (2015). Iontophoretic delivery of lipophilic and hydrophilic drugs from lipid nanoparticles across human skin. International Journal of Pharmaceutics, 495(1), 318–328. http://doi.org/10.1016/j.ijpharm.2015.08.094

Chatterjee, D. J., Li, W. Y., & Koda, R. T. (1997). Effect of vehicles and penetration enhancers on the in vitro and in vivo percutaneous absorption of methotrexate and edatrexate through hairless mouse skin. Pharmaceutical Research, 14(8), 1058– 1065. http://doi.org/10.1023/A:1012109513643

Cheung, K., Han, T., & Das, D. B. (2014). Effect of force of microneedle insertion on the permeability of insulin in skin. In Journal of Diabetes Science and Technology, 8(3), 444–452. http://doi.org/10.1177/1932296813519720

Chu, K. R., Lee, E., Jeong, S. H., & Park, E.-S. (2012). Effect of particle size on the dissolution behaviors of poorly water-soluble drugs. Archives of Pharmacal Research, 35(7), 1187–1195. http://doi.org/10.1007/s12272-012-0709-3

Cilurzo, F., Gennari, C. G. M., & Minghetti, P. (2012). Adhesive properties: a critical issue in transdermal patch development. Expert Opinion on Drug Delivery, 9(1), 33– 45. http://doi.org/10.1517/17425247.2012.637107

Clausen, T. S., Kaastrup, P., & Stallknecht, B. (2009). Proinflammatory tissue response and recovery of adipokines during 4 days of subcutaneous large-pore microdialysis.

224

Journal of Pharmacological and Toxicological Methods, 60(3), 281–287. http://doi.org/10.1016/j.vascn.2009.03.001

Coldman, M. F., Poulsen, B. J., & Higuchi, T. (1969). Enhancement of percutaneous absorption by the use of volatile: Nonvolatile systems as vehicles. Journal of Pharmaceutical Sciences, 58(9), 1098–1102. http://doi.org/10.1002/jps.2600580912

Das, B., Nayak, A. K., & Nanda, U. (2013). Topical gels of lidocaine HCl using cashew gum and Carbopol 940: Preparation and in vitro skin permeation. International Journal of Biological Macromolecules, 62, 514–517. http://doi.org/10.1016/j.ijbiomac.2013.09.049

Davies, M. I., Cooper, J. D., Desmond, S. S., Lunte, C. E., & Lunte, S. M. (2000). Analytical considerations for microdialysis sampling. Advanced Drug Delivery Reviews, 45(2-3), 169–188. http://doi.org/10.1016/S0169-409X(00)00114-9

De Clercq, E. (2017). Role of tenofovir alafenamide (TAF) in the treatment and prophylaxis of HIV and HBV infections. Biochemical Pharmacology, 153, 2-11. http://doi.org/10.1016/j.bcp.2017.11.023

Deeks, E. D. (2014). Elvitegravir: A review of its use in adults with HIV-1 infection. Drugs, 74(6), 687–697. http://doi.org/10.1007/s40265-014-0206-8

Denet, A. R., Vanbever, R., & Préat, V. (2004). Skin electroporation for transdermal and topical delivery. Advanced Drug Delivery Reviews, 56(5), 659-674. http://doi.org/10.1016/j.addr.2003.10.027

Desimmie, B. A., Schrijvers, R., & Debyser, Z. (2012). Elvitegravir: A once daily alternative to raltegravir. The Lancet Infectious Diseases, 12(1), 3-5. http://doi.org/10.1016/S1473-3099(11)70277-8

Dhankhar, P. (2014). Homogenization Fundamentals. IOSR Journal of Ebgineering (IOSRJEN), 4(5), 1–8. http://doi.org/10.9790/3021-04540108

Dhurat, R., Sukesh, M., Avhad, G., Dandale, A., Pal, A., & Pund, P. (2013). A randomized evaluator blinded study of effect of microneedling in androgenetic alopecia: A pilot study. International Journal of Trichology, 5(1), 6. http://doi.org/10.4103/0974-7753.114700

Domingo, D. S., Camouse, M. M., Hsia, A. H., Matsui, M., Maes, D., Ward, N. L., … Baron, E. D. (2010). Anti-angiogenic effects of epigallocatechin-3-gallate in human skin. International Journal of Clinical and Experimental Pathology, 3(7), 705–9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/20830241

Donnelly, R. F., Morrow, D. I. J., Fay, F., Scott, C. J., Abdelghany, S., Singh, R. R. T.,

225

… David Woolfson, A. (2010). Microneedle-mediated intradermal nanoparticle delivery: Potential for enhanced local administration of hydrophobic pre-formed photosensitisers. Photodiagnosis and Photodynamic Therapy, 7(4), 222–231. http://doi.org/10.1016/j.pdpdt.2010.09.001

Donnelly, R. F., Morrow, D. I. J., McCarron, P. A., Woolfson, A. D., Morrissey, A., Juzenas, P., … Moan, J. (2008). Microneedle-mediated intradermal delivery of 5- aminolevulinic acid: Potential for enhanced topical photodynamic therapy. Journal of Controlled Release, 129(3), 154–162. http://doi.org/10.1016/j.jconrel.2008.05.002

Donnelly, R. F., Singh, T. R. R., Garland, M. J., Migalska, K., Majithiya, R., McCrudden, C. M., … Woolfson, A. D. (2012). Hydrogel-forming microneedle arrays for enhanced transdermal drug delivery. Advanced Functional Materials, 22(23), 4879– 4890. http://doi.org/10.1002/adfm.201200864

Dou, H., Destache, C. J., Morehead, J. R., Mosley, R. L., Boska, M. D., Kingsley, J., … Gendelman, H. E. (2006). Development of a macrophage-based nanoparticle platform for antiretroviral drug delivery. Blood, 108(8), 2827–2835. http://doi.org/10.1182/blood-2006-03-012534

Eisingerich, A. B., Wheelock, A., Gomez, G. B., Garnett, G. P., Dybul, M. R., & Piot, P. K. (2012). Attitudes and acceptance of oral and parenteral HIV preexposure prophylaxis among potential user groups: A multinational study. PLoS ONE, 7(1). e28238. http://doi.org/10.1371/journal.pone.0028238

Erdo, F., Hashimoto, N., Karvaly, G., Nakamichi, N., & Kato, Y. (2016). Critical evaluation and methodological positioning of the transdermal microdialysis technique. A review. Journal of Controlled Release, 233, 147–161. http://doi.org/10.1016/j.jconrel.2016.05.035

European Meedicines Agency. (2014). Guideline on quality of transdermal patches Guideline on quality of transdermal patches. European Medicines Agency, 44(August), 1–28. Retrieved from www.ema.europa.eu/contact

Fang, J.-Y., Hwang, T.-L., Huang, Y.-L., & Fang, C.-L. (2006). Enhancement of the transdermal delivery of catechins by liposomes incorporating anionic surfactants and ethanol. International Journal of Pharmaceutics, 310(1-2), 131–138. http://doi.org/10.1016/j.ijpharm.2005.12.004

Fang, J.-Y., Tsai, T.-H., Lin, Y.-Y., Wong, W.-W., Wang, M.-N., & Huang, J.-F. (2007). Transdermal delivery of tea catechins and theophylline enhanced by terpenes: a mechanistic study. Biological & Pharmaceutical Bulletin, 30(2), 343–349. http://doi.org/10.1248/bpb.30.343

FDA. (2013). Inactive Ingredient Search for Approved Drug Products Inactive Ingredient Search for Approved Drug Products Page 2 of 2, 5–6. Retrieved from

226

https://www.accessdata.fda.gov/scripts/cder/iig/index.cfm

Fraser, I. S., Weisberg, E., Kumar, N., Kumar, S., Humberstone, A. J., McCrossin, L., … Sitruk-Ware, R. (2007). An initial pharmacokinetic study with a Metered Dose Transdermal System® for delivery of the progestogen Nestorone® as a possible future contraceptive. Contraception, 76(6), 432–438. http://doi.org/10.1016/j.contraception.2007.08.006

Funke, A. P., Schiller, R., Motzkus, H. W., Ginther, C., Miller, R. H., & Lipp, R. (2002). Transdermal delivery of highly lipophilic drugs: In vitro fluxes of antiestrogens, permeation enhancers, and solvents from liquid formulations. Pharmaceutical Research, 19(5), 661–668. http://doi.org/10.1023/A:1015314314796

Ganti, S. S., & Banga, A. K. (2016). Non-Ablative Fractional Laser to Facilitate Transdermal Delivery. Journal of Pharmaceutical Sciences, 105(11), 3324–3332. http://doi.org/10.1016/j.xphs.2016.07.023

Gee, C. M., Nicolazzo, J. A., Watkinson, A. C., & Finnin, B. C. (2012). Assessment of the lateral diffusion and penetration of topically applied drugs in humans using a novel concentric tape stripping design. Pharmaceutical Research, 29(8), 2035–2046. https://doi.org/10.1007/s11095-012-0731-7

Ghosh, T. K., Pfister, W. R., & Yum, S. I. (1997). Transdermal and topical drug delivery systems. Interpharm Press.

Golden, G. M., McKie, J. E., & Potts, R. O. (1987). Role of stratum corneum lipid fluidity in transdermal drug flux. Journal of Pharmaceutical Sciences, 76(1), 25–28. http://doi.org/10.1002/jps.2600760108

Gomaa, Y. A., El-Khordagui, L. K., Garland, M. J., Donnelly, R. F., McInnes, F., & Meidan, V. M. (2012). Effect of microneedle treatment on the skin permeation of a nanoencapsulated dye. Journal of Pharmacy and Pharmacology, 64(11), 1592– 1602. http://doi.org/10.1111/j.2042-7158.2012.01557.x

Gong, K., Viboonkiat, R., Rehman, I. U., Buckton, G., & Darr, J. A. (2005). Formation and characterization of porous indomethacin-PVP coprecipitates prepared using solvent-free supercritical fluid processing. Journal of Pharmaceutical Sciences, 94(12), 2583–2590. http://doi.org/10.1002/jps.20474

Gong, Y., Chowdhury, P., Midde, N. M., Rahman, M. A., Yallapu, M. M., & Kumar, S. (2017). Novel elvitegravir nanoformulation approach to suppress the viral load in HIV-infected macrophages. Biochemistry and Biophysics Reports, 12, 214–219. http://doi.org/10.1016/j.bbrep.2017.10.005

Goodman, M., & Barry, B. W. (1989). Lipid-protein-partitioning (LPP) theory of skin enhancer activity: finite dose technique. International Journal of Pharmaceutics,

227

57(1), 29–40. http://doi.org/10.1016/0378-5173(89)90260-3

Gotter, B., Faubel, W., & Neubert, R. H. H. (2010). FTIR microscopy and confocal Raman microscopy for studying lateral drug diffusion from a semisolid formulation. European Journal of Pharmaceutics and Biopharmaceutics, 74(1), 14–20. http://doi.org/10.1016/j.ejpb.2009.07.006

Grabowski, J., Shearer, J., Merrill, J., & Negus, S. S. (2004). Agonist-like, replacement pharmacotherapy for stimulant abuse and dependence. In Addictive Behaviors. 29(7), 1439–1464. http://doi.org/10.1016/j.addbeh.2004.06.018

Grant, R. M., Lama, J. R., Anderson, P. L., McMahan, V., Liu, A. Y., Vargas, L., … Glidden, D. V. (2010). Preexposure Chemoprophylaxis for HIV Prevention in Men Who Have Sex with Men. New England Journal of Medicine, 363(27), 2587–2599. http://doi.org/10.1056/NEJMoa1011205

Gujjar, M., & Banga, A. K. (2014). Iontophoretic and microneedle mediated transdermal delivery of glycopyrrolate. Pharmaceutics, 6(4), 663–671. http://doi.org/10.3390/pharmaceutics6040663

Gunawardana, M., Remedios-Chan, M., Miller, C. S., Fanter, R., Yang, F., Marzinke, M. A., … Baum, M. M. (2015). Pharmacokinetics of long-acting tenofovir alafenamide (GS-7340) subdermal implant for HIV prophylaxis. Antimicrobial Agents and Chemotherapy, 59(7), 3913–3919. http://doi.org/10.1128/AAC.00656-15

Guy, R. H. (2010). Transdermal Drug Delivery. In Handbook of experimental pharmacology (pp. 399–410). http://doi.org/10.1007/978-3-642-00477-3_13

Gwak, H. S., & Chun, I. K. (2002). Effect of vehicles and penetration enhancers on the in vitro percutaneous absorption of tenoxicam through hairless mouse skin. International Journal of Pharmaceutics, 236(1-2), 57–64. http://doi.org/10.1016/S0378-5173(02)00009-1

Ham, A. S., & Buckheit, R. W. (2015). Current and emerging formulation strategies for the effective transdermal delivery of HIV inhibitors. Therapeutic Delivery, 6(2), 217–229. http://doi.org/10.4155/tde.14.110

Harada, K., Murakami, T., Yata, N., & Yamamoto, S. (1992). Role of intercellular lipids in stratum corneum in the percutaneous permeation of drugs. Journal of Investigative Dermatology, 99(3), 278–282. http://doi.org/10.1111/1523- 1747.ep12616623

Hartmann, M., Hanh, B. D., Podhaisky, H., Wensch, J., Bodzenta, J., Wartewig, S., & Neubert, R. H. H. (2004). A new FTIR-ATR cell for drug diffusion studies. In Analyst.129(10), 902–905. http://doi.org/10.1039/b408940p

228

Heneine, W., & Kashuba, A. (2012). HIV prevention by oral preexposure prophylaxis. Cold Spring Harbor Perspectives in Medicine, 2(3), a007419. http://doi.org/10.1101/cshperspect.a007419

Hoelgaard, A., & Møllgaard, B. (1985). Dermal drug delivery - Improvement by choice of vehicle or drug derivative. Journal of Controlled Release, 2(C), 111–120. http://doi.org/10.1016/0168-3659(85)90037-9

Holmgaard, R., Nielsen, J. B., & Benfeldt, E. (2010). Microdialysis sampling for investigations of bioavailability and bioequivalence of topically administered drugs: Current state and future perspectives. Skin Pharmacology and Physiology, 23(5), 225-243. http://doi.org/10.1159/000314698

Hsu, S. (2005). Green tea and the skin. Journal of the American Academy of Dermatology, 52(6), 1049–1059. http://doi.org/10.1016/j.jaad.2004.12.044

Islam, M. T., Rodríguez-Hornedo, N., Ciotti, S., & Ackermann, C. (2004). Rheological characterization of topical carbomer gels neutralized to different pH. Pharmaceutical Research, 21(7), 1192–1199. http://doi.org/10.1023/B:PHAM.0000033006.11619.07

Ita, K. B. (2014). Transdermal drug delivery: progress and challenges. Journal of Drug Delivery Science and Technology, 24(3), 245–250. http://doi.org/10.1016/S1773- 2247(14)50041-X

Jackson, A., & McGowan, I. (2015). Long-acting rilpivirine for HIV prevention. Current Opinion in HIV and AIDS, 10(4), 253-257. http://doi.org/10.1097/COH.0000000000000160

Jackson, J. K., Zhao, J., Wong, W., & Burt, H. M. (2010). The inhibition of collagenase induced degradation of collagen by the galloyl-containing polyphenols tannic acid, epigallocatechin gallate and epicatechin gallate. Journal of Materials Science: Materials in Medicine, 21(5), 1435–1443. http://doi.org/10.1007/s10856-010-4019-3

Jacobi, U., Weigmann, H. J., Baumann, M., Reiche, A. I., Sterry, W., & Lademann, J. (2004). Lateral Spreading of Topically Applied UV Filter Substances Investigated by Tape Stripping. Skin Pharmacology and Physiology, 17(1), 17–22. http://doi.org/10.1159/000074058

Jadoul, A., Bouwstra, J., & Préat, V. (1999). Effects of iontophoresis and electroporation on the stratum corneum. Review of the biophysical studies. Advanced Drug Delivery Reviews, 35(1), 89-105. http://doi.org/10.1016/S0169-409X(98)00065-9

Jadoul, A., Doucet, J., Durand, D., & Préat, V. (1996). Modifications induced on stratum corneum structure after in vitro iontophoresis: ATR-FTIR and X-ray scattering studies. Journal of Controlled Release, 42(2), 165–173. http://doi.org/10.1016/0168-

229

3659(96)01452-6

Jain, P., & Banga, A. K. (2013). Induction and inhibition of crystallization in drug-in- adhesive-type transdermal patches. Pharmaceutical Research, 30(2), 562–571. http://doi.org/10.1007/s11095-012-0901-7

Jay, J. S., & Gostin, L. O. (2012). Ethical Challenges of Preexposure Prophylaxis for HIV. JAMA, 308(9), 867-8. http://doi.org/10.1001/2012.jama.10158

Johnson, M. E., Berk, D. a, Blankschtein, D., Golan, D. E., Jain, R. K., & Langer, R. S. (1996). Lateral diffusion of small compounds in human stratum corneum and model lipid bilayer systems. Biophysical Journal, 71(5), 2656–2668. http://doi.org/10.1016/S0006-3495(96)79457-2

Johnson, M. E., Blankschtein, D., & Langer, R. (1997). Evaluation of solute permeation through the stratum corneum: Lateral bilayer diffusion as the primary transport mechanism. Journal of Pharmaceutical Sciences, 86(10), 1162–1172. http://doi.org/10.1021/js960198e

Juluri, A., & Murthy, S. N. (2014). Transdermal iontophoretic delivery of a liquid lipophilic drug by complexation with an anionic cyclodextrin. Journal of Controlled Release, 189, 11–18. http://doi.org/10.1016/j.jconrel.2014.06.014

Kalia, Y. N., Naik, A., Garrison, J., & Guy, R. H. (2004). Iontophoretic drug delivery. Advanced Drug Delivery Reviews, 56(5), 619-658. http://doi.org/10.1016/j.addr.2003.10.026

Kalluri, H., Banga, A. K., Lewis, L., Agalloco, J., Lambert, B., Madsen, R., … Banga, A. K. (2011). Transdermal delivery of proteins. AAPS PharmSciTech, 12(1), 431–441. http://doi.org/10.1208/s12249-011-9601-6

Kandárová, H., Hayden, P., Klausner, M., Kubilus, J., & Sheasgreen, J. (2009). An in vitro skin irritation test (SIT) using the EpiDerm reconstructed human epidermal (RHE) model. Journal of Visualized Experiments : JoVE, (29), 1366. http://doi.org/10.3791/1366

Kandavilli, S., Nair, V., & Panchagnula, R. (2002). Polymers in Transdermal Drug Delivery Systems. Pharmaceutical Technology, 1, 62–80.

Kang, S. M., Song, J. M., & Kim, Y. C. (2012). Microneedle and mucosal delivery of influenza vaccines. Expert Review of Vaccines, 11(5), 547-560. http://doi.org/10.1586/erv.12.25

Karande, P., & Mitragotri, S. (2009). Enhancement of transdermal drug delivery via synergistic action of chemicals. Biochimica et Biophysica Acta (BBA) - Biomembranes, 1788(11), 2362–2373. http://doi.org/10.1016/j.bbamem.2009.08.015

230

Karim, Q. A., Karim, S. S. A., Frohlich, J. A., Grobler, A. C., Baxter, C., Mansoor, L. E., … Taylor, D. (2010). Effectiveness and safety of tenofovir gel, an antiretroviral microbicide, for the prevention of HIV infection in women. Science, 329(5996), 1168–1174. http://doi.org/10.1126/science.1193748

Karim, S. S. A., Kashuba, A. D., Werner, L., & Karim, Q. A. (2011). Drug concentrations after topical and oral antiretroviral pre-exposure prophylaxis: Implications for HIV prevention in women. The Lancet, 378(9787), 279-281. http://doi.org/10.1016/S0140-6736(11)60878-7

Kassis, V., & Søndergaard, J. (1982). Heat-separation of normal human skin for epidermal and dermal prostaglandin analysis. Archives of Dermatological Research, 273(3-4), 301–306. http://doi.org/10.1007/BF00409259

Katiyar, S. K. (2011). Green tea prevents non-melanoma skin cancer by enhancing DNA repair. Archives of Biochemistry and Biophysics, 508(2), 152–8. http://doi.org/10.1016/j.abb.2010.11.015

Katiyar, S. K., Afaq, F., Azizuddin, K., & Mukhtar, H. (2001). Inhibition of UVB- Induced Oxidative Stress-Mediated Phosphorylation of Mitogen-Activated Protein Kinase Signaling Pathways in Cultured Human Epidermal Keratinocytes by Green Tea Polyphenol (−)-Epigallocatechin-3-gallate. Toxicology and Applied Pharmacology, 176(2), 110–117. http://doi.org/10.1006/taap.2001.9276

Kaur, M., Ita, K. B., Popova, I. E., Parikh, S. J., & Bair, D. A. (2014). Microneedle- assisted delivery of verapamil hydrochloride and amlodipine besylate. European Journal of Pharmaceutics and Biopharmaceutics, 86(2), 284–291. http://doi.org/10.1016/j.ejpb.2013.10.007

Kelessidis, V. C., Poulakakis, E., & Chatzistamou, V. (2011). Use of Carbopol 980 and carboxymethyl cellulose polymers as rheology modifiers of sodium-bentonite water dispersions. Applied Clay Science, 54(1), 63–69. http://doi.org/10.1016/j.clay.2011.07.013

Kenneth Miller, Govil, S., & Bhatia, K. (2008, October 30). Fentanyl suspension-based silicone adhesive formulations and devices for transdermal delivery of fentanyl. Retrieved from https://patents.google.com/patent/US20040086551

Khavari, P. A. (1997). Therapeutic gene delivery to the skin. Molecular Medicine Today, 3(12), 533–538. http://doi.org/10.1016/S1357-4310(97)01143-X

Kim, J. Y., Song, J. Y., Lee, E. J., & Park, S. K. (2003). Rheological properties and microstructures of Carbopol gel network system. Colloid and Polymer Science, 281(7), 614–623. http://doi.org/10.1007/s00396-002-0808-7

231

Kim, N. W., Lee, M. S., Kim, K. R., Lee, J. E., Lee, K., Park, J. S., … Jeong, J. H. (2014). Polyplex-releasing microneedles for enhanced cutaneous delivery of DNA vaccine. Journal of Controlled Release, 179(1), 11–17. http://doi.org/10.1016/j.jconrel.2014.01.016

Kim, Y. C., Park, J. H., & Prausnitz, M. R. (2012). Microneedles for drug and vaccine delivery. Advanced Drug Delivery Reviews, 64(14), 1547-1568. http://doi.org/10.1016/j.addr.2012.04.005

Kolli, C. S., & Banga, A. K. (2008). Characterization of solid maltose microneedles and their use for transdermal delivery. Pharmaceutical Research, 25(1), 104–113. http://doi.org/10.1007/s11095-007-9350-0

Kondo, H., Park, S.-H., Watanabe, K., Yamamoto, Y., & Akashi, M. (2004). Polyphenol (−)-epigallocatechin gallate inhibits apoptosis induced by irradiation in human HaCaT keratinocytes. Biochemical and Biophysical Research Communications, 316(1), 59–64. http://doi.org/10.1016/j.bbrc.2004.01.175

Kreyden, O. P. (2004). Iontophoresis for palmoplantar hyperhidrosis. Journal of Cosmetic Dermatology, 3(4), 211–214. http://doi.org/10.1111/j.1473- 2130.2004.00126.x

Kuroda, Y., & Hara, Y. (1999). Antimutagenic and anticarcinogenic activity of tea polyphenols. Mutation Research, 436(1), 69–97. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9878691

Lai-Cheong, J. E., & McGrath, J. A. (2009). Structure and function of skin, hair and nails. Medicine, 37(5), 223–226. http://doi.org/10.1016/J.MPMED.2009.03.002

Lakhani, P., Rishabh, B., & Bafna, P. (2015). Transdermal Patches: Physiochemical and In Vitro evaluation Methods. International Journal of Pharmaceutical Science and Research, 6(5), 1826-36. http://doi.org/10.13040/IJPSR.0975-8232.6(5).

Lampe, M. A., Burlingame, A. L., Whitney, J., Williams, M. L., Brown, B. E., Roitman, E., & Elias, P. M. (1983). Human stratum corneum lipids: characterization and regional variations. Journal of Lipid Research, 24(2), 120–30. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/6833889

Lane, M. E. (2013). Skin penetration enhancers. International Journal of Pharmaceutics, 447(1-2), 12-21. http://doi.org/10.1016/j.ijpharm.2013.02.040

Lawrence, J. N. (1997). Electrical resistance and tritiated water permeability as indicators of barrier integrity of in vitro human skin. Toxicology in Vitro, 11(3), 241–249. http://doi.org/10.1016/S0887-2333(97)00015-5

Lecumberri, E., Dupertuis, Y. M., Miralbell, R., & Pichard, C. (2013). Green tea

232

polyphenol epigallocatechin-3-gallate (EGCG) as adjuvant in cancer therapy. Clinical Nutrition, 32(6), 894–903. http://doi.org/10.1016/j.clnu.2013.03.008

Lee, C. H., Moturi, V., & Lee, Y. (2009). Thixotropic property in pharmaceutical formulations. Journal of Controlled Release, 136(2), 88-98. http://doi.org/10.1016/j.jconrel.2009.02.013

Lee, W. R., Shen, S. C., Lai, H. H., Hu, C. H., & Fang, J. Y. (2001). Transdermal drug delivery enhanced and controlled by erbium:YAG laser: A comparative study of lipophilic and hydrophilic drugs. Journal of Controlled Release, 75(1-2), 155–166. http://doi.org/10.1016/S0168-3659(01)00391-1

Lee, W. R., Shen, S. C., Liu, C. R., Fang, C. L., Hu, C. H., & Fang, J. Y. (2006). Erbium:YAG laser-mediated oligonucleotide and DNA delivery via the skin: An animal study. Journal of Controlled Release, 115(3), 344–353. http://doi.org/10.1016/j.jconrel.2006.08.012

Lee, W. R., Shen, S. C., Wang, K. H., Hu, C. H., & Fang, J. Y. (2003). Lasers and Microdermabrasion Enhance and Control Topical Delivery of Vitamin C. Journal of Investigative Dermatology, 121(5), 1118–1125. http://doi.org/10.1046/j.1523- 1747.2003.12537.x

Lee, Y. B., Eun, Y. S., Lee, J. H., Cheon, M. S., Park, Y. G., Cho, B. K., & Park, H. J. (2013). Effects of topical application of growth factors followed by microneedle therapy in women with female pattern hair loss: A pilot study. Journal of Dermatology, 40(1), 81-83. http://doi.org/10.1111/j.1346-8138.2012.01680.x

Li, Y., Guo, L., & Lu, W. (2013). Laser ablation-enhanced transdermal drug delivery. Photonics and Lasers in Medicine, 2(4), 315-322. http://doi.org/10.1515/plm-2013- 0028

Lieb, W. R., & Stein, W. D. (1986). Non-stokesian nature of transverse diffusion within human red cell membranes. The Journal of Membrane Biology, 92(2), 111-119. http://doi.org/10.1007/BF01870701

Liu, W., Hu, M., Liu, W., Xue, C., Xu, H., & Yang, X. L. (2008). Investigation of the carbopol gel of solid lipid nanoparticles for the transdermal iontophoretic delivery of triamcinolone acetonide acetate. International Journal of Pharmaceutics, 364(1), 135–141. http://doi.org/10.1016/j.ijpharm.2008.08.013

Lobo, S., Sachdeva, S., & Goswami, T. (2016). Role of pressure-sensitive adhesives in transdermal drug delivery systems. Therapeutic Delivery, 7(1), 33–48. http://doi.org/10.4155/tde.15.87

Lu, C.-C., & Yen, G.-C. (2015). Antioxidative and anti-inflammatory activity of functional foods. Current Opinion in Food Science, 2, 1–8.

233

http://doi.org/10.1016/J.COFS.2014.11.002

Lu, Z., & Fassihi, R. (2015). Influence of Colloidal Silicon Dioxide on Gel Strength, Robustness, and Adhesive Properties of Diclofenac Gel Formulation for Topical Application. AAPS PharmSciTech, 16(3), 636–644. http://doi.org/10.1208/s12249- 014-0253-1

Ma, X., Taw, J., & Chiang, C. M. (1996). Control of drug crystallization in transdermal matrix system. International Journal of Pharmaceutics, 142(1), 115–119. http://doi.org/10.1016/0378-5173(96)04647-9

Maartens, G., Celum, C., & Lewin, S. R. (2014). HIV infection: Epidemiology, pathogenesis, treatment, and prevention. In The Lancet, 384(9939), 258–271. http://doi.org/10.1016/S0140-6736(14)60164-1

Maibach, H. I., & Feldmann, R. J. (1967). The effect of DMSO on percutaneous penetration of hydrocortisone and testosterone in man. Annals of the New York Academy of Sciences, 141(1), 423–427. http://doi.org/10.1111/j.1749- 6632.1967.tb34906.x

Maillard-Salin, D. G., Bécourt, P., & Couarraze, G. (2000). Physical evaluation of a new patch made of a progestomimetic in a silicone matrix. International Journal of Pharmaceutics, 199(1), 29–38. http://doi.org/10.1016/S0378-5173(00)00357-4

Mallipeddi, R., & Rohan, L. C. (2010). Nanoparticle-based vaginal drug delivery systems for HIV prevention. Expert Opinion on Drug Delivery, 7(1), 37–48. http://doi.org/10.1517/17425240903338055

Manstein, D., Herron, G. S., Sink, R. K., Tanner, H., & Anderson, R. R. (2004). Fractional photothermolysis: A new concept for cutaneous remodeling using microscopic patterns of thermal injury. Lasers in Surgery and Medicine, 34(5), 426– 438. http://doi.org/10.1002/lsm.20048

Margolis, D. A., González-García, J., Stellbrink, H.-J., Eron, J. J., Yazdanpanah, Y., Griffith, S., … Spreen, W. (2016). Cabotegravir + Rilpivirine as Long-Acting Maintenance Therapy : LATTE-2 Week 32 Results. Conference on Retroviruses and Opportunistic Infections, 2–4. Retrieved from http://www.croiconference.org/sessions/cabotegravirrilpivirine-long-acting- maintenance-therapy-latte-2-week-32-results

Massud, I., Martin, A., Dinh, C., Mitchell, J., Jenkins, L., Heneine, W., … Gerardo Garcí-Lerma, J. (2014). Pharmacokinetic profile of raltegravir, elvitegravir and dolutegravir in plasma and mucosal secretions in rhesus macaques. Journal of Antimicrobial Chemotherapy, 70(5), 1473–1481. http://doi.org/10.1093/jac/dku556

Mayer, K. H., Jones, D., Oldenburg, C., Jain, S., Gelman, M., Zaslow, S., … Mimiaga,

234

M. J. (2017). Optimal HIV Postexposure Prophylaxis Regimen Completion With Single Tablet Daily Elvitegravir/ Cobicistat/Tenofovir Disoproxil Fumarate/Emtricitabine Compared With More Frequent Dosing Regimens. J Acquir Immune Defic Syndr, 75(5):535-539. Retrieved from www.jaids.com

Mayes, S., & Ferrone, M. (2006). Fentanyl HCl patient-controlled iontophoretic transdermal systems for the management of acute postoperative pain. Annals of Pharmacotherapy, 40(12), 2178-2185. http://doi.org/10.1345/aph.1H135

McCormack, S., Dunn, D. T., Desai, M., Dolling, D. I., Gafos, M., Gilson, R., … Gill, O. N. (2016). Pre-exposure prophylaxis to prevent the acquisition of HIV-1 infection (PROUD): effectiveness results from the pilot phase of a pragmatic open-label randomised trial. The Lancet, 387(10013), 53–60. http://doi.org/10.1016/S0140- 6736(15)00056-2

McEnery, R. (2011). Oral tenofovir arm of VOICE trial discontinued early. IAVI Report : Newsletter on International AIDS Vaccine Research, 15(5), 21. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/22121550

Mehdizadeh, A., Toliate, T., Rouini, M. R., Abashzadeh, S., & Dorkoosh, F. (2004). Design and in vitro evaluation of new drug-in-adhesive formulations of fentanyl transdermal patches. Acta Pharmaceutica (Zagreb, Croatia), 54(4), 301–317. https://www.ncbi.nlm.nih.gov/pubmed/15634614

Milewski, M., Yerramreddy, T. R., Ghosh, P., Crooks, P. A., & Stinchcomb, A. L. (2010). In vitro permeation of a pegylated naltrexone prodrug across microneedle- treated skin. Journal of Controlled Release, 146(1), 37–44. http://doi.org/10.1016/j.jconrel.2010.05.034

Min, K., & Kwon, T. K. (2014). Anticancer effects and molecular mechanisms of epigallocatechin-3-gallate. Integrative Medicine Research, 3(1), 16–24. http://doi.org/10.1016/j.imr.2013.12.001

Mitragotri, S. (2000). Synergistic effect of enhancers for transdermal drug delivery. Pharmaceutical Research, 17(11), 1354-1359. http://doi.org/10.1023/A:1007522114438

Mitragotri, S., & Kost, J. (2004). Low-frequency sonophoresis. Advanced Drug Delivery Reviews, 56(5), 589–601. http://doi.org/10.1016/j.addr.2003.10.024

Mizutani, Y., Mitsutake, S., Tsuji, K., Kihara, A., & Igarashi, Y. (2009). Ceramide biosynthesis in keratinocyte and its role in skin function. Biochimie, 91(6), 784-790. http://doi.org/10.1016/j.biochi.2009.04.001

Moghimipour, E., Rezaee, S., & Omidi, A. (2011). The effect of formulation factors on the release of oxybutynin hydrochloride from transdermal polymeric patches.

235

Journal of Applied Pharmaceutical Science, 1(10), 73–76.

Mohammed, D., Hirata, K., Hadgraft, J., & Lane, M. E. (2014). Influence of skin penetration enhancers on skin barrier function and skin protease activity. European Journal of Pharmaceutical Sciences, 51(1), 118–122. http://doi.org/10.1016/j.ejps.2013.09.009

Morrow, K. M., Fava, J. L., Rosen, R. K., Vargas, S., Shaw, J. G., Kojic, E. M., … Katz, and The Project LINK Study Te, D. F. (2014). Designing Preclinical Perceptibility Measures to Evaluate Topical Vaginal Gel Formulations: Relating User Sensory Perceptions and Experiences to Formulation Properties. AIDS Research and Human Retroviruses, 30(1), 78–91. http://doi.org/10.1089/aid.2013.0099

Morrow, K. M., Underhill, K., van den Berg, J. J., Vargas, S., Rosen, R. K., & Katz, D. F. (2014). User-identified gel characteristics: a qualitative exploration of perceived product efficacy of topical vaginal microbicides. Archives of Sexual Behavior, 43(7), 1459–1467. http://doi.org/10.1007/s10508-013-0235-5

Murthy, S. N., Sen, A., Zhao, Y. L., & Hui, S. W. (2003). pH influences the postpulse permeability state of skin after electroporation. Journal of Controlled Release, 93(1), 49–57. http://doi.org/10.1016/j.jconrel.2003.08.002

Nagle, D. G., Ferreira, D., & Zhou, Y.-D. (2006). Epigallocatechin-3-gallate (EGCG): chemical and biomedical perspectives. Phytochemistry, 67(17), 1849–55. http://doi.org/10.1016/j.phytochem.2006.06.020

Naik, Kalia, & Guy. (2000). Transdermal drug delivery: overcoming the skin’s barrier function. Pharmaceutical Science & Technology Today, 3(9), 318–326. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10996573

Nakajima-Kambe, T., Shigeno-Akutsu, Y., Nomura, N., Onuma, F., & Nakahara, T. (1999). Microbial degradation of polyurethane, polyester polyurethanes and polyether polyurethanes. Applied Microbiology and Biotechnology, 51(2), 134-140. http://doi.org/10.1007/s002530051373

Nayak, A., Short, L., & Das, D. B. (2015). Lidocaine permeation from a lidocaine NaCMC/gel microgel formulation in microneedle-pierced skin: vertical (depth averaged) and horizontal permeation profiles. Drug Delivery and Translational Research, 5(4), 372–386. http://doi.org/10.1007/s13346-015-0229-z

Negus, S. S., & Henningfield, J. (2015). Agonist medications for the treatment of cocaine use disorder. Neuropsychopharmacology, 40(8), 1815-1825. http://doi.org/10.1038/npp.2014.322

Negus, S. S., & Mello, N. K. (2003a). Effects of chronic d-amphetamine treatment on cocaine- and food-maintained responding under a second-order schedule in rhesus

236

monkeys. Drug and Alcohol Dependence, 70(1), 39–52. http://doi.org/10.1016/S0376-8716(02)00339-3

Negus, S. S., & Mello, N. K. (2003b). Effects of chronic d-amphetamine treatment on cocaine-and food-maintained responding under a progressive-ratio schedule in rhesus monkeys. Psychopharmacology, 167(3), 324–332. http://doi.org/10.1007/s00213-003-1409-y

Nelson, A. G., Zhang, X., Ganapathi, U., Szekely, Z., Flexner, C. W., Owen, A., & Sinko, P. J. (2015). Drug delivery strategies and systems for HIV/AIDS preexposure prophylaxis and treatment. Journal of Controlled Release, 219, 669–680. http://doi.org/10.1016/j.jconrel.2015.08.042

Nguyen, H. X., & Banga, A. K. (2015). Enhanced skin delivery of vismodegib by microneedle treatment. Drug Delivery and Translational Research, 5(4), 407–423. http://doi.org/10.1007/s13346-015-0241-3

Nguyen, H. X., & Banga, A. K. (2017). Fabrication, characterization and application of sugar microneedles for transdermal drug delivery. Therapeutic Delivery, 8(5), 249– 264. http://doi.org/10.4155/tde-2016-0096

Nguyen, H. X., & Banga, A. K. (2018a). Delivery of Methotrexate and Characterization of Skin Treated by Fabricated PLGA Microneedles and Fractional Ablative Laser. Pharmaceutical Research, 35(3) 68. http://doi.org/10.1007/s11095-018-2369-6

Nguyen, H. X., & Banga, A. K. (2018b). Effect of ablative laser on in vitro transungual delivery. International Journal of Pharmaceutics, 544(2), 402–414. http://doi.org/10.1016/j.ijpharm.2017.09.048

Nguyen, H. X., Puri, A., Bhattaccharjee, S. A., & Banga, A. K. (2018). Qualitative and quantitative analysis of lateral diffusion of drugs in human skin. International Journal of Pharmaceutics, 544(1), 62–74. http://doi.org/10.1016/j.ijpharm.2018.04.013

Nicolazzo, J. A., Morgan, T. M., Reed, B. L., & Finnin, B. C. (2005). Synergistic enhancement of testosterone transdermal delivery. Journal of Controlled Release, 103(3), 577–585. http://doi.org/10.1016/j.jconrel.2004.12.007

Nicoli, S., Penna, E., Padula, C., Colombo, P., & Santi, P. (2006). New transdermal bioadhesive film containing oxybutynin: In vitro permeation across rabbit ear skin. International Journal of Pharmaceutics, 325(1-2), 2–7. http://doi.org/10.1016/j.ijpharm.2006.06.010

Nutan, & Gupta, S. K. (2011). Microbicides: A new hope for HIV prevention. Indian Journal of Medical Research, 134(12), 939-949. http://doi.org/10.4103/0971- 5916.92639

237

Ohm, A. (2000). Interaction of Bay t 3839 coprecipitates with insoluble excipients. European Journal of Pharmaceutics and Biopharmaceutics, 49(2), 183–189. http://doi.org/10.1016/S0939-6411(99)00080-6

Osborne, D. U., Ward, A. J. I., & O’neill, K. J. (1988). Microemulsions as Topical Drug Delivery Vehicles. I. Characterization of a Model System. Drug Development and Industrial Pharmacy, 14(9), 1203–1219. http://doi.org/10.3109/03639048809151929

Ostrenga, J., Steinmetz, C., Poulsen, B., & Yett, S. (1971). Significance of vehicle composition II: Prediction of optimal vehicle composition. Journal of Pharmaceutical Sciences, 60(8), 1180–1183. http://doi.org/10.1002/jps.2600600813

Park, K. Y., Kim, D. H., Jeong, M. S., Li, K., & Seo, S. J. (2010). Changes of antimicrobial peptides and transepidermal water loss after topical application of tacrolimus and ceramide-dominant emollient in patients with atopic dermatitis. Journal of Korean Medical Science, 25(5), 766–771. http://doi.org/10.3346/jkms.2010.25.5.766

Patel, K. N., Patel, H. K., & Patel, V. A. (2012). Formulation and characterization of drug in adhesive transdermal patches of diclofenac acid. International Journal of Pharmacy and Pharmaceutical Sciences, 4(1), 296–299.

Pathan, I. B., & Setty, C. M. (2009). Chemical penetration enhancers for transdermal drug delivery systems. Tropical Journal of Pharmaceutical Research, 8(2), 173-179. http://doi.org/10.4314/tjpr.v8i2.44527

Pattani, A., McKay, P. F., Garland, M. J., Curran, R. M., Migalska, K., Cassidy, C. M., … Donnelly, R. F. (2012). Microneedle mediated intradermal delivery of adjuvanted recombinant HIV-1 CN54gp140 effectively primes mucosal boost inoculations. Journal of Controlled Release, 162(3), 529–537. http://doi.org/10.1016/j.jconrel.2012.07.039

Pettit, D. K., & Gombotz, W. R. (1998). The development of site-specific drug-delivery systems for protein and peptide biopharmaceuticals. Trends in Biotechnology, 16(8), 343-349. http://doi.org/10.1016/S0167-7799(98)01186-X

Pfennig, A. (1995). Kirk-Othmer Encyclopedia of Chemical Technology, 4th Ed., Vol. 10. M. Howe-Grant (Editor). John Wiley & Sons, New York 1993. 1022 Chemie Ingenieur Technik, 67(3), 352–353. http://doi.org/10.1002/cite.330670323

Pierre, M. B. R., & dos Santos Miranda Costa, I. (2011). Liposomal systems as drug delivery vehicles for dermal and transdermal applications. Archives of Dermatological Research, 303(9), 607–621. http://doi.org/10.1007/s00403-011- 1166-4

238

Pikal, M. J. (2001). The role of electroosmotic flow in transdermal iontophoresis. Advanced Drug Delivery Reviews, 46(1-3), 281-305. http://doi.org/10.1016/S0169- 409X(00)00138-1

Pinnell, S. R. (2003). Cutaneous photodamage, oxidative stress, and topical antioxidant protection. Journal of the American Academy of Dermatology, 48(1), 1-19. http://doi.org/10.1067/mjd.2003.16

Prausnitz, M. R. (1996). The effects of electric current applied to skin: A review for transdermal drug delivery. Advanced Drug Delivery Reviews, 18(3), 395–425. http://doi.org/10.1016/0169-409X(95)00081-H

Prausnitz, M. R. (2004). Microneedles for transdermal drug delivery. Advanced Drug Delivery Reviews, 56(5), 581-587. http://doi.org/10.1016/j.addr.2003.10.023

Prausnitz, M. R., & Langer, R. (2008). Transdermal drug delivery. Nature Biotechnology, 26(11), 1261-1268. http://doi.org/10.1038/nbt.1504

Prausnitz, M. R., Mitragotri, S., & Langer, R. S. (2004). Current status and future potential of transdermal drug delivery. Nature Reviews, 3(2), 115–124. http://doi.org/10.1038/nrd1304

Puri, A., Murnane, K. S., Blough, B. E., & Banga, A. K. (2017). Effects of chemical and physical enhancement techniques on transdermal delivery of 3-fluoroamphetamine hydrochloride. International Journal of Pharmaceutics, 528(1-2), 452–462. http://doi.org/10.1016/j.ijpharm.2017.06.041

Puri, A., Nguyen, H. X., & Banga, A. K. (2016). Microneedle-mediated intradermal delivery of epigallocatechin-3-gallate. International Journal of Cosmetic Science, 38(5), 512–523. http://doi.org/10.1111/ics.12320

Puri, A., Sivaraman, A., Zhang, W., Clark, M. R., & Banga, A. K. (2017). Expanding the domain of drug delivery for HIV prevention: exploration of the transdermal route. Critical Reviews in Therapeutic Drug Carrier Systems, 34(6), 551–587. http://doi.org/10.1615/CritRevTherDrugCarrierSyst.2017020147

Quan, D., & Deshpanday, N. A. (2016). Triacetin as a penetration enhancer for Triacetin transdermal delivery of a basic drug, 1–10. Retrieved from https://patents.google.com/patent/US5834010

Rajan, P., & Grimes, P. E. (2002). Skin barrier changes induced by aluminum oxide and sodium chloride microdermabrasion. Dermatologic Surgery, 28(5), 390–393. http://doi.org/10.1046/j.1524-4725.2002.01239.x

Ramana, L. N., Anand, A. R., Sethuraman, S., & Krishnan, U. M. (2014). Targeting

239

strategies for delivery of anti-HIV drugs. Journal of Controlled Release, 192, 271- 283. http://doi.org/10.1016/j.jconrel.2014.08.003

Ramanathan, S., Mathias, A. A., German, P., & Kearney, B. P. (2011). Clinical pharmacokinetic and pharmacodynamic profile of the HIV integrase inhibitor elvitegravir. Clinical Pharmacokinetics, 50(4), 229-244. http://doi.org/10.2165/11584570-000000000-00000

Rao, G., Guy, R. H., Glikfeld, P., LaCourse, W. R., Leung, L., Tamada, J., … Azimi, N. (1995). Reverse Iontophoresis: Noninvasive Glucose Monitoring in Vivo in Humans. Pharmaceutical Research: An Official Journal of the American Association of Pharmaceutical Scientists, 12(12), 1869–1873. http://doi.org/10.1023/A:1016271301814

Rawat, S., Vengurlekar, S., Rakesh, B., Jain, S., & Srikarti, G. (2008). Transdermal delivery by iontophoresis. Indian Journal of Pharmaceutical Sciences, 70(1), 5. http://doi.org/10.4103/0250-474X.40324

Ray, A. S., Fordyce, M. W., & Hitchcock, M. J. M. (2016). Tenofovir alafenamide: A novel prodrug of tenofovir for the treatment of Human Immunodeficiency Virus. Antiviral Research, 125 (63-70). http://doi.org/10.1016/j.antiviral.2015.11.009

Rivera-Gonzalez, G., Shook, B., & Horsley, V. (2014). Adipocytes in skin health and disease. Cold Spring Harbor Perspectives in Medicine, 4(3), a015271. http://doi.org/10.1101/cshperspect.a015271

Rodríguez Bayón, A. M., & Guy, R. H. (1996). Iontophoresis of nafarelin across human skin in vitro. Pharmaceutical Research, 13(5), 798–800. http://doi.org/10.1023/A:1016072205371

Rohan, L. C., Devlin, B., & Yang, H. (2014). Microbicide dosage forms. Current Topics in Microbiology and Immunology, 383, 27–54. http://doi.org/10.1007/82_2013_357

Rost, J., & Rapoport, S. (1964). Reduction-potential of Glutathione. Nature, 201(4915), 185–185. http://doi.org/10.1038/201185a0

Sachdeva, V., Bai, Y., Kydonieus, A., & Banga, A. K. (2013). Formulation and optimization of desogestrel transdermal contraceptive patch using crystallization studies. International Journal of Pharmaceutics, 441(1-2), 9–18. http://doi.org/10.1016/j.ijpharm.2012.12.014

Sachdeva, V., Siddoju, S., Yu, Y. Y., Kim, H. D., Friden, P. M., & Banga, A. K. (2010). Transdermal iontophoretic delivery of terbinafine hydrochloride: Quantitation of drug levels in stratum corneum and underlying skin. International Journal of Pharmaceutics, 388(1-2), 24–31. http://doi.org/10.1016/j.ijpharm.2009.12.029

240

Saini, S., Baghel, S., & Agrawal, S. S. (2014). Recent development in Penetration Enhancers and Techniques in Transdermal Drug Delivery System. J. Adv. Pharm. Edu. & Res., 4(1), 31–40.

Sánchez-Rodríguez, J., Vacas-Córdoba, E., Gómez, R., De La Mata, F. J., & Muñoz- Fernández, M. Á. (2015). Nanotech-derived topical microbicides for HIV prevention: The road to clinical development. Antiviral Research, 113 (33-48). http://doi.org/10.1016/j.antiviral.2014.10.014

Sang, S., Lee, M.-J., Hou, Z., Ho, C.-T., & Yang, C. S. (2005). Stability of Tea Polyphenol (−)-Epigallocatechin-3-gallate and Formation of Dimers and Epimers under Common Experimental Conditions. Journal of Agricultural and Food Chemistry, 53(24), 9478–9484. http://doi.org/10.1021/jf0519055

Santoro, A., Rovati, L. C., Lanzini, R., & Setnikar, I. (2000). Pharmaceutical development and characteristics of a new glyceryl trinitrate transdermal patch. Arzneimittelforschung., 50(10), 897–903.

Santoyo, S., Arellano, A., Ygartua, P., & Martin, C. (1995). Penetration enhancer effects on the in vitro percutaneous absorption of piroxicam through rat skin. International Journal of Pharmaceutics, 117(2), 219–224. http://doi.org/10.1016/0378- 5173(94)00344-5

Scalia, S., Marchetti, N., & Bianchi, A. (2013). Comparative evaluation of different co- Antioxidants on the photochemical- and functional-stability of epigallocatechin-3- Gallate in topical creams exposed to simulated sunlight. Molecules, 18(1), 574–587. http://doi.org/10.3390/molecules18010574

Scalia, S., Trotta, V., & Bianchi, A. (2014). In vivo human skin penetration of (-)- epigallocatechin-3-gallate from topical formulations. Acta Pharmaceutica (Zagreb, Croatia), 64(2), 257–65. http://doi.org/10.2478/acph-2014-0017

Scheuplein, R. J. (1965). Mechanism of percutaneous adsorption. I. Routes of penetration and the influence of solubility. The Journal of Investigative Dermatology, 45(5), 334–346. http://doi.org/10.1038/jid.1965.140

Scheuplein, R. J., & Blank, I. H. (1971). Permeability of the skin. Physiological Reviews, 51(4), 702–747. http://doi.org/10.1152/physrev.1971.51.4.702

Schicksnus, G., & Muller-Goymann, C. C. (2004). Lateral diffusion of ibuprofen in human skin during permeation studies. Skin Pharmacol.Physiol, 17(2), 84–90. https://www.ncbi.nlm.nih.gov/pubmed/14976385

Schlesinger, E., Johengen, D., Luecke, E., Rothrock, G., McGowan, I., van der Straten, A., & Desai, T. (2016). A Tunable, Biodegradable, Thin-Film Polymer Device as a Long-Acting Implant Delivering Tenofovir Alafenamide Fumarate for HIV Pre-

241

exposure Prophylaxis. Pharmaceutical Research, 33(7), 1649–1656. http://doi.org/10.1007/s11095-016-1904-6

Schulz, R., Yamamoto, K., Klossek, A., Flesch, R., Hönzke, S., Rancan, F., … Netz, R. R. (2017). Data-based modeling of drug penetration relates human skin barrier function to the interplay of diffusivity and free-energy profiles. Proceedings of the National Academy of Sciences, 114(14), 3631–3636. http://doi.org/10.1073/pnas.1620636114

Schwindt, D. A., Wilhelm, K. P., & Maibach, H. I. (1998). Water diffusion characteristics of human stratum corneum at different anatomical sites in vivo. Journal of Investigative Dermatology, 111(3), 385–389. http://doi.org/10.1046/j.1523-1747.1998.00321.x

Sekikawa, H., Nakano, M., & Arita, T. (1978). Inhibitory effect of polyvinylpyrrolidone on the crystallization of drugs. Chemical & Pharmaceutical Bulletin, 26(1), 118– 126. http://doi.org/10.1248/cpb.26.118

Sekkat, N., Kalia, Y. N., & Guy, R. H. (2002). Biophysical study of porcine ear skin in vitro and its comparison to human skin in vivo. Journal of Pharmaceutical Sciences, 91(11), 2376–2381. http://doi.org/10.1002/jps.10220

Sharma, A., Gupta, S., Sarethy, I. P., Dang, S., & Gabrani, R. (2012). Green tea extract: Possible mechanism and antibacterial activity on skin pathogens. Food Chemistry, 135(2), 672–675. http://doi.org/10.1016/j.foodchem.2012.04.143

Sharma, N., Madan, P., & Lin, S. (2016). Effect of process and formulation variables on the preparation of parenteral paclitaxel-loaded biodegradable polymeric nanoparticles: A co-surfactant study. Asian Journal of Pharmaceutical Sciences, 11(3), 404–416. http://doi.org/10.1016/j.ajps.2015.09.004

Shukla, C., Bashaw, E. D., Stagni, G., & Benfeldt, E. (2014). Applications of dermal microdialysis: A review. Journal of Drug Delivery Science and Technology, 24(3), 259-269. http://doi.org/10.1016/S1773-2247(14)50044-5

Simonsen, L., Petersen, M. B., Benfeldt, E., & Serup, J. (2002). Development of an in vivo animal model for skin penetration in hairless rats assessed by mass balance. Skin Pharmacol Appl Skin Physiol, 15(6), 414–424. http://doi.org/10.1159/000066455

Singh, B. N., Shankar, S., & Srivastava, R. K. (2011). Green tea catechin, epigallocatechin-3-gallate (EGCG): Mechanisms, perspectives and clinical applications. Biochemical Pharmacology, 82(12), 1807–1821. http://doi.org/10.1016/j.bcp.2011.07.093

Singh, G., Ghosh, B., Kaushalkumar, D., & Somsekhar, V. (2008). Screening of

242

Venlafaxine Hydrochloride for Transdermal Delivery: Passive Diffusion and Iontophoresis. AAPS PharmSciTech, 9(3), 791–797. http://doi.org/10.1208/s12249- 008-9111-3

Singh, I., & Morris, A. P. (2011). Performance of transdermal therapeutic systems: Effects of biological factors. International Journal of Pharmaceutical Investigation, 1(1), 4–9. http://doi.org/10.4103/2230-973X.76721

Singh, V. K., Anis, A., Banerjee, I., Pramanik, K., Bhattacharya, M. K., & Pal, K. (2014). Preparation and characterization of novel carbopol based bigels for topical delivery of metronidazole for the treatment of bacterial vaginosis. Materials Science and Engineering C, 44, 151–158. http://doi.org/10.1016/j.msec.2014.08.026

Sivamani, R. K., Liepmann, D., & Maibach, H. I. (2007). Microneedles and transdermal applications. Expert Opinion on Drug Delivery, 4(1), 19–25. http://doi.org/10.1517/17425247.4.1.19

Sjögren, F., Svensson, C., & Anderson, C. (2002). Technical prerequisites for in vivo microdialysis determination of interleukin-6 in human dermis. British Journal of Dermatology, 146(3), 375–382. http://doi.org/10.1046/j.1365-2133.2002.04621.x

Snopková, S., Havlíčková, K., & Husa, P. (2016). [Tenofovir alafenamide fumarate - a new generation of tenofovir]. Klinicka Mikrobiologie a Infekcni Lekarstvi, 22(3), 111–117. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/27907972

Song, Y., Hemmady, K., Puri, A., & Banga, A. K. (2018). Transdermal delivery of human growth hormone via laser-generated micropores. Drug Delivery and Translational Research, 8(2), 450-460. http://doi.org/10.1007/s13346-017-0370-y

Soriani, M., Rice-Evans, C., & Tyrrell, R. M. (1998). Modulation of the UVA activation of haem oxygenase, collagenase and cyclooxygenase gene expression by epigallocatechin in human skin cells. FEBS Letters, 439(3), 253–7. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9845332

Syed, T. (2007). Phase III management of papulopustular rosacea with 2% green tea extract (-)EGCg in a hydrophilic cream. A placebo-controlled, double- blind study. Journal of the Amercian Academy of Dermatology, 56(2), AB15. http://doi.org/10.1016/j.jaad.2006.10.119

Tamburic, S., & Craig, D. Q. M. (1995). An investigation into the rheological, dielectric and mucoadhesive properties of poly(acrylic acid) gel systems. Journal of Controlled Release, 37(1-2), 59–68. http://doi.org/10.1016/0168-3659(95)00064-F

Tanojo, H., Junginger, H. E., & Boddé, H. E. (1997). In vivo human skin permeability enhancement by oleic acid: Transepidermal water loss and Fourier-transform infrared spectroscopy studies. Journal of Controlled Release, 47(1), 31–39.

243

http://doi.org/10.1016/S0168-3659(96)01613-6

Tanuja Yadav, Mishra, S., Das, S., Aggarwal, S., & Rani, V. (2015). Anticedants and natural prevention of environmental toxicants induced accelerated aging of skin. Environmental Toxicology and Pharmacology, 39(1), 384–91. http://doi.org/10.1016/j.etap.2014.11.003

Tanzi, E. L., Lupton, J. R., & Alster, T. S. (2003). Lasers in dermatology: Four decades of progress. Journal of the American Academy of Dermatology, 49(1), 1-31. http://doi.org/10.1067/mjd.2003.582

Tas, C., Joyce, J. C., Nguyen, H. X., Eangoor, P., Knaack, J. S., Banga, A. K., & Prausnitz, M. R. (2017). Dihydroergotamine mesylate-loaded dissolving microneedle patch made of polyvinylpyrrolidone for management of acute migraine therapy. Journal of Controlled Release, 268, 159–165. http://doi.org/10.1016/j.jconrel.2017.10.021

Taudorf, E. H., Haak, C. S., Erlendsson, A. M., Philipsen, P. A., Anderson, R. R., Paasch, U., & Haedersdal, M. (2014). Fractional ablative erbium YAG laser: Histological characterization of relationships between laser settings and micropore dimensions. Lasers in Surgery and Medicine, 46(4), 281–289. http://doi.org/10.1002/lsm.22228

Taudorf, E. H., Lerche, C. M., Erlendsson, A. M., Philipsen, P. A., Hansen, S. H., Janfelt, C., … Haedersdal, M. (2016). Fractional laser-assisted drug delivery: Laser channel depth influences biodistribution and skin deposition of methotrexate. Lasers in Surgery and Medicine, 48(5), 519–529. http://doi.org/10.1002/lsm.22484

Taudorf, E. H., Lerche, C. M., Vissing, A.-C., Philipsen, P. A., Hannibal, J., D’Alvise, J., … Haedersdal, M. (2015). Topically applied methotrexate is rapidly delivered into skin by fractional laser ablation. Expert Opinion on Drug Delivery, 12(7), 1059– 1069. http://doi.org/10.1517/17425247.2015.1031216

Tettey-Amlalo, R. N. O., Kanfer, I., Skinner, M. F., Benfeldt, E., & Verbeeck, R. K. (2009). Application of dermal microdialysis for the evaluation of bioequivalence of a ketoprofen topical gel. European Journal of Pharmaceutical Sciences, 36(2-3), 219–225. http://doi.org/10.1016/j.ejps.2008.09.002

Thigpen, M. C., Kebaabetswe, P. M., Paxton, L. A., Smith, D. K., Rose, C. E., Segolodi, T. M., … Brooks, J. T. (2012). Antiretroviral Preexposure Prophylaxis for Heterosexual HIV Transmission in Botswana. New England Journal of Medicine, 367(5), 423–434. http://doi.org/10.1056/NEJMoa1110711

Torrisi, B. M., Zarnitsyn, V., Prausnitz, M. R., Anstey, A., Gateley, C., Birchall, J. C., & Coulman, S. A. (2013). Pocketed microneedles for rapid delivery of a liquid-state botulinum toxin A formulation into human skin. Journal of Controlled Release, 165(2), 146–152. http://doi.org/10.1016/j.jconrel.2012.11.010

244

Touitou, E., Alkabes, M., Memoli, A., & Alhaique, F. (1996). Glutathione stabilizes ascorbic acid in aqueous solution. International Journal of Pharmaceutics, 133(1-2), 85–88. http://doi.org/10.1016/0378-5173(95)04419-1

Tsague, L., & Abrams, E. J. (2014). Commentary: Antiretroviral treatment for pregnant and breastfeeding women – the shifting paradigm. AIDS, 28(2), S119–S121. http://doi.org/10.1097/QAD.0000000000000234

Tuan-Mahmood, T. M., McCrudden, M. T. C., Torrisi, B. M., McAlister, E., Garland, M. J., Singh, T. R. R., & Donnelly, R. F. (2013). Microneedles for intradermal and transdermal drug delivery. European Journal of Pharmaceutical Sciences, 50(5), 623-637. http://doi.org/10.1016/j.ejps.2013.05.005

Van Damme, L., Corneli, A., Ahmed, K., Agot, K., Lombaard, J., Kapiga, S., … Taylor, D. (2012). Preexposure Prophylaxis for HIV Infection among African Women. New England Journal of Medicine, 367(5), 411–422. http://doi.org/10.1056/NEJMoa1202614

Van Der Straten, A., Van Damme, L., Haberer, J. E., & Bangsberg, D. R. (2012). Unraveling the divergent results of pre-exposure prophylaxis trials for HIV prevention. AIDS, 26(7), F13-9. http://doi.org/10.1097/QAD.0b013e3283522272

Van Smeden, J., Hoppel, L., van der Heijden, R., Hankemeier, T., Vreeken, R. J., & Bouwstra, J. A. (2011). LC/MS analysis of stratum corneum lipids: ceramide profiling and discovery. Journal of Lipid Research, 52(6), 1211–1221. http://doi.org/10.1194/jlr.M014456

Vedha Hari, B. N., Devendharan, K., & Narayanan, N. (2013). Approaches of novel drug delivery systems for Anti-HIV agents. International Journal of Drug Development and Research, 5(4), 16-24.

Vemulapalli, V., Yang, Y., Friden, P. M., & Banga, A. K. (2008). Synergistic effect of iontophoresis and soluble microneedles for transdermal delivery of methotrexate. Journal of Pharmacy and Pharmacology, 60(1), 27–33. http://doi.org/10.1211/jpp.60.1.0004

Venuganti, V. V., Sahdev, P., Hildreth, M., Guan, X., & Perumal, O. (2011). Structure- skin permeability relationship of dendrimers. Pharmaceutical Research, 28(9), 2246–2260. http://doi.org/10.1007/s11095-011-0455-0

Veselinovic, M., Yang, K. H., LeCureux, J., Sykes, C., Remling-Mulder, L., Kashuba, A. D. M., & Akkina, R. (2014). HIV pre-exposure prophylaxis: Mucosal tissue drug distribution of RT inhibitor Tenofovir and entry inhibitor Maraviroc in a humanized mouse model. Virology, 464-465(1), 253–263. http://doi.org/10.1016/j.virol.2014.07.008

245

Vocci, F. J., Acri, J., & Elkashef, A. (2005). Medication development for addictive disorders: The state of the science. American Journal of Psychiatry, 152(8), 1432- 1440. http://doi.org/10.1176/appi.ajp.162.8.1432

Vocci, F. J., & Appel, N. M. (2007). Approaches to the development of medications for the treatment of methamphetamine dependence. Addiction, 102(suppl 1), 96-106. http://doi.org/10.1111/j.1360-0443.2007.01772.x

Volkow, N. D., & Li, T.-K. (2004). Science and Society: Drug addiction: the neurobiology of behaviour gone awry. Nature Reviews Neuroscience, 5(12), 963– 970. http://doi.org/10.1038/nrn1539

Walters, K. A., & Hadgraft, J. (1993). Pharmaceutical skin penetration enhancement. M. Dekker.

Weigmann, H. J., Lademann, J., Pelchrzim, R. V., Sterry, W., Hagemeister, T., Molzahn, R., … Shah, V. P. (1999). Bioavailability of clobetasol propionate - Quantification of drug concentrations in the stratum corneum by dermatopharmacokinetics using tape stripping. Skin Pharmacology and Applied Skin Physiology, 12(1-2), 46–53. http://doi.org/10.1159/000029845

Weng, W., Quan, P., Liu, C., Zhao, H., & Fang, L. (2016). Design of a Drug-in-Adhesive Transdermal Patch for Risperidone: Effect of Drug-Additive Interactions on the Crystallization Inhibition and In Vitro/In Vivo Correlation Study. Journal of Pharmaceutical Sciences, 105(10), 3153–3161. http://doi.org/10.1016/j.xphs.2016.07.003

Wermeling, D. P., Banks, S. L., Hudson, D. A., Gill, H. S., Gupta, J., Prausnitz, M. R., & Stinchcomb, A. L. (2008). Microneedles permit transdermal delivery of a skin- impermeant medication to humans. Proceedings of the National Academy of Sciences, 105(6), 2058–2063. http://doi.org/10.1073/pnas.0710355105

WHO | Definition of key terms. (2013). WHO. Retrieved from http://www.who.int/hiv/pub/guidelines/arv2013/intro/keyterms/en/

Williams, A. (2003). Transdermal and topical drug delivery from theory to clinical practice. Pharmaceutical Press.

Williams, A. C., & Barry, B. W. (2004). Penetration enhancers. Advanced Drug Delivery Reviews, 56(5), 603-618. http://doi.org/10.1016/j.addr.2003.10.025

Williams, A. C., & Barry, B. W. (2012). Penetration enhancers. Advanced Drug Delivery Reviews, 64, 128-137. http://doi.org/10.1016/j.addr.2012.09.032

Wing, D., Prausnitz, M. R., & Buono, M. J. (2013). Skin pretreatment with microneedles

246

prior to pilocarpine iontophoresis increases sweat production. Clinical Physiology and Functional Imaging, 33(6), 436–440. http://doi.org/10.1111/cpf.12053

Wokovich, A. M., Prodduturi, S., Doub, W. H., Hussain, A. S., & Buhse, L. F. (2006). Transdermal drug delivery system (TDDS) adhesion as a critical safety, efficacy and quality attribute. European Journal of Pharmaceutics and Biopharmaceutics, 64(1), 1-8. http://doi.org/10.1016/j.ejpb.2006.03.009

Wolf, J. M., Dungan, S. R., McCarthy, M. J., Lim, V., & Phillips, R. J. (2014). Vibration- induced geometric patterns of persistent holes in Carbopol gels. Journal of Non- Newtonian Fluid Mechanics, 220, 99–107. http://doi.org/10.1016/j.jnnfm.2014.10.001

Woodward, J. T., & Metevia, V. L. (1985, September 26). Transdermal drug delivery devices with amine-resistant silicone adhesives. Retrieved from https://patents.google.com/patent/US4655767A/en

Wu, J., & Nyborg, W. (2006). Emerging Therapeutic Ultrasound. World Scientific. http://doi.org/10.1142/6047

Yang, C. S., Lambert, J. D., Ju, J., Lu, G., & Sang, S. (2007). Tea and cancer prevention: Molecular mechanisms and human relevance. Toxicology and Applied Pharmacology, 224(3), 265-273. http://doi.org/10.1016/j.taap.2006.11.024

Yang, Y., Kalluri, H., & Banga, A. K. (2011). Effects of chemical and physical enhancement techniques on transdermal delivery of cyanocobalamin (vitamin B12) in vitro. Pharmaceutics, 3(3), 474–484. http://doi.org/10.3390/pharmaceutics3030474

Yerramsetty, K. M., Rachakonda, V. K., Neely, B. J., Madihally, S. V., & Gasem, K. A. M. (2010). Effect of different enhancers on the transdermal permeation of insulin analog. International Journal of Pharmaceutics, 398(1-2), 83–92. http://doi.org/10.1016/j.ijpharm.2010.07.029

Yoshino, S., Mitoma, T., Tsuruta, K., Todo, H., & Sugibayashi, K. (2014). Effect of emulsification on the skin permeation and UV protection of catechin. Pharmaceutical Development and Technology, 19(4), 395–400. http://doi.org/10.3109/10837450.2013.788512

Yu, J., Bachhav, Y. G., Summer, S., Heinrich, A., Bragagna, T., Böhler, C., & Kalia, Y. N. (2010). Using controlled laser-microporation to increase transdermal delivery of prednisone. Journal of Controlled Release, 148(1), e71–e73. http://doi.org/10.1016/j.jconrel.2010.07.032

Zempsky, W. T., Sullivan, J., Paulson, D. M., & Hoath, S. B. (2004). Evaluation of a low-dose lidocaine iontophoresis system for topical anesthesia in adults and

247

children: a randomized, controlled trial. Clinical Therapeutics, 26(7), 1110–9. http://doi.org/S014929180490183X [pii]

Zhang, Y., Brown, K., Siebenaler, K., Determan, A., Dohmeier, D., & Hansen, K. (2012). Development of lidocaine-coated microneedle product for rapid, safe, and prolonged local analgesic action. Pharmaceutical Research, 29(1), 170–177. http://doi.org/10.1007/s11095-011-0524-4

Zhu, Y., Wang, S., Lin, F., Li, Q., & Xu, A. (2014). The therapeutic effects of EGCG on vitiligo. Fitoterapia, 99, 243–51. http://doi.org/10.1016/j.fitote.2014.08.007

Zillich, O. V., Schweiggert-Weisz, U., Hasenkopf, K., Eisner, P., & Kerscher, M. (2013). Antioxidant activity, lipophilicity and extractability of polyphenols from pig skin - development of analytical methods for skin permeation studies. Biomedical Chromatography, 27(11), 1444–1451. http://doi.org/10.1002/bmc.2941