Design and development of soft contact lenses for diabetic eye

'HVLJQDQGGHYHORSPHQWRIVRIWFRQWDFWOHQVHVIRU GLDEHWLFH\H INTERNATIONAL DOCTORAL SCHOOL OF THE USC $QD)LOLSD 3HUHLUDGD0RWD 7KLV3K'7KHVLVZDVIRFXVHGRQWKHGHVLJQRIVRIW&/VVXLWDEOH IRU WKH ORDGLQJ DQG UHOHDVH RI WKHUDSHXWLF DPRXQWV RI GUXJV XVHIXO IRU WKH WUHDWPHQW RI GLDEHWLFUHODWHG RFXODU GLVHDVHV 7ZR W\SHV RI FRPELQDWLRQ SURGXFWV KDYH EHHQ GHYHORSHG VRIW &/V ZLWK DWRUYDVWDWLQ DQG VRIW &/V ZLWK SUDYDVWDWLQ5HOHDVHSURGLOHVRIGUXJORDGHG&/VLQD'SULQWHG LQYLWURPRGHOZHUHFRPSDUHGZLWKGUXJOHYHOVLQWKHWHDUGOXLG RI LQ YLYR DQLPDO PRGHOV 0HODQLQ ZDV LQFRUSRUDWHG LQWR &/V ZLWK WKH GRXEOH DLP RI LQFRUSRUDWLQJ GUXJV LQWR &/V DQG HQGRZLQJ &/V ZLWK SKRWRWKHUPDO FDSDFLW\ 7KLV ZRUN UHQGHU GUXJ&/FRPELQDWLRQSURGXFWVIRUWRSLFDORFXODUGUXJGHOLYHU\ WKDW SURYLGH VXVWDLQHG GUXJ OHYHOV LQ WKH RFXODU VXUIDFH 0RUHRYHULQYLWURWHVWPHWKRGVWKDWPLPLFGUXJUHOHDVHRQWKH RFXODUVXUIDFHKDYHEHHQLGHQWLGLHG 7+(6,6 Ana Filipa Pereira da Mota 2022 3K'7KHVLV 'HVLJQDQGGHYHORSPHQWRIVRIW FRQWDFWOHQVHVIRUGLDEHWLFH\H 6DQWLDJRGH&RPSRVWHOD Doctoral Programme in Drug Research and Development DOCTORAL THESIS DESIGN AND DEVELOPMENT OF SOFT CONTACT LENSES FOR DIABETIC EYE Ana Filipa Pereira da Mota INTERNATIONAL DOCTORAL SCHOOL OF THE UNIVERSITY OF SANTIAGO DE COMPOSTELA DOCTORAL PROGRAM IN DRUG RESEARCH AND DEVELOPMENT SANTIAGO DE COMPOSTELA 2022 AUTHORIZATION OF THESIS SUPERVISOR Design and development of soft contact lenses for diabetic eye Prof. Carmen Álvarez Lorenzo REPORT: That the present Thesis, corresponds to the work carried out by Miss Ana Filipa Pereira da Mota, under my supervision, and that I authorize its presentation considering it gathers the necessary requirements of article 34 of the USC Doctoral Studies Regulation, and that as supervisor of this Thesis, it does not incur in the abstention causes established by the law 40/2015 In accordance with the Doctoral Studies Regulations, I also declare that this doctoral thesis is suitable to be defended as a Monographic with reproduction of publications, in which the participation of the doctoral student was decisive for its elaboration and the publications are in accordance with the Research Plan. At Santiago de Compostela, on December 19th, 2022 PhD CANDIDATE STATEMENT Miss. Ana Filipa Pereira da Mota PhD Thesis: Design and development of soft contact lenses for diabetic eye I submit my Doctoral Thesis, following the procedure according to the Regulation, stating that: 1. This Thesis gathers the results corresponding to my work. 2. When necessary, explicit mention is given to the collaborations the work may have had. 3. The present document is the final version submitted for its defense and coincides with the document sent in electronic format. 4. I confirm that the Thesis does not incur in any plagiarism of any other authors or documents submitted by me for obtaining other degrees. At Santiago de Compostela, on December 19th, 2022 Sgd. Ana Filipa Pereira da Mota Aos meus pais e irmã. 3. ATORVASTATIN-ELUTING CONTACT LENSES: EFFECTS OF MOLECULAR IMPRINTING AND STERILIZATION ON DRUG LOADING AND RELEASE ..... 175 3.1 INTRODUCTION ................................................................... 175 3.2 MATERIALS AND METHODS ............................................ 181 3.2.1 Materials ........................................................................... 181 3.2.2 Computational Modeling .................................................. 182 3.2.3 Hydrogel Preparation ....................................................... 183 3.2.4 Hydrogel Characterization ............................................... 185 3.2.5 Atorvastatin Loading and Release .................................... 186 3.2.6 HET-CAM Test ................................................................. 187 3.2.7 Sterilization ....................................................................... 187 3.2.8 FTIR-ATR Analysis ........................................................... 188 3.2.9 Cytocompatibility Studies ................................................. 189 3.2.10 Cornea and Sclera Permeability and Accumulation ...... 190 3.2.11 Light Stability of Atorvastatin Calcium Solution ............ 191 3.2.12 Statistical Analysis .......................................................... 192 3.3 RESULTS AND DISCUSSION ............................................. 192 3.3.1 Hydrogels Synthesis and Conditioning ............................. 192 3.3.2 Hydrogels Characterization ............................................. 195 3.3.3 Atorvastatin Loading and Release .................................... 197 3.3.4 HET-CAM Test and Cytocompatibility ............................. 202 3.3.5 Effects of Sterilization ....................................................... 204 3.3.6 Cornea and Sclera Permeability and Accumulation ........ 216 3.3.7 Stability of Atorvastatin Calcium Solution ....................... 218 3.4 CONCLUSIONS .................................................................... 219 3.5 REFERENCES ....................................................................... 220 4. CONTACT LENSES FOR PRAVASTATIN DELIVERY TO EYE SEGMENTS: DESIGN AND IN VIVO-IN VITRO CORRELATIONS .......................................................................... 235 4.1 INTRODUCTION .................................................................. 235 4.2 MATERIALS AND METHODS ............................................ 240 4.2.1 Materials .......................................................................... 240 4.2.2 Synthesis of Hydrogel Discs and CLs .............................. 241 4.2.3 Hydrogel Characterization .............................................. 242 4.2.4 Pravastatin Loading and Release .................................... 244 4.2.4.1 Pravastatin Release from Sterile CLs ........................ 245 4.2.5 HET-CAM Test ................................................................. 246 4.2.6 Cytocompatibility and Anti-inflammatory Activity .......... 246 4.2.7 Ex vivo Cornea and Sclera Permeability ......................... 248 4.2.8 Sterilization and Stability of Pravastatin-Loaded Hydrogels ................................................................................................... 250 4.2.9 In vivo Experiments: Drug Release and Tissue Accumulation ............................................................................. 250 4.2.9.1 In vivo Release ........................................................... 253 4.2.9.2 Quantification of Pravastatin in Tear Fluid .............. 254 4.2.9.3 Quantification of Pravastatin in Ocular Tissues ....... 254 4.2.10 Statistical Analysis ......................................................... 255 4.3 RESULTS AND DISCUSSION ............................................. 256 4.3.1 Hydrogels Synthesis and Characterization ...................... 256 4.3.2 Pravastatin Loading and In vitro Release ........................ 259 4.3.3 Cytocompatibility and HET-CAM .................................... 261 4.3.4 Effects of Sterilization and Light Exposition on Drug Stability and Release .................................................................. 264 4.3.5 Effects of SLF Volume and Proteins on In vitro Drug Release Profiles ......................................................................... 266 4.3.6 Ex vivo Corneal and Scleral Permeability ....................... 270 4.3.7 Anti-inflammatory Activity ................................................ 273 4.3.8 In vivo Studies ................................................................... 276 4.3.9 In vitro-in vivo Correlations ............................................. 281 4.4 CONCLUSIONS ..................................................................... 284 4.5 REFERENCES ........................................................................ 284 5. IN VITRO – IN VIVO CORRELATION OF DRUG RELEASE PROFILES FROM MEDICATED CONTACT LENSES USING AN IN VITRO EYE BLINK MODEL .......................................... 299 5.1 INTRODUCTION ................................................................... 299 5.2 MATERIALS AND METHODS ............................................ 303 5.2.1 Materials ........................................................................... 303 5.2.2 Contact Lens Preparation................................................. 304 5.2.3 Drug Loading.................................................................... 305 5.2.3.1 Pravastatin Sodium .................................................... 305 5.2.3.2 Resveratrol ................................................................. 306 5.2.4 Eye Blink Model ................................................................ 306 5.2.4.1 Eyeball and Collection Unit ....................................... 307 5.2.4.2 Eyelid ......................................................................... 308 5.2.4.3 Flow Rate and Blinking ............................................. 309 5.2.4.4 Temperature and Humidity ........................................ 309 5.2.4.5 Release Sampling ....................................................... 309 5.2.4.6 Drug Extraction from the Eyelid and CLs ................. 310 5.2.5 Release in a Vial ............................................................... 310 5.2.6 Drug Quantification Methods .......................................... 311 5.2.7 Statistical Analysis ........................................................... 312 5.3 RESULTS AND DISCUSSION ............................................. 312 5.3.1 3D Eye Blink Model ......................................................... 312 5.3.1.1 Pravastatin ................................................................ 312 5.3.1.2 Resveratrol ................................................................ 315 5.3.2 In vivo Release – Eye Blink Model Comparisons ............ 317 5.3.2.1 Pravastatin ................................................................ 318 5.3.2.2 Resveratrol ................................................................ 323 5.4 CONCLUSIONS .................................................................... 326 5.5 REFERENCES ....................................................................... 327 6. BIOINSPIRED MELANIN CONTACT LENSES WITH ANTIBACTERIAL PROPERTIES .............................................. 335 6.1 INTRODUCTION .................................................................. 335 6.2 MATERIALS AND METHODS ............................................ 340 6.2.1 Materials .......................................................................... 340 6.2.2 Silicone Hydrogel Discs Preparation .............................. 341 6.2.3 Melanin Synthesis and Incorporation on the Hydrogels .. 342 6.2.4 Characterization of Melanin Discs and CLs .................... 343 6.2.4.1 Raman Spectroscopy.................................................. 343 6.2.5 Ciprofloxacin Loading and Release ................................. 343 6.2.6 Light Transmittance and Photothermal Activity ............... 344 6.2.7 Antibacterial Performance by Isothermal Microcalorimetry ................................................................................................... 344 6.3 RESULTS AND DISCUSSION ............................................. 345 6.3.1 Melanin Incorporation...................................................... 345 6.3.2 Light Transmittance and Photothermal Activity ............... 349 6.3.3 Ciprofloxacin Loading and Release ................................. 354 6.3.4 Antibacterial Performance ............................................... 357 6.4 CONCLUSIONS ..................................................................... 360 6.5 REFERENCES ........................................................................ 361 7. CONCLUSIONS ......................................................................... 369 ANNEXES........................................................................................ 375 RESUMO 3 RESUMO A visión é un dos sentidos que xoga un papel máis relevante na interacción social das persoas desde o momento do nacemento ata a vellez. Factores xenéticos, o envellecemento, determinados estilos de vida e comportamentos, as infeccións e unha serie de condicións de saúde son factores de risco importantes para ao desenvolvemento de afeccións oculares que poden comprometer a capacidade de visión. A diabetes mellitus é un dos maiores desafíos de saúde e está a acadar proporcións epidémicas. A falta de control da glicemia provoca unha variedade de danos en múltiples estruturas dos segmentos anterior e posterior do ollo, desencadeando o desenvolvemento de queratopatía diabética, síndrome do ollo seco, cataratas, glaucoma, retinopatía diabética e edema macular. A concienciación pública sobre as enfermidades oculares está aumentando, o que permite un diagnóstico precoz e o início temprano do tratamento das patoloxías oculares, e xenera unha maior demanda de solucións máis eficaces e amigábeis para o paciente. Atopar un tratamento eficaz a nivel ocular é un desafío, xa que hai moitos factores a ter en conta para que o fármaco alcance a súa diana. O ollo é un órgano complexo equipado con varias barreiras anatómicas e fisiolóxicas fronte a penetración de fármacos, incluíndo mecanismos de defensa dinámica (parpadeo ou lagrimeo), drenaxe nasolacrimal, os ANA FILIPA PEREIRA DA MOTA 4 compoñentes e dinámica da película lacrimal e barreiras anatómicas estáticas (corneal, conxuntival e sanguínea). Un dos retos máis importantes da administración ocular é manter a concentración óptima do fármaco no lugar de acción durante períodos prolongados de tempo. Segundo a zona do ollo afectada (segmento anterior ou posterior), a administración presenta máis ou menos dificultades. A vía sistémica (oral, intramuscular ou intravenosa) úsase raramente, xa que son necesarias doses elevadas de fármaco para alcanzar niveis efectivos nas estruturas oculares, polo tanto asócianse cun un elevado risco de producir efectos secundarios. A administración tópica en forma de colirios é a máis común entre os métodos clásicos de administración de fármacos ao ollo. Trátase dun procedemento de administración non invasivo, indoloro, cun aceptable grao de cumprimento, que permite conseguir un efecto inmediato. A súa principal desvantaxe é que a biodispoñibilidade ocular dos fármacos administrados é inferior ao 5%, por mor dás barreiras escritas con anterioridade. Isto obriga a efectuar instilacións repetidas de disolucións concentradas a intervalos de tempo curtos, coas consecuentes molestias para o doente e o incremento do risco de que aparezan efectos secundarios a nivel sistémico, xa que parte da dose instilada pode acceder ao torrente circulatorio a través da conxuntiva, mucosa nasal ou tracto gastrointestinal. Entre os enfoques que se propuxeron para dar resposta ao reto de manter concentracións de fármacos suficientemente elevadas na superficie corneal e promover a penetración nas estruturas oculares internas, o uso de lentes de contacto brandas está a gañar un interese cada vez maior. As lentes de contacto están formadas por entramados poliméricos altamente biocompatibles e son usadas por máis de 100 millóns de persoas para corrixir problemas de visión. As lentes de contacto brandas Resumo 5 son hidroxeles altamente porosos e flexibles que poden absorber grandes cantidades de auga. O uso de lentes de contacto como plataformas para a cesión controlada de fármacos oftálmicos foi ideado en 1961 por Otto Wichterle e os seus colaboradores. A partires deste ano, e despois de case 60 anos, Johnson & Johnson Vision logrou a aprobación por parte das respectivas axencias reguladoras para a comercialización da primeira lente de contacto con propiedades antihistamínicas no Xapón e Canadá. Unha vez colocadas no ollo, a lente de contacto atrasa a renovación do fluido na película lacrimal posterior á lente, incrementando o gradiente de concentración de fármaco na córnea e favorecendo o fluxo no tecido. Como resultado, as lentes de contacto minimizan a perda improdutiva de fármacos ao tempo que aumentan a biodispoñibilidade ocular a máis do 50%. A personalización dos tratamentos, a mellora da adherencia do paciente o tratamento e mesmo o control da liberación de fármacos en resposta a estímulos son outras vantaxes potenciais. A pesar das vantaxes aparentes dun dispositivo de administración de fármacos baseado en lentes de contacto fronte ás gotas tópicas, transformar unha lente de contacto nunha plataforma de liberación de fármacos afronta numerosos desafíos científicos, tecnolóxicos e regulatorios. A maioría dos materiais de lentes de contacto dispoñibles carecen de afinidade polos fármacos oftálmicos. Cando unha lente de contacto se mergulla nunha disolución de fármaco, as moléculas deste difunden á fase acuosa da lente de contacto ata que a concentración tanto na disolución de carga como na fase acuosa do hidroxel este en equilibrio. Se non hai ningunha interacción ou unión do fármaco á rede de polímeros das lentes de contacto, os fármacos libéranse rápidamente cando a lente de contacto cargada de fármacos é colocada na superficie ocular. Nesta situación, non hai vantaxes da combinación de lente de contacto-fármaco en comparación cun colirio. Exploráronse estratexias ANA FILIPA PEREIRA DA MOTA 12 e cesión de pravastatina. Tamén se avaliaron os efectos da esterilización mediante alta presión hidrostática e por calor húmido en autoclave. As lentes de contacto que mellores resultados ofreceron cargaron >6 mg/g in vitro e a esterilización a alta presión hidrostática non afectou a estabilidade do fármaco nen do entramado do hidroxel. Fíxose unha análise preliminar de citocompatibilidade e tolerancia ocular con fibroblastos Balb/3T3 e un ensaio da membrana coroalantoidea de ovo de galiña fecundado (HET-CAM). A permeabilidade da pravastatina foi avaliada ex vivo usando tecidos de córnea e esclera porcina. Tamén se investigou a actividade antiinflamatoria dos hidroxeles cargados con pravastatina, mostrando capacidade anti-inflamatoria (TNF-α). A continuación, unha vez superados con éxito os ensaios preliminares de biocompatibilidade e cumprindo cos principios das 3Rs, planificouse un ensaio in vivo en coellos brancos de New Zealand para determinar a seguridade e a tolerancia ocular das lentes de contacto deseñadas e a acumulación de pravastatina nos distintos tecidos oculares. Como control empregouse unha disolucion de pravastatina coa mesma dose de fármaco que o liberado polas lentes de contacto. Esta parte do traballo levouse a cabo en colaboración co Profesor Gonzalo Carracedo en Ocupharm, Universidade Complutense de Madrid. Os protocolos foron aprobados polo Comité de Experimentación Animal da Universidade Complutense de Madrid, co correspondente permiso do Comité de Ética e cumprindo cas directrices ARVO e ARRIVE. Un obxectivo adicional dos estudos in vivo foi identificar as condicións de liberación in vitro que poidan proporcionar correlacións in vitro-in vivo (IVIVC). As probas in vivo confirmaron que, en comparación cos colirios as lentes de contacto proporcionaron niveis significativamente máis altos Resumo 13 de pravastatina na lagrima entre 1 e 7 horas despois do uso. Ademais, despois de 8 h, a pravastatina estaba presente na córnea, esclera, humor acuoso e humor vítreo. Finalmente, a construcción do gráfico de Levy co fin de por de manifesto correlacións in vivo-in vitro (IVIVC), mostrou a existencia de fortes correlacións entre as porcentaxes de fármaco liberado in vitro e in vivo. A presenza de proteínas no medio de liberación in vitro favoreceu as IVIVC. Os resultados desta segunda parte da Tese foron publicados na revista Journal of Controlled Release (DOI: 10.1016/j.jconrel.2022.06.001). 3. Predicir o comportamento in vivo de lentes de contacto medicadas mediante un modelo in vitro de ollo impreso en 3D que simula o fluxo e o pestanexo ocular A maioría dos estudos in vitro para avaliar a cesión de fármacos das lentes de contacto empregan un modelo de vial de volume fixo sen protocolos estandarizados. Como consecuencia, non se dispón dunha guía normalizada para levar a cabo os ensaios in vitro nen se ten información sobre en qué medida os resultados in vitro poden servir para predicir os resultados in vivo. Isto significa que aínda son necesarios probas in vivo en modelos animais e estudos clínicos en humanos, mesmo para avaliar produtos combinados de fármacos-lentes de contacto en fase inicial, o que fai que o desenvolvemento sexa moi custoso en tempo e recursos. Os modelos de ollos in vitro que imitan o escenario in vivo e os parámetros oculares chave teñen recibido atención nos últimos anos. Recentemente desenvolvéronse modelos de ollos in vitro impresos en ANA FILIPA PEREIRA DA MOTA 14 3D para avaliar o comportamento das lentes de contacto simulando os efectos do fluxo, volume do fluido lacrimal, a exposición ao aire e a frecuencia de pestanexo. O modelo de ollo impreso en 3D utilizado nesta terceira parte da Tese foi desenvolvido polo Profesor Lyndon Jones e o Doutor Chau Minh-Phan no CORE (Centre of Ocular Research and Education) da Universidade de Waterloo. O modelo consiste nun globo ocular feito de resina que simula a superficie ocular do ollo e unha pálpebra de alcohol polivinílico que se desliza sobre o globo ocular ou sobre as lentes de contacto adaptadas, espallando a solución lacrimal (subministrada a través dun tubo que está unido ao soporte da pálpebra). A solución de saída recóllese na unidade de recollida situada debaixo do globo ocular. No estudo desenvolvido, probouse a liberación in vitro no modelo de ollo a dous fluxos diferentes (5 e 10 µL/min) e unha velocidade de parpadeo de 1 pestanexo/10 s. Para coñecer as correlacións in vitro-in vivo, o obxectivo deste traballo foi analizar os perfís de liberación das lentes de contacto cargadas con fármacos neste modelo in vitro impreso 3D e comparar os resultados obtidos coa liberación dun vial e cos niveis de fármaco detectados no fluido lacrimal in vivo en modelo animal. Probáronse lentes de contacto cargadas con dous fármacos, pravastatina e resveratrol, de propiedades fisicoquímicas distintas. A cesión de ambos fármacos foi máis sostida e menor no modelo 3D en comparación coa liberación in vitro en viais. Ambos fármacos presentaron perfís de cesión similares no modelo 3D e in vivo, aínda que a cantidade total de fármaco liberado no modelo 3D foi significativamente menor que in vivo, especialmente para o resveratrol. Atopáronse fortes correlacións entre as porcentaxes de pravastatina liberada no modelo 3D (fluxo de 10 µl/min) e in vivo. A información recollida no presente estudo pode servir para coñecer os parámetros Resumo 15 fisiolóxicos relevantes que inflúen na liberación de fármaco a partir de lentes de contacto in vivo, como a composición do fluído lacrimal, o fluxo lacrimal, a temperatura do sistema e a composición da superficie do globo ocular. Os resultados obtidos poden servir de guía para mellorar o modelo 3D de pestanexo. Os experimentos con modelos de ollo 3D realizáronse en colaboración co Profesor Lyndon Jones e o Doutor Chau Minh-Phan no CORE (Centre for Ocular Research & Education) na Universidade de Waterloo durante una estadía de tres meses, e o correspondente manuscrito foi aceptado para publicación na revista Drug Delivery and Translational Research (DOI: 10.1007/s13346-022-01276-6). 4. Uso de melanina para comunicar as lentes de contacto de afinidade por fármacos específicos e dotalas de capacidade de resposta fototérmica A melanina é un pigmento natural que está presente en microorganismos, animais e plantas e presenta unha variedade de funcionalidades, incluíndo pigmentación, eliminación de radicais, protección contra a radiación e regulación térmica. A melanina a nivel ocular (eumelanina) actúa como un reservorio de unión química ao retener os fármacos dentro das células pigmentadas e influí na súa biodistribución. Ademais, a melanina ten unha característica importante para converter a luz absorbida en calor. O aumento rápido e local da temperatura através da melanina demostrou ser un método prometedor para matar bacterias sen risco de resistencia bacteriana o efectos secundarios. ANA FILIPA PEREIRA DA MOTA 16 A diabetes mellitus leva a cambios significativos en varios tecidos e estruturas do ollo. Os pacientes diabéticos teñen unha barreira epitelial corneal debilitada que pode levar a infeccións bacterianas da córnea e da conxuntiva. Na última parte da Tese, abordouse o obxectivo de incorporar melanina as lentes de contacto coa dobre intencionalidade de: avaliar un procedimiento alternativo de incorporar fármacos as lentes de contacto a través da afinidade fármaco-melanina, e de tirar proveito da capacidade da melanina de absorber luz e dar lugar a incrementos localizados de temperatura para conseguir efectos fototérmicos. Incorporouse melanina en discos de hidroxel de silicona e dous tipos de lentes de contacto comerciais, e realizáronse probas adicionais para identificar a presenza de melanina nestes hidroxeles. A incorporación de melanina fíxose somerxendo os discos ou lentes nunha disolución de L-DOPA (L-3,4 dihidroxifenilalanina) oxidada por tirosinasa por distintos tempos de incubación. Avaliouse a capacidade fototérmica da melanina nos discos e lentes e observouse un aumento máximo de 20 graos en 10 segundos na zona tratada nos discos incubados en melanina por 16 h. Os discos de silicona e lentes de contacto foron cargados con ciprofloxacino, un antibiótico, e a cantidade de ciprofloxacino cedida foi maior nos discos de silicona con melanina incorporada. Ademais, a cantidade de ciprofloxacino liberada dos discos de hidroxel de silicona promoveu unha inhibición significativa do crecemento de S. aureus rexistrado mediante microcalorimetría isotérmica. Resumo 17 En conxunto, as estratexias bioinspiradas desenvolvidas nesta Tese de Doutoramento permiten obter produtos combinados fármaco-lente de contacto que proporcionan niveis sostidos de fármacos na superficie ocular e tamén permiten unha terapia fototérmica factible de enfermidades do segmento anterior do ollo. Ademais, identificáronse procedementos de ensaio in vitro que imitan a cesión de fármacos na superficie ocular, o que pode facilitar a implementación de metodoloxías cas que se poidan acadar correlacións in vitro-in vivo relevantes. Esta información debe server para simplificar o desenvolvemento de lentes de contacto medicadas. INTRODUCTION The work described in this chapter was published in Testing drug release from medicated contact lenses: The missing link to predict in vivo performance, Journal of Controlled Release 343, 2022, 672-702, authored by Ana F. Pereira-da-Motaa, Chau Minh-Phanb,c, Angel Concheiroa Lyndon Jonesb,c, and Carmen Alvarez-Lorenzoa aDepartamento de Farmacología, Farmacia y Tecnología Farmacéutica, I+D Farma Group (GI-1645), Facultad de Farmacia, Instituto de Materiales (iMATUS) and Health Research Institute of Santiago de Compostela (IDIS), Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain. bCentre for Ocular Research & Education (CORE), School of Optometry and Vision Science, University of Waterloo, Waterloo, ON, Canada. cCentre for Eye and Vision Research (CEVR), 17W, Hong Kong, Science Park, Hong Kong. 21 1. INTRODUCTION The world’s first drug-releasing contact lens (CL) was recently approved by the appropriate regulatory bodies in both Japan and Canada [FDA News, 2021; Johnson & Johnson Vision Care, 2021. Although this approval was only specific to a drug (ketotifen)-CL combination, it nonetheless represents a significant milestone in the already long journey to obtain drug-CL combination products. Indeed, 2021 marked the 60th anniversary of the patent by Otto Wichterle and colleagues on the preparation of poly(2-hydroxyethylmethacrylate), pHEMA, soft CLs by the spin casting process [Wichterle, 1961]. pHEMA hydrogels revealed excellent optical properties and biocompatibility [Wichterle and Lim, 1960; Kopeček, 2009], and their moderate water content immediately caught the attention of Wichterle and Lim as a suitable compartment to host drugs, as stated in their US Patent 3,220,960 “medicinally active substances (…) may be dissolved in the aqueous constituent of the hydrogels to provide medication over an extended period” [Wichterle and Lim, 1965]. In these pioneering patents, the potential for using soft CL materials as drug delivery devices that can both act as an optical correction device and also provide therapeutic treatment for ocular diseases was implicit. However, the commercialization of both applications was not so straightforward. After the patent publication in 1961 it took Bausch + Lomb a further 10 years to bring soft CLs to the ANA FILIPA PEREIRA DA MOTA 28 provides the primary mode of action (PMOA). The agency center with the primary jurisdiction then works together with the other agency centers to carry out the adequate premarket review. In the case of a drug-releasing CL used for vision correction that simultaneously delivers a drug for an ocular disease (device-led combination product), substantial equivalence to a predicate product (e.g., the CL device solely) does not apply, since the addition of the drug results in a new intended use and/or implies a change in the technological characteristics, which in turn raises additional questions regarding safety and effectiveness [FDA, 2019a]. If the drug makes the greatest contribution, for example, a plano power bandage CL that does not correct any residual refractive error but releases a drug onto the ocular surface, then the drug-led combination product would follow the “new drug application” pathway and demonstrate the safety and the effectiveness of the product in the new conditions of use. Although examples of drug releasing CLs are not included in the FDA guidelines, illustrative examples that refer to related device-led and drug-led combination products may be useful to understand the long regulatory pathway for a drug-CL combination product [FDA, 2019a]. 1.1.2 Drug-Eluting CLs in Clinical Trials The lack of predicate products implies that the safety and efficacy of a drug-releasing CL should be demonstrated in clinical trials. The search in ClinicalTrials.gov of “contact lens AND drug” (April 2021) led to 344 outcomes. Most clinical trials referred to safety studies of new CL or drugs after ocular or oral administration when they were administered sequentially, not as a combination product. Only 16 clinical trials referred to true drug-CL combination products (Table 1. Introduction 29 1.1). In addition to the antihistamine/mast cell stabilizer ketotifen, other drugs of interest include those intended for the treatment of macular edema (dexamethasone), glaucoma and ocular hypertension (e.g., latanoprost, timolol and dorzolamide), pain management after photorefractive keratectomy, dry eye symptoms, and persistent epithelial defects. Although in most clinical studies, the changes in CL composition or technologies to prepare the drug-releasing CL are not disclosed, the application of novel technologies is evident in some cases. As an example, NCT04747808 refers to a Phase 2a study of safety, tolerability, and efficacy of CLs that have been loaded with bimatoprost by means of a proprietary printing technology (MediPrint™). The company disclosed that using an FDA approved drug and an FDA approved CL led to a shorter regulatory course via a 505 (b) (2) pathway [MediPrint™, 2022]. Of note, the average age of the five patients in the trial was 77.4 years old and none had previously worn CLs. The bimatoprost-loaded CLs were well tolerated during the seven days of continuous wearing, which supports the short-term safety of these lenses [BusinessWire, 2021]. Patient and practitioner acceptability is a critical issue for the success of drug-releasing CLs in the marketplace. Physiological conditions associated with old age and vision disorders may hinder the insertion and removal of the CLs [Novack and Barnett, 2020]. In a recent survey, most patients suffering from an ophthalmic disease indicated that they would prefer to use drug releasing CLs instead of eye drops, if the combination product was less time consuming and reduced the required frequency of application [Ghazal et al., 2018]. Another important consideration is that CL-fitters in some countries are licensed to prescribe CLs but not drugs. Therefore, it is not entirely ANA FILIPA PEREIRA DA MOTA 30 clear who would be permitted to prescribe drug-releasing CLs, and the regulations will differ between various countries. In addition to legal issues, there are still many other issues to be solved, including the stability of the drug once loaded in the CL, the expiration date, the sterilization protocol for labile drugs, assurance of sterility, disposable procedures, and how to reload the drugs [Lanier et al., 2020; Tipnis and Burgess, 2018; Pereira-da-Mota et al., 2021]. 31 Table 1.1. Clinical trials of CLs loaded with a drug or an active substance according to a ClinicalTrials.gov search for “contact lens AND drug” after refinement for “not yet recruiting, recruiting, enrolling by invitation, active not recruiting, terminated, or completed” (April 2021). Clinical Study Status Study Title Contact Lens Drug Drug Loading Condition/Outcome Measures Phase NCT04225611 Not yet recruiting Therapeutic Contact Lens Drug Delivery System (TCL-DDS) in Patients With Recurrent Cystoid Macular Edema Methafilcon (Kontur Kontact Lens Company, Hercules, CA) Dexamethasone TCL-DDS consisted of a Dexamethasonepolymer film encapsulated inside a CL Cystoid macular edema. Occurrence of CL related ocular infection; corneal epithelial defect; ocular hypertension greater than 28; changes in OCT macular thickness; visual acuity. 1/2 NCT00432757 Completed Evaluation of Efficacy and Safety of an AntiAllergy Drug with a Contact Lens in Allergic Conjunctivitis. Etafilcon A (1-Day ACUVUE, Vistakon, Florida, USA) Ketotifen CLs loaded 0.019 mg of Ketotifen Allergic conjunctivitis. Ocular itching; conjunctival, ciliary, and episcleral redness; chemosis and mucous discharge; tearing and lid swelling. 3 NCT00445874 Completed Evaluation of Efficacy and Safety of an Anti-Allergy Drug With a Contact Lens in the Treatment of Allergic Conjunctivitis Etafilcon A (1-Day ACUVUE, Vistakon, Florida, USA) Ketotifen CLs loaded 0.019 mg of Ketotifen Allergic conjunctivitis. Ocular itching; conjunctival, ciliary, and episcleral redness; chemosis and mucous discharge; tearing and lid swelling. 3 NCT04500574 Not yet recruiting Latanoprost Eluting Contact Lens for Treating Glaucoma and Ocular Hypertension Methafilcon (Kontur Kontact Lens Company, Hercules, California) Latanoprost Latanoprost eluting CLs consisted in a thin film of latanoprostpolymer film encapsulated Glaucoma. Ocular hypertension. Adverse events as assessed by ocular infection, corneal epithelial defects, or cystoid macular edema; changes in intraocular 1 1. Introduction 32 within the periphery of the lens pressure; tolerability and comfort. NCT04747808 Completed Study of LL-BMT1 in Patients With Elevated Intraocular Pressure 7-days continuous wearing CLs Bimatoprost Drug-printed CLs Primary open angle glaucoma and ocular hypertension. Adverse event rate; IOP elevation and changes. 2 NCT00889252 Completed Safety Study of a Contact Lens With Ketotifen in Healthy,Normal Volunteers - Ketotifen - Allergic conjunctivitis. Ocular itching; conjunctival, ciliary, and episcleral redness; chemosis and mucous discharge; tearing and lid swelling. 3 NCT00569777 Completed Safety Study of a Contact Lens With Ketotifen in Healthy, Normal Volunteers - Ketotifen - Allergic conjunctivitis. Ocular itching; conjunctival, ciliary, and episcleral redness; chemosis and mucous discharge; tearing and lid swelling. 3 NCT02852057 Recruiting Effectiveness and Safety of Timolol and Dorzolamide Loaded Contact Lenses Senofilcon A (ACUVUE Oasys, Vistakon, Fl, USA) Timolol maleate and dorzolamide hydrochloride CLs contained vitamin E as an additive for achieving extended release of the drugs Glaucoma. IOP changes. 1 NCT03848221 Completed Direct Application of Systane Complete to Contact Lenses Daily disposable contact lens Systane Complete; Sensitive Eyes Rewetting Drops Direct application of Systane Complete to CL Dry eye. Contact lens complication. Ocular surface damage; identifying dry eye disease. 4 NCT03026257 Completed Clinical Assessment of a HYDRAGLYDE® Regimen. Lotrafilcon B (AIR OPTIX® plus Polyoxyethylene polyoxybutylene ; CLs were packaged in a blister solution Myopia. Hypermetropia. Refractive errors. - ANA FILIPA PEREIRA DA MOTA 33 HYDRAGLYD E®, Alcon, A Novartis Division) EOBO containing the wetting agent polyoxyethylene polyoxybutylene Ex vivo total cholesterol uptake. NCT03392532 Completed Comparison of Two Silicone Hydrogel Toric Contact Lenses Lotrafilcon B (AIR OPTIX® plus HYDRAGLYD E®, and AIR OPTIX® for Astigmatism , Alcon, A Novartis Division) Polyoxyethylene polyoxybutylene ; EOBO - Astigmatism. Percentage of lenses with axis orientation within ±30 degrees from the 90 degree axis. - NCT01918410 Completed Effect of Contact Lens With Alginic Acid in Dry Eye Patients SEED 1dayPure moisture and SEED 2weekPure (SEED Co., Ltd., Tokyo, Japan) Alginic acid CLs were stored in a solution of alginic acid. Dry eyes . Difference of visual analogue scale for the ocular discomfort; tear lipid layer thickness; schirmer and tear breakup time tests; corneal fluorescein staining, and tear proteomic analysis. - NCT04283331 Recruiting Anesthetic Impregnated Bandage Soft Contact Lens (BSCL) in Pain Management After Photorefractive Keratectomy (PRK) Bandage contact lens (BCL) Proparacaine The BSCL were soaked in proparacaine hydrochloride 0.5% Myopia . Hypermetropia. Refractive errors. Astigmatism. Daily pain score; complete re-epithelialization and final refraction at postoperative. 4 NCT03388138 Completed Clinical Evaluation of Etafilcon A With Etafilcon A (1 - Day Ketotifen CLs loaded 0.019 mg of Visual acuity . 2 1. Introduction 34 Ketotifen ACUVUE, Vistakon, Florida, USA) ketotifen Monocular contact lens corrected distance visual acuity; eyes with clinically significant slit lamp findings and with unacceptable lens fitting. NCT03653650 Recruting Autologous Platelet - rich Plasma in the Treatment of Persistent Epithelial Defects Bandage contact lens (BCL) Autologous plateletrich plasma Bandage CLs plus autologous platelet-rich plasma eye drops Persistent epithelial defect . Persistent epithelial defect healing time; change in corneal sensitivity; uncorrected visual acuity, best corrected visual acuity; ocular pain; ocular surface symptoms, and frequency of adverse effects. - NCT04553432 Recruting Dry Eye OmniLenz Application of Omnigen Research Study (DOORS) OmniLenz® (NuVision®, Nottingham, UK) Omnigen (Amnionic membrane) OminLenz allows easy delivery and comfortable retention of Omnigen at the ocular surface Dry eye . Change in dry eye symptoms, visual acuity, meniscus height, noninvasive tear breakup time, ocular surface staining and ocular redness. 4 ANA FILIPA PEREIRA DA MOTA 1. Introduction 35 1.2 GUIDELINES AND PHARMACOPOEIA METHODS FOR TESTING OCULAR DRUG RELEASE The safety and efficacy of a drug product or a drug-device combo product are ultimately demonstrated in a clinical trial. However, studies with human participants are the final step after numerous preclinical studies involving laboratory testing and a variety of in vivo tests. In vivo testing in animal models should follow very strict ethical rules and adhere to the 3R´s principles; namely, ‘replace’ (search for alternatives), ‘reduce’ (minimize the experiments), and ‘refine’ (avoid distress) [European Commission, 2021]. Moreover, animal testing can be expensive and not suitable for early-stage testing. In vitro drug release tests are very valuable for quality assurance since they reveal the robustness and reproducibility of a drug product. Additionally, these tests help identify the critical elements of a drug-device that are important for in vivo performance [Kostewicz et al., 2014]. Therefore, in vitro models that better predict the performance of these devices may significantly facilitate the research and development process, thereby reducing both time and cost to bring these products to market. 1.2.1 Determining Drug Bioavailability from Drug-CLs using Mathematical Models The advantages of drug-releasing CLs as ocular drug delivery systems are commonly compared with conventional eye drop administration. A simple mathematical model to explain the concentration of drugs on the ocular surface following an eye drop instillation was developed by Lang and Stiemke [Lang and Stiemke, 1996]. This model assumes that the tear fluid layer on the ocular surface behaves as a continuously stirred reactor (Figure 1.1). Once an eye drop ANA FILIPA PEREIRA DA MOTA 36 is deposited on the eye, the volume that cannot be contained in the tear layer spills over, and the rest of the volume is rapidly mixed with the tear fluid (Figure 1.2). If no overflow occurs (i.e., drop volume is very small), the initial drug concentration on the ocular surface (CR(0)) can be estimated as the ratio of the dose applied (Qi) to the volume of the tear fluid layer (VR). Since there is a hydrodynamic flow of tear fluid from temporal to the nasal segments of the eye (𝑉󰇗 , the flow entrance of new tear fluid is the same as the flow of drainage) [Doane, 1989], the total amount of drug in the tear layer decreases exponentially according to 𝑄(𝑡)=𝑄·exp󰇡− 󰇗 ∙𝑡󰇢 (Eq. 1.1) In this equation QR(t) represents the amount of drug at time t in the tear fluid, Qi the initial amount (dose) instilled, 𝑉󰇗 the tear flow, and VR the volume of tear fluid. Similarly, drug concentration in the tear fluid layer decreases along time as follows 𝐶(𝑡)=() =𝐶(0)·exp󰇡− 󰇗 ∙𝑡󰇢 (Eq. 1.2) In addition to drainage through the lacrimal puncta, tear film losses occur due to evaporation and absorption through the cornea, but the contributions of these two factors under healthy conditions are limited to 15-20% of total tear fluid losses [Joshi et al., 1996]. In other words, the amount of drug that is transported from the eye mostly enters into the nasolacrimal duct at a rate of 𝑄(𝑡)=𝑄· 󰇡1 − exp (− 󰇗 ∙𝑡)󰇢 (Eq. 1.3) 1. Introduction 37 These equations do not consider drug penetration into eye tissue, but for most drugs that show poor ocular bioavailability, the model is still valid. Assuming that the drug penetrates through all ocular tissues by diffusion, the flux of drug from the tear fluid layer (J) depends on drug diffusion through the eye tissue (D), the oil-water partition coefficient (K), the drug concentration (CR(t)), and the thickness of the diffusional barrier (h) as follows 𝐽 = ∙𝐶(𝑡) (Eq. 1.4) Thus, the rate of absorption can be calculated as 𝑄󰇗(𝑡)=𝐽∙𝐴= ∙𝐶(𝑡) (Eq. 1.5) The amount absorbed at a given time is estimated by integration to be () ∙󰇗∙()· 󰇗·() (Eq. 1.6) The typical values of the permeability coefficient, P, of ophthalmic drugs are quite low [Loch et al., 2012], and therefore the fraction absorbed is predicted to also be minimal. An example simulating the fate of the drug after instillation of 50 μL of a 0.1% drug solution is shown in Figure 1.1. Assumptions included that only 10 μL are effectively mixed with the tear fluid (Qi = 0.01 mg), the drug has a quite high permeability coefficient (0.5⋅10-6 cm/s), the corneal surface available for absorption is 0.5 cm2 [CavasMartínez et al., 2014], and that the tear flow is 1.2 μL/min [Lang and Stiemke, 1996]. Figure 1.1 evidences the rapid clearance of the drug from the tear fluid layer and the small amount that can penetrate the ocular tissues. This model also explains that a drug-induced tearing ANA FILIPA PEREIRA DA MOTA 44 performance. Particularly, for ophthalmic non-solution products, traditional systemic pharmacokinetics studies (using blood sampling) do not inform on bioavailability at the site of action (inner eye structures) [Luke and Kozak, 2021]. Regulatory bodies have published recommendations on bioequivalence studies that involve clinical endpoint studies (mostly for glaucoma, ocular hypertension, and pain), pharmacokinetic studies in aqueous humor (e.g., topical corticosteroids prior to cataract extraction), microbial kill rate assay (e.g., antimicrobial drugs) and in vitro release studies [Choi and Lionberger, 2016]. Clinical endpoints provide semiqualitative information and may be affected by the patient health conditions. Since their capability to discriminate similar products is low, clinical endpoint assessment requires a fairly large sample size. Pharmacokinetic studies in the aqueous humor offer quantitative data of transcorneal ocular bioavailability, but usually only one sampling per patient is feasible. Moreover, drug levels in the aqueous humor may be affected by several interindividual differences (e.g., age, ethnicity, ocular illness) and thus a very large sample size is required for statistical significance [Luke and Kozak, 2021]. Therefore, so far, in vitro studies are considered more reliable to assess differences among ophthalmic formulations. Researchers and regulatory bodies have been attempting to establish methods for in vitro-in vivo correlations (IVIVC). Strong efforts are being made to develop ocular physiologically relevant pharmacokinetic (PBPK) in silico models (i.e., via computer simulation) that can predict drug ocular bioavailability and tear film breakup time based on the physicochemical properties of the drug and the formulation. For instance, GastroPlusTM has recently been implemented with an Ocular Compartmental Absorption and Transit (OCAT™) model [FDA, 2019b]. Various other modeling approaches 1. Introduction 45 to simulate drug transport through the cornea are also currently under investigation [Pak et al., 2018]. Some product-specific guidances are available for ophthalmic nonsolution products, such as suspensions and ointments. A great deal of attention has been focused on identifying drug release tests that can detect changes in formulation or manufacturing process having the same drug and the same inactive ingredients. The most common apparatus to monitor drug release from non-solution ophthalmic products are those included in official pharmacopeias for oral dosage forms, with some modifications to be adapted to semisolid products. Some key examples of the apparatus include the Franz diffusion cell, USP apparatus 2 with enhancer cells, and USP apparatus 4 with semisolid adapters (described in Figure 1.4) [Bao and Burgess, 2018]. Nevertheless, non-compendial dissolution methods that can be more biorelevant are under investigation, such as devices simulating lowvolume flow-through [Luke and Kozak, 2021]. ANA FILIPA PEREIRA DA MOTA 46 Figure 1.4. (A) Scheme of some of the apparatus considered in the USP chapter <1724> for in vitro drug release from semisolid drug products. Franz (vertical) diffusion cells consists of two compartments (donor and receptor) separated from each other by a membrane onto which, typically, 200 mg of the product under test is placed. A heating jacket is used to regulate the temperature and the experiment is carried out for 4-6 h. The modified USP apparatus 2 consists of a cell filled with the product (300 mg - 2 g covered with a permeable membrane) which is placed at the bottom of the glass flat vessel. The release medium is added and maintained under stirring with the help of a small paddle positioned at a certain distance from the product. The flow-through cell USP apparatus 4 consists of a reservoir (semisolid adapter) filled with the product (also covered with a permeable membrane) through which a sinusoidal flow of release medium is impelled. Reprinted from Bao and Burgess [Bao and Burgess, 2018] by permission from Springer Nature; (B) Linear correlation (semi-logarithmic mode) between critical parameters of loteprednol etabonate ophthalmic ointments (crossover modulus CM and power law consistency index K) and drug transcorneal flux; and (C) linear correlation between in vitro release rate recorded using the three apparatus described in (A) and ex vivo transcorneal flux (receptor filled with 5 mL of artificial lachrymal fluid with 9% hydroxypropyl-β-cyclodextrin, 600 rpm). Reprinted from Bao et al. [Bao and Burgess, 2018] with permission from Elsevier. 1. Introduction 47 There is an increasing number of reports devoted to identifying the apparatus and the release conditions that allow for improved discrimination among similar ophthalmic formulations of a given drug [Bao et al., 2017]. In vitro release tests in simulated lachrymal fluid and ex vivo transcorneal flux correlations have also been attempted. As an example, four loteprednol etabonate ophthalmic ointments considered as qualitatively and quantitatively equivalent formulations (the same components in the same concentration; Q1/Q2) were evaluated in terms of rheological behavior, in vitro drug release in artificial lachrymal fluid and ex vivo (rabbit) transcorneal flux [Bao et al., 2018]. The USP apparatus 4 (1.54 cm2 exposed area, 50 mL medium capacity, flow rate 8 mL/min) clearly showed differences in release rate among the formulations, which were not detected or were less evident using the Franz diffusion cells (1.77 cm2 exposed area, 12 mL receptor medium, 600 rpm) or the USP apparatus 2 with enhancer cells (4 cm2 exposed area, 40 mL release medium, 150 rpm). The highest release rates, but the less reproducible ones, were recorded in the Franz diffusion cells. Inverse correlations were found between the rheological properties (crossover modulus and power law consistency index) and the release rate (Higuchi model) of the four ointments. Moreover, ex vivo transcorneal permeation coefficients recorded in Franz diffusion cells also revealed differences among the formulations, which correlated with their rheological features (Figure 1.4A). Compared to the drug release rates calculated in vitro, the drug permeation rates were lower, but the rank order of the formulations was the same; namely, a direct correlation was found between in vitro release rate and ex vivo release flux (Figure 1.4B). This study demonstrated the usefulness of the in vitro release tests as predictive tools of ex vivo drug transcorneal flux, which may correlate with the flux in vivo, although this has not been demonstrated yet. Additionally, particular attention should be paid to ANA FILIPA PEREIRA DA MOTA 48 the preservation of corneal structure and epithelium integrity during the test [Bao et al., 2018]. In any case, such in vitro-ex vivo correlations are particularly useful when a predicate product has already demonstrated safety and effectiveness, and a generic product is intended to be developed. Nevertheless, prediction of in vivo performance does not depend only on the release rate, but also on other formulation properties, such as retention time on the ocular surface. Release tests carried out with solid ophthalmic formulations, namely inserts to be placed in the conjunctival fornix, are also quite heterogeneous [Del Amo and Urtti, 2018; Bertens et al., 2018; Miller and Eaton, 2021; Saettone and Salminen, 1995]. Ocusert-like inserts prepared with poly(lactic coglycolic) acid and polyethylene glycol (PEG) and loaded with brimonidine tartrate showed almost constant release rate for one month when placed in phosphate buffered saline (PBS) solution at 37 °C [Mealy et al., 2014]. The release medium was completely replaced at each sampling point, but total release medium volume and stirring conditions were not disclosed. There are few other inserts for drug delivery that are under preclinical or clinical phase. A bimatroprost ocular ring (formerly HeliosTM), which contains the drug onto a non-biodegradable support and coated with silicone, was reportedly ‘safe’ in preclinical tests, but the in vitro release profiles were not disclosed [Miller and Eaton, 2021]. Ciprofloxacin release from OphthaCoil inserts (stainless steel coated with SlipSkin®) was evaluated by placing the insert in a silicon tube (inner diameter of 1 mm), through which simulated lachrymal fluid was pumped at a rate of 100 μL/min. Fractions of 150 μL were collected for release kinetics evaluation. In general, the profiles showed fast release in the first minutes and decreased amount released in the next few hours [Pijls et al., 2007a]. Pradofloxacin-loaded OphthaCoil inserts were similarly tested in vitro and compared with the release in vivo in a dog model 1. Introduction 49 after insertion in the lower conjunctival fornix of the eye [Pijls et al., 2007b]. The in vivo release was notably slower and thus more prolonged, which was attributed to the in vitro high flow of release medium (100 μL/min) compared to physiological tear turnover (1-5 μL/min). Overall, the information reported for topical ocular solid formulations also reveals a disparity of release media and testing conditions, without conclusive proposals on the in vitro test setup that may mimic in vivo release. The lack of standardized in vitro test models has also been pointed out as a relevant concern for the development of ocular implants for posterior eye segment [Braga et al., 2010]. A variety of setups, including configurations similar to those depicted in Figure 1.4, have been tested, but finding IVIVC remains elusive. The high complexity of the ocular tissues, comprising a diversity of metabolic and physiological barriers, makes the in vitro mimicking of the in vivo processes of drug release, absorption and distribution much more challenging than for any other route. 1.3 IN VITRO TESTING OF DRUG RELEASE FROM MEDICATED CLS Since there is not an officially approved method for testing drug release from CLs, authors have used a myriad of different approaches. A search was performed in Web of Science database with the keywords “drug AND contact lens” within the 2000 to 2021 time frame. The search was further refined using the word “release” and to papers published in English. In total, 455 results were generated. This search included ocular drugs and macromolecular demulcents. Further refinement manually removed contributions that did not match the ANA FILIPA PEREIRA DA MOTA 50 appropriate outcomes searched for, resulting in 251 original papers being analyzed. The in vitro setups can be categorized into three main groups (Figure 1.5): (i) release in a beaker with or without replacement of the release medium at a pre-stablished time; (ii) release under fluid flow in specially designed microfluidic chambers; (iii) release under blinking-mimic conditions Most reports refer to in vitro release profiles recorded in small beakers, but using widely variable volume, medium composition, stirring and replacement. For exhaustive information the reader is referred to Table 1.3. Although analysis of this non-harmonized data is not easy, beakers in which the release medium is not replaced are usually filled with more volume than those that undergo partial or complete replacement of the release medium at each sampling time (Figure 1.5A1). The selection of the release medium volume is a highly relevant issue, as evidenced in a previous comprehensive report [Tieppo et al., 2014]. If a too small volume is chosen, then the release process may become controlled by drug solubility, and therefore the capability of the CL to regulate drug release is overestimated. In comparison, if too large a volume is chosen or a small volume is frequently replaced, then a sink effect accompanied by a large concentration gradient may forcedly accelerate the release process, and the CL may become exhausted much more rapidly than under in vivo conditions. Clearly, CLs that are able to sustain the release under the in vitro more challenging conditions are also expected to control the release under in vivo conditions. However, as evidenced in section 1.4, these in vitro conditions do not ensure IVIVC because the release in vivo may be 1. Introduction 51 slower. Interestingly, an overall analysis of the time that CLs sustain drug release in vitro suggests that, disregarding whether the release medium is replaced or not, conventional commercially available CLs release the drug much faster than CLs designed ad hoc or coated with components that exhibit affinity for the drug. According to the data compiled in Table 1.3, for these latter CLs, the duration of the release could be prolonged for 149 (s.d. 99), 319 (s.d. 338), 226 (s.d. 278) and 290 (s.d. 328) hours in PBS without replacement, in PBS with replacement, in artificial tears without replacement, and in artificial tears with replacement, respectively. The box plot showing the median values and the upper and lower quartiles are depicted in Figure1.5A2 and A3. No in vivo release profiles are available for reports collected in Table 1.3. Attempts to better mimic the complex scenario of in vivo release have mainly focused on the design of microfluidic chambers that allow for regulation of both the flow and the total volume of medium that bathes the CL (Figure 1.5B). The microfluidic device depicted in Figure 1.5B1, which may hold 175 μL in the inner chamber and pumps the release medium at a flow rate of 3 μL/min, has been shown to be useful for discrimination of the capability of silicone hydrogel CLs to sustain dexamethasone release [Kaczmarek et al., 2014]. Compared to the release in a large-volume beaker, CLs loaded with a variety of drugs and demulcents showed slower release profiles when tested in the microfluidic device, which were prolonged for weeks [Tieppo et al., 2012a; White et al., 2016; White et al., 2011]. The microfluidic device depicted in Figure 1.5B2 consisted of two 3D printed molds resembling the corneal/scleral curvature and the eyelid. The CL is placed between these structures, and the space available for the release medium is 100 μL. For a flow rate of 3.3 μL/min, commercially available CLs sustained the release of ciprofloxacin, moxifloxacin and fluconazole for ANA FILIPA PEREIRA DA MOTA 52 several hours compared to the few minutes recorded when the test was carried out in a vial with 4.8 mL PBS, partially replaced at each sampling time [Bajgrowicz et al., 2015; Phan et al., 2016a]. The microfluidic device depicted in Figure 1.5B3 is the one with the lowest volume in the inner chamber (45 μL, which is quite close to the maximum tear film volume in the eye) and has 8 outlets that converge in a common collector [Pimenta et al., 2016a]. Once again, release profiles recorded using this microfluidic device showed remarkably slower rates than the release in a beaker [Alvarez-Rivera et al., 2019; Silva et al., 2021]. At this time, the microfluidic devices are still to be validated with appropriate in vivo data. The reports on devices that can mimic the pressure exerted by the eyelid on the CL are still incipient (Figure 1.5C). Repetitive load and friction cycles (16 kPa) onto CLs placed in a Simublink device (Figure 1.5C1) have been shown to accelerate the release of levofloxacin when directly loaded in the bulk of the CL [Galante et al., 2015], but the effect was negligible when the drug was encapsulated in liposomes coating the CL [Paradiso et al., 2017]. A whole in vitro eye model device constructed using 3D printing has recently been proposed to evaluate the effects of air exposure, flow rate, and blinking frequency on drug release rate (Figure 1.5C2) [Phan et al., 2021; Phan et al., 2016b; Phan et al., 2019a]. The model considers the coating of the front surface of the eyeball with a silicone material to prevent unspecific binding of the tested drug and allows for quantifying the amount of drug that penetrates into the polyvinyl alcohol (PVA)-mimicking eyelid [Phan et al., 2019a]. 1. Introduction 53 Figure 1.5. Main setups used to evaluate in vitro drug release from CLs. (A) Beaker of variable volume (from small well to large tube) with (R) or without (w/oR) replacement of the release medium at a pre-established time. The box plots constructed from the data compiled in Table 1.3 demonstrate that (A1) tests made without replacement of the release medium (w/oR) usually involve large volumes, and (A2 and A3) CLs designed to control drug release (code C+) are more efficient in providing sustained release than conventional CLs (code C-) disregarding whether the release medium (PBS or artificial tears) is replaced or not during the in vitro test. (B) Devices proposed to evaluate drug release under fluid flow showing different configurations for inlet and outlet ports; (B1), (B2) and (B3) reprinted, respectively, from Tieppo et al. [Tieppo et al., 2012a] with permission from Elsevier, Bajgrowicz et al. [Bajgrowicz et al., 2015] with permission from ARVO, and Pimenta et al. [Pimenta et al., 2016a] with permission from Springer Nature. (C) Advanced prototypes in which CLs can be subjected to normal forces that may mimic the blinking conditions; (C1) reproduced from Galante et al. [Galante et al., 2015] by permission of Taylor & Francis Ltd., and (C2) reproduced from Phan et al. [Phan et al., 2021] (Creative Common CC BY license). 60 CLs was quantified 12 Bimatoprost Siloxane (100 μL), ethyleneglycol dimethacrylate (15 μL), dimethyl acrylamide (300 μL), and hydroxyl ethylmethacryla te (HEMA, up to 1 ml) Soaking in bimatoprost microemulsion (ME) or solution (SM) containing 25, 50 or 75 μg drug per mL of simulated tear fluid for 7 days 2 mL of simulated tear fluid at 34 ºC and 50 rpm; complete replacement of the release medium at each sampling point New Zealand rabbits (male and female). SM (32.6 µg) and ME (46.4 µg) CL on one eye. As control, one drop (50 μL) of 0.03% w/v bimatoprost eye drop solution was instilled on one eye. Contralateral eyes were referred as control. In vitro, SM and ME CLs showed a relevant burst and sustained drug release for 24 h and 48 h, respectively. In vivo, the release in tear fluid was prolonged for 12 and 24 h, respectively. The eye drop was cleared in less than 1 hour. Xu et al., 2019 13 Travoprost Silicone hydrogel CLs made of dimethyl acrylamide (250 µL), siloxane (150 µL), EGDMA (10 µL), and hydroxyethyl methacrylate (HEMA, up to 1000 µL) Soaking in drug microemulsion (TME) or solution (T-SM) containing 25, 50 and 75 μg drug per mL of simulated tear fluid for 10 days 2 mL of simulated tear fluid at 34 ºC and 100 rpm; complete replacement of the release medium at each sampling point New Zealand rabbits (male and female). T-SM (13.9 µg) and T-ME (26.9 µg) CL on one eye. As control, one drop (50 μL) of 0.003% w/v travoprost eye drop solution was instilled on one eye. Contralateral eyes were referred as control. In vitro and in vivo, TSM CL and T-ME CL sustained drug release for 24 and 48 h, respectively. Eye drops showed a rapid decay in drug levels in the first two hours. Xu and Liu, 2020 14 Olopatadine EGDMA (10 L), siloxane (100 L), and HEMA (up to 1 mL) Soaking (SM-OL), drug directly added during polymerization (DL-OL), or drug encapsulated in ethylcellulose microprarticles in 2 mL of simulated tear fluid at 34 ºC and 100 rpm; complete replacement of the release medium at New Zealand rabbits of either sex. One DNT-OL CL (260 µg dose). As control, one drop (50 µg olopatadine HCl) was instilled on one eye. Contralateral In vitro, DNT-OL CL attenuated the burst but most drug was still released in the first 12 h. In vivo, olopatadine levels in tear fluid lasted few hours after eye drop Xue et al., 2020 ANA FILIPA PEREIRA DA MOTA 61 doughnut CLs (DNT-OL) each sampling time eyes were referred as control. instillation and were prolonged for 24 h with the CLs. 15 Betaxolol HCl Silicone hydrogel CLs made of HEMA:NVP:TRIS 20:40:40 w/w/w Drug loaded pH - responsive film (cellulose acetate-Eudragit S100) embedded in the CL 10 mL of simulated tear fluid or PBS at 35 ºC and 100 rpm; 2 mL replacement of the release medium at each sampling point Male Nippon albino rabbit. Soaked CL (121 µg dose), filmembedded CL (700 µg dose), or one eye drop (100 µg dose) on one eye. Contralateral eyes were referred as control. In vitro , drug release was minimum in PBS pH6.8 and prolonged for 10 h in simulated tear fluid. In vivo, drug levels in tear fluid were measurable for several days. IVIVC found. Zhu et al., 2018a 16 Diclofenac sodium HEMA and EGDMA (0.5 %) Drug loaded pH - responsive film (ethylcelluloseEudragit S100) embedded in the CL 10 mL of simulated tear fluid or PBS at 35 ºC and 100 rpm; 2 mL replacement of the release medium at each sampling point Male Nippon albino rabbit. Soaked CL (100 µg dose), filmembedded CL (121 µg dose), or one eye drop (150 µg dose) on one eye. Contralateral eyes were referred as control. In vitro , drug release was minimum in PBS pH6.8. The release rate in simulated tear fluid depended on thickness and molecular weight of polymer film. In vivo, drug levels in tear fluid were measurable for 12 h. IVIVC found. Z hu et al., 2018b 17 Betaxolol HCl Silicone hydrogel CLs made of HEMA:NVP:TRIS 20:40:40 w/w/w Drug loaded ion - responsive film (poly(styrenedivinyl benzene) in cellulose acetate) embedded in the CL 10 mL of simulated tear fluid at 35 ºC and 100 rpm; 2 mL replacement of the release medium at each sampling point Male Nippon albino rabbit. Filmembedded CL (700 µg dose), or one eye drop of drug-resin complex (100 µg dose) on one eye. Contralateral eyes were referred as control. In vitro , drug release was sustained for one week. In vivo, drug levels in tear fluid were measurable for several days. The resin eye drops remained for 8h in tear fluid. IVIVC found. Zhu et al., 2018c 1. Introduction 62 18 Timolol ACUVUE® TruEye™ (narafilcon A) silicone hydrogel contact lenses Soaking in 3 mL of 20 or 50 mg/mL vitamin E in ethanol for 24 h, washed with water, and then placed in 3.5 ml of 8.0 mg/ml timolol maleatePBS solution for 21 days. CLs without vitamin E were soaked in 2.5 mg/mL drug solution for 7 days 2 mL of PBS (room temperature) Beagle dog model of glaucoma. One eye drop (150 g) was applied to one eye twice a day for 4 days (total amount of timolol delivered 1200 g). The other eye was referred as control. One control CL (60 g dose) to be worn for 24 h, and replaced daily with a freshly drugloaded CL for 4 days. One control CL (200 g dose) to be worn for 4 days. One drug-loaded vit E-pretreated CL (200 g dose) to be worn for 4 days. In vitro, vit Epretreated CLs sustained drug release up to 84 h compared to the 4 h of control CLs. In vivo, vit Epretreated CLs were more efficient in regulating IOP than control CLs wore for 4 days. Peng et al., 2012a 19 Timolol and dorzolamide Senofilcon A Soaking in 3 mL of 40 mg/mL vitamin E in ethanol for 24 h, washed with water, and then placed in 3.5 ml of 0.8 mg/ml timolol maleatePBS solution or 3.5 ml of 0.75 mg/ml dorzolamide hydrochloride 2 mL of PBS (room temperature) Beagle dog model of glaucoma. One eye drop (205 g of timolol and 670 g of dorzolamide) was applied to one eye twice a day for 4 days. The other eye was referred as control. One drug-loaded commercial CL Dually-loaded commercial CLs released 90% timolol in 1.2 h, and 90% dorzolamide in 3 h. Vit E-pretreated CLs sustained drugs release for 42 h. Dually loaded CLs exhibited superior IOP reduction compared to eye drops with 4Hsu et al., 2015a ANA FILIPA PEREIRA DA MOTA 63 solution (4 days). For dual loading, vit E pretreated CLs were soaked in 3.5 ml of PBS containing timolol (12.75 mg/ml) and dorzolamide (20 mg/ml) (non - pretreated with vitamin E) (60 g of timolol and 220 g of dorzolamide) was worn for 24 h and replaced daily for 5 days. One drug-loaded vit E-pretreated CL (200 g of timolol and 680 g of dorzolamide) was worn for 48 h and replaced once with a similar CL. The treatment was stopped at 96 h. to 6 - fold lower drug loading. Continuous wear of dually-loaded vitamin E-pretreated CLs reduced IOP during the 4 days of wear time and for another 8 days after removal of the CLs. 20 Timolol and bimatoprost EGDMA (10 L), DMA (310 L), NVP (10 L), siloxane (100 L), and HEMA (up to 1 mL) Three small implants loaded with timolol (100 mg), bimatoprost (75 mg) and HA (60 mg), respectively, were included in the CLs 2 mL of simulated tear fluid at 34 ºC and 100 rpm; complete replacement of the release medium at each sampling time New Zealand rabbits of either sex. One CL on one eye (50 µg bimatoprost and 85 µg timolol). As control, one drop (15 µg bimatoprost and 250 µg timolol) was instilled on one eye. Contralateral eyes were referred as control. In vitro and in vivo , implant-loaded CLs sustained drug release for 24 h and lowered IOP for 72 h. CLs loaded by soaking showed higher burst release both in vitro and in vivo, although therapeutic levels were recorded in tear fluid for 12 h and low IOP was maintained 48 h. Desai et al., 2020 21 Timolol Acuvue Oasys Soaking for 24 h in 3% (w/w) dispersion of timolol - loaded Storage in 1 ml of packaging solution (PBS) Beagle dog model of glaucoma. One CL in one eye. The other In vitro , freshly loaded CLs sustained drug release for two weeks. CLs stored in Jung et al., 2013 1. Introduction 64 propoxylated glyceryl triacrylate nanoparticles in ethanol for 2 weeks at 4 ºC. Release was then tested in 1.753.5 mL PBS at room temperature and 40 ºC eye was referred as control. packaging solution evidenced leakage of drug-loaded nanoparticles, which led to lower amount released. In vivo, decrease in IOP was observed on days 2, 3 and 4 of wearing. 22 Timolol HEMA-based ring containing timolol-ethyl cellulose nanoparticles and sandwiched in HEMA-CL During synthesis 2 mL of simulated tear fluid at 34 ºC under shaking; complete replacement of the release medium at each sampling time New Zealand rabbits of either sex. One CL with ring (150 µg dose) placed on one eye. As control, one drop (250 µg) was instilled on one eye. Contralateral eyes were referred as control. In vitro, the CLs sustained drug release for two days. In vivo, CL showed Cmax of 6.79 µg/mL in 5 min, followed by steady release for several days. The drop led to Cmax of 132.6 µg/mL in 5 min and rapid decrease of drug in tera fluid. Maulvi et al., 2016b 23 Timolol and hyaluronic acid HEMA (669 L), EGDMA (5 L), DMA (310 L), TRIS (1 L), NVP (10 L) Semi-circular implants containing timolol or HA 2 mL of simulated tear fluid at 34 ºC and 100 rpm; complete replacement of the release medium at each sampling point New Zealand white rabbits of either sex. One CL (148 μg timolol) for 7 days wearing. One drop of 0.5% timolol maleate (250 μg timolol). Contralateral eyes were referred as control. CLs sustained the release of timolol and HA for 96 h in vitro and 72 h in vivo and decreased IOP during 144 h. Desai et al., 2018 24 Timolol base and latanoprost HEMA (580 L) and EGDMA (15 L) Micelles containing both drugs (0.4 mL) Franz diffusion cell. Donor compartment Male Nippon albino rabbits, healthy and glaucoma model. One CL (100 μg In vitro, CLs sustained timolol and latanoprost release for up to 120 h and 96 Xu et al., 2019 ANA FILIPA PEREIRA DA MOTA 65 were added to the monomers with one CL and 1 mL STF, and receptor with 7 mL STF at 35 ºC and 50 rpm; 1 mL replacement of the release medium at each sampling point. timolol and 1 μg latanoprost) on one eye. As reference, 50 μL of Xalacom® eye drops (250 μg timolol and 2.5 μg latanoprost) were instilled. Contralateral eyes were referred as control. h, respectively. In vivo, CLs increased MRT (79.6-fold and 122.2-fold) and bioavailability (2.2fold and 7.3-fold) for both timolol and latanoprost compared with eye drops. IOP reduction over 168 h. 25 Timolol base Silicone hydrogel CLs made of dimethyl acrylamide (350 µL), siloxane (100 µL), and hydroxyethyl methacrylate (HEMA, up to 1000 µL) Soaking in timolol microemulsion (TB-ME-SM) or timolol solution (TB-SM) containing 1, 2 or 3 mg drug per mL of simulated tear fluid for 10 days 2 mL of simulated tear fluid at 34 ºC and 50 rpm; complete replacement of the release medium at each sampling point New Zealand rabbits (male and female). TB-SM (234.3 ± 18.5 µg) and TB-ME-SM (215.3 ± 9.1 µg) CL on one eye. As control, one drop (50 μL) of 0.5% w/v timolol eye drop solution was instilled on one eye. Contralateral eyes were referred as control. In vitro , TB - SM CL and TB-ME-SM CL released more than 90% drug in 12 h and in 48 h, respectively. In vivo, TB-SM CL and TB-ME-SM CL had Cmax at 5 min and provided measurable drug levels in tear fluid for 24 and 72 h, respectively. Eye drops showed a rapid decay in timolol levels in the first two hours. In vivo efficacy was evaluated regarding intraocular pressure (IOP). TB-ME-SM CL showed prolonged reduction in IOP values. Wei et al., 2021 26 Sparfloxacin EGDMA (10 L), siloxane (100 Immersion in 2 mL of drug solution (2 - 6 mg/mL) in 2 mL of simulated tear fluid at 34 ºC New Zealand rabbits of either sex. One CL with ring (129 µg In vitro , the ring - loaded CLs were the only able to sustain Ran et al., 2020 1. Introduction 66 L), and HEMA (up to 1 mL) 0.5% PVPsimulated tear fluid medium, autoclaving and soaking for 7 days. Alternatively, a drug-loaded ring was adapted to the CL under shaking; complete replacement of the release medium at each sampling time drug) placed on one eye. As control, one drop (150 µg) was instilled on one eye. Contralateral eyes were referred as control. Efficacy in conjunctivitis model. drug release for two days. In vivo, the ringloaded CLs provided therapeutically useful drug levels in tear fluid for 12 h and favored conjunctivitis treatment. 27 Ofloxacin Silicone hydrogel CLs made of dimethyl acrylamide (250 µL), siloxane (100 µL), and hydroxyl ethylmethacryla te (HEMA, up to 1000 µL) Soaking in ofloxacin microemulsion (Of-ME) or ofloxacin solution (Of-SM) containing 1, 2 or 3 mg drug per mL of simulated tear fluid for 7 days 2 mL of simulated tear fluid at 34 ºC and 100 rpm; complete replacement of the release medium at each sampling point New Zealand rabbits (male and female). Of-SM (191 µg) and Of-ME (358 µg) CL on one eye. As control, one drop (50 μL) of 0.3% w/v ofloxacin eye drop solution was instilled on one eye. Contralateral eyes were referred as control. In vitro, Of-SM and Of-ME CLs showed a relevant burst and sustained drug release for 24 h and 72 h, respectively. In vivo, the release in tear fluid was prolonged for 24 and 48 h, respectively. The eye drop was cleared in less than 1 hour. In vivo efficacy was tested against Staphylococcus aureus-induced conjunctivitis. Of-ME CL improved the symptoms in 24 h. Complete healing was observed after 4 days of treatment with either one CL or 0.3% w/v ofloxacin eye Li et al., 2020 ANA FILIPA PEREIRA DA MOTA 67 drop solution instilled every 4 h. 28 Gatifloxacin HEMA with methacrylic acid (MAA) (25:1 mol/mol) Soaking in 0.5 mg/L drug solution in 0.9% NaCl medium at 37 ºC until equilibrium Non - disclosed volume of 0.9% NaCl medium Sprague Dawley rats. Keratitis was induced on the right eye. Drug-loaded CLs (50 mg) were placed on the cornea and the eyelids were sutured. Saline and drug drops (5 μL; unknown concentration) were instilled every 4 h and used as controls. In vitro P(HEM A - co - MAA) (11.8 µg/mg) released 70% drug in 24 h. In vivo efficacy in a rat model of bacterial keratitis. Wearing for 48 h of P(HEMA-coMAA) favored the healing of cornea lesions caused by epithelial erosion and stromal ulceration, more efficiently than d rug drops. Shi et al., 2013 29 Gatifloxacin Dimethyl acrylamide (31%), siloxane (2.5%), NVP (1%), EGDMA (1%) and hydroxyl ethylmethacryla te (HEMA, up to 1000 µL) Drug directly added to the monomers solution, following by autoclaving and storage in 0.3% drug solution in Pluronic micelles packaging solution (GT-PL-CL). For comparison, CLs were soaked in the drugcontaining packaging solution 2 mL of simulated tear fluid at 34 ºC and 100 rpm; complete replacement of the release medium at each sampling point New Zealand rabbits (male and female). GT-PL-CL (92 g dose), SM-CL (53 g dose) or one eye drop (150 g dose) on eye. Contralateral eyes were referred as control. In vitro , all CLs showed more than 60% burst release in the first 1 h. SM-CL completed 90% release in 6 h, while GT-PL-CL extended the release for 48 h. In vivo, SM-CL and GTPL-CL prolonged the release 12 h and 24 h, respectively. Maulvi et al., 2020 1. Introduction 68 30 Hyaluron ic acid HEMA (46.7%), MAA (0.08%), EGDMA (0.5%) in water (52%) (0.1 mm) Direct addition of HA to the monomer solution 2 mL simulated tear fluid at 100 rpm at 35 ºC; complete replacement of the release medium at each sampling point New Zealand white rabbits. CL was placed on right eye. As control, one drop (50 μL) of 0.1% HA was instilled on the right eye. Left eye was referred as control. In vitro release from CLs (200 µg HA) was prolonged for 10 days, without burst. In vivo showed a burst in the first day followed by therapeutically useful values for 10 days (MRT 128 h). Eye drops disappeared in the first 3 h of treatment (MRT 0.74 h). Maulvi et al., 2015 31 Hyaluronic acid DMA (200 µL), EGDMA (10 µL), siloxane (100 µL), HEMA (up to 1 mL) HA and reduced graphene oxide (rGO) were added before polymerization in silicone CLs (HAGO-DL) or loaded by soaking in 2 mg/mL HA in simulated tear fluid (HA - GO - SM) 2 mL simulated tear fluid at 100 rpm; complete replacement of the release medium at each sampling point White New Zealand rabbits of either sex. CL was placed on right eye. As control, one drop (50 μL) of 0.1% HA was instilled on the right eye. Left eye was referred as control. In vitro HA - GO - DL (10 µg HA loaded) sustained HA release for 96 h. In vivo, HA-GO-DL provided therapeutic levels for 48 h and promoted the production of tear fluid. Huang et al., 2021 32 Hyaluronic acid DMA (200 µL), EGDMA (10 µL), siloxane (100 µL), HEMA (up to 1 mL) HA and Pluronic F127 were added before polymerization to obtain DL-HA-Pl CLs containing 20, 40 or 60 µg of HA and 20 µg of Pluronic F127. Alternatively, CLs were soaked in 1 - 2 mL simulated tear fluid at 50 rpm; complete replacement of the release medium at each sampling point White New Zealand rabbits of either sex. DL-HA-Pl CL (22.5 µg HA) and HA-SM CL (17.6 µg HA) was placed on right eye. As control, one drop (50 μL) of 0.1% HA was instilled on the right eye. Left eye In vitro , HA - SM CL showed high burst and completed the release in 12-36 h. DL-HA-Pl CL sustained HA release for 48-96 h. In vivo, HA-SM CL and DL-HA-Pl CL provided therapeutically useful values for 4 and 48 h, respectively. Wei et al., 2020a ANA FILIPA PEREIRA DA MOTA 69 3 mg/mL HA in simulated tear fluid for 7 days (HA - SM CLs) was referred as control. 33 Prednisolone and beclomethasone Lidofilcon (HEMA-based) Soaking in 1 mL of prednisolone (5 mg/mL) or beclomethasone (1 mg/mL) for 18 h at 4 ºC 1 mL of saline medium for injection (unknown temperature or stirring); complete replacement of the release medium every 3 h New Zealand white rabbits. Drug-loaded CL on both eyes and the eyes were closed with surgical tape for four hours. The treatment was applied on days 1, 2, 5, 8 and 10. The amount of drug in plasma and anterior and posterior segment tissues was analyzed on day 11. In vitro , CLs sustained drug release for 6 h. Both drugs were preferentially found in posterior segment tissues, with lower levels in vitreous humour. Schultz et al., 2011 34 Pirfenidone 11 commercially available CLs Soaking in 2 mL of 0.05%-0.5% drug solution 2 mL of PBS under shaking; complete replacement of the release medium at each sampling time New Zealand rabbits. One polymacon CL on the right eye (1147 g dose), one eye drop (30 L of 0.5%; 150 g dose) in the left eye. In vitro , polymacon CL showed sustained release for 30 min. In vivo, significant levels in tear fluid were recorded for 60 min. CL wearing provided measurable drug levels in cornea, aqueous humor and scl era. Yang et al., 2016 35 Pirfenidone Silicone hydrogel CLs (Acuvue Oasys) pretreated with vitamin E Soaking in 0.1% drug solution in PBS for 72 h 3 mL of PBS without replacement of the release medium New Zealand White rabbits of either sex. An alkali burn was induced in one eye. One group received CLs. In vitro , CLs sustained drug release up to 260 min. Gene expression of inflammatory cytokines IL-1β, TNFα, and TGF-β1 was reduced. Dixon et al., 2018 1. Introduction ANA FILIPA PEREIRA DA MOTA 76 Bimatoprost-imprinted silicone hydrogel CLs (Table 1.2, entry 4) showed minor improvements in drug loading compared to nonimprinted CLs, but more prolonged release in vitro (up to 36 h vs. 24 h) and slightly higher drug levels in tear fluid in vivo (rabbits) for 12 h [Yan et al., 2020]. Once again, although the capability of the CLs to sustain drug release in vivo was shown to be shorter than in vitro, the increase in residence time in tear fluid compared to eye drops was remarkable. Comonomers with affinity for the drug, such as N-vinylpyrrolidone (NVP), have been used to improve the capability of HEMA-based CLs to uptake puerarin and to prolong its release in vitro and in tear fluid for 4 h (Table 1.2, entry 5) [Xu et al., 2010a]. Extended in vitro release up to 10 h was observed when β-cyclodextrin was copolymerized with HEMA, which in turn allowed for prolonged permanence in tear fluid, up to 6 h (Table 1.2, entry 6) [Xu et al., 2010b]. Favorable drug-CL interactions may also occur when using commercially available CLs, particularly silicone hydrogels, which are more prone to stablish hydrophobic interactions. As an example, diquafosol, a secretagogue for dry eye treatment, can be taken up by comfilcon A and balafilcon A (up to 0.02 mg) (Table 1.2, entry 7). In in vivo studies, CLs sustained drug release in tear fluid for 240 min and increased tear secretion for 300 min, while the effects of topical instillation only lasted for 90 min [Dominguez-Godinez et al., 2018]. Loading of hydrophobic drugs can be enhanced if the drug is encapsulated in lipid nanoparticles or microemulsions. Ketotifen was encapsulated in solid lipid nanoparticles, which were pegylated (pSLNs) or not (SLNs) and then used for the drug loading of HEMAbased hydrogels (Table 1.2, entry 8) [Zhang et al., 2020]. Some hydrogels were loaded with ketotifen by addition of SLNs (DL-K-SLN- 1. Introduction 77 100) or p-SLNs (DL-K-p-SLN-100) to the monomer solution before polymerization. Another batch of hydrogels was loaded by soaking into either ketotifen solution (SM-K-50 and SM-K-100) or ketotifenencapsulated SLNs (SM-K-SLN-100) or p-SLNs (SM-K-p-SLN-100). All hydrogels were autoclaved before testing drug release. Hydrogels loaded by soaking showed a very intense burst in the in vitro release tests (Figure 1.7A1). Directly loaded CLs contained more drug and released it at slower rate. Interestingly, in vivo, these latter CLs (DL-Kp-SLN-100) provided lower Cmax (445.7 ± 85.3 μg/mL vs. 581.6 ± 152.7 μg/mL) but more prolonged drug levels in tear fluid along time than CLs that were loaded by soaking in ketotifenencapsulated p-SLNs (SMK-p-SLN-100) (Figure 1.7A2). This finding was related to stronger retention of the p-SLNs when they were incorporated into the bulk of the hydrogel. Similarly, prolonged ketotifen release in both in vitro and in vivo was recorded from CLs loaded with silica shell nanoparticles encapsulating the drug (Table 1.2, entry 9) [Maulvi et al., 2016a]. From a safety perspective, it should be noted that direct loading, in which the drugs are incorporated into a CL material before polymerization, has the inherent risk of leakage of unreacted monomers when the CL is inserted in the eye. In vitro-in vivo correlations (IVIVC) have been attempted through Levy plot analysis. Mainly, the percentage of drug released in vitro at a certain time is reported on the X-axis, and the percentage of drug released in the tear fluid at the same time on the Y-axis. The methodology for the estimation of this latter parameter was not disclosed in most papers, but it can be assumed that the percentage of drug released in vivo was estimated as [Xu et al., 2019] Drug released 𝑖𝑛 𝑣𝑖𝑣𝑜 (%) =       ×100% (Eq. 1.11) ANA FILIPA PEREIRA DA MOTA 78 which can be calculated as Drug released 𝑖𝑛 𝑣𝑖𝑣𝑜 (%)=   ×100% (Eq. 1.12) In the case of ketotifen, the Levy plot suggests moderate IVIVC (Figure 1.7A3) [Zhang et al., 2020], although this plot should be read with caution, since only four data points of drug levels in tear fluid were used for the analysis. CLs loaded with drugs encapsulated in microemulsions have also been investigated in detail. As an example, the effect of surfactant chain length (C8-sodium caprylate, C12-Tween 20, C18-Tween 80) and the molecular weight of Pluronic block copolymers (8400-PF68 and 12600-PF127) was investigated to elucidate how the stability of the microemulsion may determine cyclosporine A release kinetics from hydrogel CLs (Table 1.2, entry 10) [Maulvi et al., 2017a]. CLs loaded with cyclosporine A during polymerization (DL-100) were opaque due to drug precipitation. CLs loaded with cyclosporine A encapsulated in stable PF127-T80 microemulsions were transparent and released the drug in vitro faster than DL-100 and non-stable PF68-SC CLs (Figure 1.7B1). In this case, a prolonged release profile was not synonymous of prolonged efficacy since the release rate from drug-precipitated CLs was too slow to achieve therapeutic levels (Figure 1.7B2). Indeed, in vivo release tests showed that PF127-T80-containing CLs may supply higher drug levels and for more prolonged time than the other tested CLs. Although a dependence of the percentage of drug released to tear fluid on the percentage of drug released in vitro was observed (Figure 1.7B3), the correlation coefficients of Levy plots were far from 1, mostly because the cumulative amount released in vivo in the first time period was larger than that predicted from the in vitro release values. The authors pointed to the presence of lipophilic proteins in tears as a 1. Introduction 79 cause of the Levy plot deviations, since they bind and solubilize hydrophobic drugs, promoting the release from the CL [Maulvi et al., 2017a]. Prolonged levels of cyclosporine A in tear fluid were also found when the drug was encapsulated in Eudragit S100 nanoparticles (Table 1.2, entry 11), but IVIVC were not investigated [Maulvi et al., 2017b]. Soaking in microemulsions has also been demonstrated as a useful method to load bimatoprost and to prolong both in vitro and in vivo drug release from CLs (Table 1.2, entry 12). Compared to the loading by soaking in a bimatoprost solution, the microemulsions doubled the amount loaded and the time required for complete release both in vitro (48 h vs. 24 h) and in vivo (24 h vs. 12 h) [Xu et al., 2019]. Similar results were reported for travoprost (Table 1.2, entry 13) [Xu et al., 2020]. Hydrophobic polymeric microparticles have been tested to encapsulate olopatadine and then applied as a doughnut ring on CLs (Table 1.2, entry 14). Compared to CLs loaded by soaking (SM-OL) or to which the drug was directly added during polymerization (DLOL), the doughnut CLs (DNT-OL) avoided burst release in vitro, but still released most of the drug in the first 12 h. In vivo, DNT-OL CLs provided therapeutic levels in tear fluid for 24 h, but once again the Levy plot showed poor IVIVC as in vivo release was faster than predicted [Xue et al., 2020]. Also, bioinspired strategies for choosing the CL monomers have been shown able to enhance the affinity of CLs for olopatadine increasing the loading and providing 24 h almost constant rate release in vitro [González-Chomón et al., 2016]. Although these CLs were not tested in vivo, in cell culture they were able to efficiently inhibit the release of histamine and TNF-α from sensitized mast cells. ANA FILIPA PEREIRA DA MOTA 80 Figure 1.7. (A1) In vitro ketotifen release profiles from CLs loaded by soaking in drug solutions (SM-K-50 and SM-K-100) or drug-encapsulating non-pegylated solid lipid nanoparticles (SM-K-SLN-100) or pegylated solid lipid nanoparticles (SM-K-pSLN-100), and CLs loaded by adding the drug-encapsulating nanoparticles directly to the monomer solution (DL-K-SLN-100 and DL-K-p-SLN-100), (A2) in vivo ketotifen profiles in tear fluid during CLs wearing and after eye drop instillation (n = 6), and (A3) Levy plot depicting the percentage of ketotifen released in vivo vs. in vitro. Reprinted from Zhang et al. [Zhang et al., 2020] with permission from Elsevier. (B1) In vitro cyclosporine A release profiles from CLs prepared with the drug directly added (DL-100) or previously encapsulated in a microemulsion, (B2) in vivo release profiles (n = 6) and, in the insert, the appearance of the CLs that were transparent when loaded with the drug encapsulated in stable microemulsions, and opaque when the drug was encapsulated in unstable microemulsions and precipitated in the CL network, and (B3) Levy plots for percentage of drug released in vivo vs. in vitro. Reprinted from Maulvi et al. [Maulvi et al., 2017a] with permission from Elsevier. A variety of film-embedded CLs have also been designed. For example, pH-responsive films made of cellulose acetate and Eudragit S100 containing betaxolol hydrochloride were integrated into silicone hydrogels (Table 1.2, entry 15). The film prevented premature discharge during storage in PBS at pH 6.8 and provided sustained drug release for 10 days in simulated lachrymal fluid. Since the films 1. Introduction 81 enhanced drug loading (700 μg dose) compared to commonly soaked CLs (100 μg) remarkably higher AUC0-240h (599 vs. 26 μg⋅h/mL) and MRT (88 vs. 1.7 h) were recorded in tear fluid (Figure 1.8A1). Improvements compared to eye drops were also evident (AUC0-240h 10.5 μg⋅h/mL and MRT 0.4 h). The Levy plot revealed that after a lag time, good IVIVC was obtained (R2 0.9708) (Figure 1.8A2) [Zhu et al., 2018a]. In a related study, the same group evidenced the effects of the film components and thickness in controlling the release of diclofenac (Table 1.2, entry 16) and showed that for a similar amount of drug loaded, the film-embedded HEMA-based CL (121 μg dose) sustained the release both in vitro and in vivo for 12 h (Figure 1.8B1). In comparison, soaked CL (100 μg dose) rapidly discharged in 4 h. The improvement in the IVIVC according to the Levy plot was notably better in the case of the film-embedded CLs (Figure 1.8B2) [Zhu et al., 2018b]. The films were further modified to respond to ionic strength, preventing premature discharge when stored in water (Table 1.2, entry 17) [Zhu et al., 2018c] or to respond to both ionic strength and pH [Wei et al., 2020b]. Once again, the film-embedded CLs showed sustained release in vivo (Figure 1.8C1), which correlated quite well with the release pattern in vitro (Figure 1.8C2) [Zhu et al., 2018c]. ANA FILIPA PEREIRA DA MOTA 82 Figure 1.8. (A1) Betaxolol hydrochloride (BH) levels in tear fluid after instillation of eye drops (100 μg dose) or wearing of soaked CL (121 μg dose) and filmembedded CL (700 μg dose), and (A2) Levy plot for IVIVC (R2= 0.9708). Reprinted from Zhu et al. [Zhu et al., 2018a] with permission from Elsevier. (B1) Diclofenac sodium levels in tear fluid after instillation of eye drops (150 μg dose) or wearing of soaked CL (100 μg dose) and film-embedded CL (121 μg dose), and (B2) Levy plot for IVIVC of soaked CL (R2= 0.9019) and film-embedded CL (R2= 0.9230). Reprinted from Zhu et al. [Zhu et al., 2018b] with permission from Elsevier. (C1) Betaxolol hydrochloride (BH) levels in tear fluid after instillation of suspension eye drops (100 μg dose) or wearing of dug-resin complex film-embedded CL (700 μg dose), and (C2) Levy plot for IVIVC obtained for the CL (R2= 0.9406). Reprinted from Zhu et al. [Zhu et al., 2018c] with permission from Elsevier. 1.4.2 Comparison of In vitro Release Profiles and Therapeutic Outcome Regulation of intraocular pressure (IOP) using timolol-loaded CLs has been the aim of various studies. Since most commercially available CLs lack sufficient affinity for ocular drugs, Chauhan and coworkers developed the strategy of creating biocompatible and optically transparent diffusion barriers that rely on the hydrophobic features of 1. Introduction 83 vitamin E [Peng et al., 2012a; Peng et al., 2012b]. Both vitamin Epretreated and non-pretreated silicone hydrogel CLs were loaded with 200 μg timolol with the aim of continuous wearing for 4 days (Table 1.2, entry 18). In vitro, non-pretreated CLs released 80% of the drug in the first 4 h, while vitamin E-pretreated CLs extended drug release for up to 84 h [Peng et al., 2012a]. The CLs were tested in a Beagle dog model of glaucoma. Daily-replaced non-pretreated CLs (60 μg timolol) led to a significant decrease in IOP from the first day of treatment until one day after. A similar pattern was observed for the eye drops, but the IOP decrease was more pronounced with the daily CLs in spite of containing only 20% of the drug dose instilled with the eye drops. Nonpretreated CLs worn for 4 days (200 μg timolol) caused a decrease in IOP during the first two days only, as expected from the limited capability of the CLs to sustain drug release in vitro. In comparison, vitamin E-pretreated CLs wore for 4 days (200 μg timolol) caused a progressive decrease in IOP from day 1 to day 4, and the decrease was maintained 24 h after CL removal. These findings clearly demonstrated the advantages of sustained drug release from CLs and their capability to enhance ocular bioavailability compared to eye drops. Interestingly, commercially available senofilcon A CLs showed distinct release profiles depending on whether they were loaded with one antiglaucoma drug or simultaneously with two drugs (Table 1.2, entry 19) [Hsu et al., 2015a]. In separate, senofilcon A CLs loaded 20 μg timolol and sustained the release over 0.7 h, or loaded 122 μg of dorzolamide and sustained the release for 2.5 h. When both drugs were simultaneously loaded, the loading increased up to 60 μg timolol and 218 μg dorzolamide, and the release was extended to 1.2 h for timolol and 3.0 h for dorzolamide. The increase in drug affinity could be due to favorable drug-drug interactions through hydrogen bonding. Pretreatment of CLs with vitamin E increased drug uptake and led to more ANA FILIPA PEREIRA DA MOTA 84 controlled release (Figure 1.9A). Timolol release was sustained for 24.6 h in single loaded CLs (18 μg dose) and for 42.2 h in dually loaded CLs (193 μg). Similarly, dorzolamide release was prolonged for 36.0 h in single loaded CLs (122 μg dose) and for 42.3 h in dually loaded CLs (680 μg) [Hsu et al., 2015a]. In vivo efficacy was evaluated in terms of decrease in IOP in a Beagle dog model of glaucoma. Compared to eye drops that required frequent administration to maintain low IOP values in the treated eye and that altered the IOP values in the contralateral eye (Figure 1.9B), dually-loaded CLs showed more pronounced and sustained decrease in IOP using a lower dose (4to 6-fold lower) and avoiding effects on the control eye (Figure 1.9C and D). Dually-loaded vitamin E-pre-treated CLs maintained the therapeutic effect for two days and, after treatment, the decrease in IOP was maintained for approx. one week after the CLs were removed (Figure 1.9D). These findings confirm once again that CLs can increase ocular drug bioavailability and decrease systemic absorption of drugs through the conjunctiva and nasolacrimal duct, potentially avoiding side effects. These results also suggest a correlation between the time the drug can be released in vitro for a sustained period and the time that therapeutic effects can be maintained in vivo. It has been hypothesized that the prolonged IOP reduction may be a consequence of the creation of drug depots in the ocular tissues and, in particular, drug partition into corneal epithelial cells [Hsu et al., 2015a]. When hydrophilic drugs are administered as eye drops, the precorneal residence time is very short and the drug molecules may penetrate the epithelium by diffusion in between the cells (through the tight junctions), reach the stroma and then diffuse across the endothelium to the aqueous humor. In contrast, the transcellular pathway is slow and requires prolonged contact time for the drug to partition into the epithelium cells. Therefore, sustained release from CLs for several days may facilitate drug accumulation into 1. Introduction 85 epithelium cells, which may subsequently act as drug depots. Once the CL is removed, the accumulated drug molecules may slowly release from the cells, maintaining the therapeutic effect. Figure 1.9. (A) Timolol and dorzolamide release profiles from dually loaded senofilcon A CLs that were pre-treated (20% VE) or not (0% VE) with vitamin E. In the insert the solid lines represent the fitting to the square root kinetics (n = 3). (B, C and D) Measurement of IOPs of 10 beagle dogs treated with (B) Cosopt® eye drops (one eye twice a day for 4 days) for 79 h (indicated by the dash line); (C) dually-loaded CLs worn for 24 h and replaced daily for 5 days; and (D) duallyloaded vitamin E pre-treated CLs worn for 48 h and replaced once with a similar CL (n = 10). Reprinted from Hsu et al. [Hsu et al., 2015a] with permission from Elsevier. Dual delivery of timolol and bimatoprost has been investigated in the form of small implants (partial rings) attached to the outer periphery of silicone hydrogel CLs [Desai et al., 2020]. This approach had the drawback of rapid discharge during wet sterilization. Therefore, omplant-containing CLs could only be sterilized using radiation ANA FILIPA PEREIRA DA MOTA 92 injection, the release in vitro was sustained for 6 h. In vivo evaluation consisted of placing drug-loaded CLs on both eyes, which were kept closed with surgical tape for 4 h. Prednisolone was detected in the posterior segment ocular tissue of all rabbits, with concentrations ranging 26–166 ng/g. Prednisolone was found in the vitreous humor of three out of eight eyes and only detected in the plasma of one animal (out of six). Its metabolite prednisone (inactive) was below the quantification limit in all cases. Beclomethasone dipropionate was also detected in the posterior segment tissue, but not in vitreous humour and plasma. Its active metabolite, 17-beclomethasone, was detected in the posterior segment tissue and in most vitreous humor samples. Both prednisolone and beclomethasone dipropionate are small (< 600 g/mol) hydrophobic molecules. If they had penetrated the eye through the cornea and then migrated towards the macula and retina, the parent drugs should also be detected in the vitreous humor at similar levels. Since this was not the case, the drug molecules released from the CLs to the cornea and limbal areas may have entered through a non-corneal route. They could have reached the local vasculature and then been transported toward the posterior segment. The absence of drug in plasma discarded the notion that access to the posterior segment was via systemic circulation [Schultz et al., 2011]. Screening of eleven commercially available CL materials pointed to polymacon (polyHEMA) CLs as the most appropriate ones for loading of pirfenidone, an anti-inflammatory drug that favors the healing of the ocular surface (Table 1.2, entry 34) [Yang et al., 2016]. All tested CLs released the drug rapidly in vitro (Figure 1.11A), with polymacon having the best release profile (~30 min release duration). In rabbit eyes, the CLs (1147 μg dose) provided relevant pirfenidone levels in tear fluid for 60 min, while eye drops (150 μg dose) disappeared in the first 15 min (Figure 1.11B). Drug-loaded polymacon 1. Introduction 93 CLs led to significantly higher drug levels in the cornea, aqueous humor and sclera at all data points assessed (Figure 1.11C, D and F). Importantly, the drug levels in the conjunctiva only showed a minor increase in the first 60 min (Figure 1.11E), which can be correlated with the preferential release of the drug towards the post-lens lachrymal fluid. Therefore, low levels in the conjunctiva may be the result of rapid clearance, as well as systemic absorption. Simultaneous improvements in pirfenidone loading and controlled release (80-260 min) were observed for silicone hydrogel CLs (Acuvue Oasys®) pre-treated with vitamin E (Table 1.2, entry 35). In an animal model of alkali burn, pirfenidone-loaded CLs efficiently down-regulated the gene expression of several inflammatory cytokines IL-1β, TNF-α, and TGF-β1 in the cornea [Dixon et al., 2018]. ANA FILIPA PEREIRA DA MOTA 94 Figure 1.11. ( A ) Percentage of pirfenidone (PFD) released in PBS from commercially available CLs that were loaded by soaking in 2 mL of 0.5 mg/mL drug, (B) levels of PFD in tear fluid (rabbit model) after instillation of one drop of 0.5% PFD eye drop and during wearing of PFD-loaded polymacon CLs; (C-F) PFD levels in different ocular tissues at different times after eye drop instillation and CL wearing. Reprinted from Yang et al. [Yang et al., 2016] with permission by Taylor & Francis Ltd. 1. Introduction 95 In a subsequent study, CLs showing prolonged release of pirfenidone were designed by embedding a drug insert into two layers of a silicone elastomer (Table 1.2, entry 36) [Wu et al., 2021]. The CLs exhibited a very low water content (11%) and 10-times lower dose (15 μg) than the amount instilled using eye drops. In vitro, CLs released 52% of the loaded dose in the first hour, followed by ten-times lower release rate in the next 15 hours. In vivo, CLs sustained drug levels for 8 h in tear fluid, which led to a higher drug concentration in the aqueous humor after 2 h of wear compared to the eye drop instillation, despite releasing less drug [Wu et al., 2021]. Pegylated (p-SLNs) and non-pegylated (SLNs) solid lipid nanoparticles have also been tested to enhance the loading of epalrestat (an aldose reductase inhibitor used for the treatment of diabetic neuropathy) for delivery to the retina (Table 1.2, entry 37). In vitro release profiles in simulated tear fluid revealed that CLs directly loaded with epalrestat encapsulated p-SLNs (DL-EP-p-SLN) had a smaller burst and sustained the release up to 196 h, compared to CLs loaded by soaking (SM-EP-p-SLN), which prolonged the release up to 144 h. DLEP-p-SLN CLs and SM-EP-p-SLN CLs provided measurable drug levels in tear fluid (rabbit model) for 96 h and 48 h, respectively [Zhu and Sheng, 2020]. Epalrestat accumulation in various ocular tissues after wearing of DL-EP-p-SLN CLs for 24 h were in rank order of lens (7.28 μg/g) > cornea (6.34 μg/g) > aqueous humor (4.84 μg/g) > retina (0.21 μg/g). In contrast, after 24 h of an eye drop instillation, the only measurable levels of epalrestat were detected in the lens (0.87 μg/g) and aqueous humor (0.82 μg/g). These findings support that CLs may facilitate drug penetration to the back of the eye. Efficient drug delivery to the posterior segment was investigated for the antibiotic ofloxacin loaded in HEMA-based corneal CL, scleral/ ANA FILIPA PEREIRA DA MOTA 96 corneal CL (S/CL) and rings (Table 1.2, entry 38) [Shikamura et al., 2016]. The composition was the same in all cases; the only change referred to the size and shape of the device, which led to different weights: 44 mg CL, 261 mg S/CL, and 183 and 72 mg for ring 1 and 2, respectively. After soaking in the antibiotic solution, the amounts loaded were 282, 1715, 1200 and 565 μg per device, respectively. Drug release was evaluated in vivo (rabbit) in terms of the difference between dose and amount of drug remaining in the device after certain wearing periods (Figure 1.12A1). As expected from their higher loading, S/CL and ring 1 released more drug, although the percentage released after 1 h of wear was slightly lower (47.5% and 40.9%, respectively) than that recorded for the CL (59.6%). Preferential biodistribution of ofloxacin to anterior segment tissues was observed for CLs, which provided the highest drug accumulation in the cornea (Figure 1.12A2). S/CL led to high drug levels both in anterior and posterior segment tissues, while the rings facilitated drug accumulation in the posterior segment (Figure 1.12A3). All devices were much more efficient than eye drops in drug biodistribution and provided therapeutically useful drug levels in all tissues, including the retina-choroid after 1 h of wear (0.4 μg/g with CL, and ~4 μg/g with S/CL and rings) [Shikamura et al., 2016]. There are only a few studies that report on in vitro and in vivo release kinetics together with therapeutic outcomes and drug accumulation in ocular tissues [Maulvi et al., 2019]. HEMA-based CLs were used to test the potential value of using gold nanoparticles (65 nm) to increase the loading and slow the release of timolol (Table 1.2, entry 39). Two different approaches were investigated: (i) addition of gold nanoparticles to the monomers before polymerization and; (ii) addition of gold nanoparticles to the timolol solution in which the hydrogels were soaked. These approaches slightly increased the capability of the hydrogels to uptake timolol (284 and 277 μg, respectively) compared 1. Introduction 97 to the same hydrogels processed in the absence of gold nanoparticles (253 μg). Regardless of the procedure, all hydrogels released most of the drugs within one hour when tested in vitro. Unexpectedly, in vivo (rabbit) results showed that timolol levels in tear fluid were quantifiable for 60 h. A significant decrease in IOP values were recorded in the first 24 h of wear. Thus, once again the therapeutic effect appears to be more prolonged than the release profiles recorded in vitro under sink conditions. This finding can be related to drug accumulation in various ocular tissues, as observed in Figure 1.12B1 and B2 [Maulvi et al., 2019]. Specifically, accumulation of timolol in the ciliary muscle, where most β-receptors are located [Wax and Molinoff, 1987], may explain prolonged IOP decrease. Unfortunately, a similar analysis for the drug instilled using eye drops was not available. ANA FILIPA PEREIRA DA MOTA 98 Figure 1.12. ( A1 ) Ofloxacin released from HEMA - based corneal CL, scleral/corneal CL (S/CL) and rings once placed in Japanese albino rabbits (nictitating membranes were removed); the S/CL fell out of the eyes after 2 h; and (A2 and A3) ofloxacin levels in ocular tissues after 1 h wearing compared to the levels achieved after one drop instillation of ofloxacin ophthalmic solution (OOS) (n = 3; *p < 0.05, **p < 0.01, ***p < 0.005). Reprinted from Shikamura et al. [Shikamura et al., 2016] with permission from Taylor & Francis Ltd.; (B1 and B2) Evolution of timolol levels in various ocular tissues after wearing of CL with (0.025 mM-GNP-CL-4 mg; 277 μg of timolol) or without (Blank-4 mg; 253 μg of timolol) gold nanoparticles (mean ± SD; n = 3; # p < 0.05). Reprinted from Maulvi et al. [Maulvi et al., 2019] with permission from Elsevier. 1. Introduction 99 Latanoprost-eluting CLs have been designed by encapsulating a drug-loaded poly(lactic-co-glycolic)acid (PLGA) film in methafilcon (Table 1.2, entry 40) [Ciolino et al., 2014]. Two high molecular weight (118 kDa) PLGA of 65 glycolide:35 L-lactide ratio and 85 glycolide:15 L-lactide ratio were used, and three sets of CLs were prepared containing films of thickness 20, 40 and 45 μm. CL65:35, 20, CL65:35, 40 and CL85:15, 45 contained 89, 178 and 178 μg latanoprost, respectively, and released in vitro 90%, 48% and 45% of the drug dose in the first three days. CL65:35, 400 and CL85:15, 45 sustained drug release for 20 days more (Figure 1.13A1). In vivo (rabbits) concentration of latanoprost was continuously monitored in the aqueous humor for 4 weeks (Figure 1.13A2). Compared to the low Cmax (54 ng/mL) and the rapid concentration decrease recorded after one drop instillation, CL65:35, 20, CL65:35, 40 and CL85:15, 45 led to Cmax of 970, 854 and 1473 ng/mL, respectively, and average steady concentration (Css) of 5.6, 39.6 and 21.0 ng/mL. Interestingly, good IVIVC was observed for the percentage of total drug absorbed in vivo (percentage of AUC0-28 days in aqueous humor) with respect to the percentage of drug released in vitro. Moreover, the correlation coefficient became closer to 1 (R2 = 0.98) when CL85:15, 45 was presoaked for 1 or 3 days in PBS to remove the drug released as a burst before wearing (Figure 1.13A3) [Ciolino et al., 2014]. This finding opens the possibility of predicting drug levels in the aqueous humor from the in vitro release values. Since latanoprost does not decrease IOP in rabbits, the therapeutic efficacy of similarly designed CLs (PLGA 50:50 and 147 μg drug) was demonstrated in vivo in glaucomatous eyes of cynomolgus monkeys [Ciolino et al., 2016] (Figure 1.13A4). Recently, methafilcon CLs encapsulating PLGA 85:15 films loaded with dexamethasone were shown to sustain drug release in vitro over 7 days and to provide therapeutic levels in various ocular tissues, ANA FILIPA PEREIRA DA MOTA 100 including the retina (Table 1.2, entry 41); (Figure 1.13B1 and B2) [Ross et al., 2019]. In the aqueous humor (rabbit model) the drug levels were sustained for 7 days, and the CLs effectively prevented sutureinduced corneal neovascularization and inflammation (Figure 1.13B3) and also lipopolysaccharide-induced anterior uveitis [Bengani et al., 2020]. Some other in vivo studies have been carried out with one optimized formulation and the drug concentration in tissues and the therapeutic response evaluated. Since only one CL type was reported, the effect of a change in the release rate on the in vivo levels and therapeutic outcomes cannot be evaluated. Nevertheless, all the studies evidenced that compared to eye drops, CLs favor drug accumulation and prolong the therapeutic levels, which in turn leads to improved therapeutic response with either less dose or less frequent administration. These findings have also been reported for CLs loaded with cyclosporine A impregnated in nanoporous silica by means of supercritical CO2 (Table 1.2, entry 42) [Choi et al., 2019]. In vitro, CLs sustained drug release for 8 h, while in vivo, high levels of drug were detected for 48 h in cornea and conjunctiva. CLs increased tear volume and stabilized tear film after 1 and 2 weeks of treatment. 1. Introduction 101 Figure 1.13. (A1) Latanoprost release profiles in vitro from CL embedding drugloaded films of PLGA 65:35 of thickness 20 μm (CL65:35, 20) and 40 μm (CL65:35, 40) and PLGA 85:15 of thickness 45 μm (CL85:15, 45); (A2) latanoprost concentration in aqueous humor (rabbits) during CL wearing; and (A3) Levy plots for CL85:15, 45 without pre-conditioning (R2 = 0.875) and with pre-conditioning for 108 Atropine and roscovitine DMA, TRIS, TRIS-OH, EGDMA (Thickness - 1 mm) Incorporation the drugs during the synthesis of the hydrogel materials or soaking in 1 mL of 2 mg/mL of roscovitine solution for 7 days PBS pH 7.4 (1 mL, replaced completely at predetermined timepoints) 37 ºC (Rotation not specified) Atropine – about 23 days. Roscovit ine – about 50 days Lasowski and Sheardow n, 2016 Atropine, pirenzepine Commercially available silicone (Narafilcon A), and conventional hydrogel CLs (Etafilcon A, Oculfilcon B, Omafilcon A) (Thickness – mean of 0.08 mm) Soaking in 2 mL of 10 mg/mL of atropine, and 1 mg/mL of pirenzepine for 24 h PBS (4 mL, 100 µl removed at predetermined timepoints) 34 ºC, shaking 24 hours Hui et al., 2017 Azulene HEMA, MAPTAC, anionic monomer, and EDMA (Thickness – not specified) Soaking in 5 mL of 0.2 wt.% azulene for 48 h Physiological saline solution pH 7.0 (5mL, replaced completely every 2 hours) 37 ºC (Rotation not specified) 8 hours Uchida et al., 2003 Betaine, dexpanthenol Commercially available silicone hydrogel CLs (Narafilcon A, Senofilcon A) (Thickness – not specified) Soaking in 3 mL of 8 or 80 mg/mL of dexpanthenol or betaine, respectively for 1 or 2 days PBS (2 mL) (Replacement not specified) Not specified Until 40 hours Hsu et al., 2015b Betamethasone Commercially available silicone hydrogel CLs (Lotrafilcon B, Senofilcon A, Comfilcon A) (Thickness – 0.07 and 0.08 mm) Soaking in 2 mL of 2 mM of betamethasone until achieve the equilibrium at 25 ºC Artificial tear solution pH 8 (2 mL, replaced completely at predetermined timepoints) 37 ºC (Rotation not specified) Until 41 days Rad et al., 2016a ANA FILIPA PEREIRA DA MOTA 109 Bimatoprost and latanoprost Commercially available silicone hydrogel CL (Senofilcon A, Narafilcon A) (Thickness – 0.1 mm) Soaking in 3 mL of 0.125 mg/mL of bimatoprost, and 62.4 µg/mL of latanoprost until achieve the equilibrium PBS (Volume and replacement not specified) Not specified About 10 days Sekar et al., 2019 Brimonidine HEMA, MAA, MAAM, 4VP, and EGDMA (Thickness – 0.4 mm) Soaking in 10 mL of 0.2 mM of brimonidine aqueous solution for 24 h at room temperature 0.9% NaCl and artificial tear solutions (10 mL, 1 mL replaced at predetermined timepoints) 37 ºC (Rotation not specified) 48 h Omranipo ur et al., 2015 Bupivacaine, lidocaine, tetracaine Commercially available silicone hydrogel CL (Lotrafilcon B). TRIS, DMA, DBE-U12, NVP, and EGDMA (Thickness – 0.2 mm). Soaking in 3 mL of 5.0, 2.5, and 1.0 mg/mL of lidocaine, bupivacaine, and tetracaine solutions, respectively for 7 days PBS (2 mL, without replacement) Not specified Until about 90 hours Peng et al., 2012c Chlorhexidine, levofloxacin, and diclofenac HEMA, PVP, TRIS, NVP, EGDMA (Thickness – 0.25 and 0.3 mm) Soaking in 1 mL of 1 mg/mL of chlorhexidine, levofloxacin, and diclofenac until achieve the equilibrium at 25 ºC PBS or water (3 mL, 200 µl replaced at predetermined timepoints) 38 ºC, 180 rpm 4 days Pimenta et al., 2016b Ciprofloxacin Ciprofloxacin - PLGA films coated with pHEMA (Thickness – 450 μm) Ciprofloxacin films were created by solvent casting, using a 1:1 ratio of medication to PLGA PBS pH 7.4 (15 mL, replaced completely at each sampling point) 37 ºC, continuous shaking 4 weeks Ciolino et al., 2009 Ciprofloxacin HEMA, TRIS, EGDMA, acetic acid or acrylic acid (Thickness – not specified) Soaking in 2 mL of ciprofloxacin (9.06 mM, 0.10 mM and 0.025 mM) for one week Artificial tear solution (2 mL, without replacement) 34 ºC, with continuous shaking 14 days Hui et al., 2012 Ciprofloxacin Commercially available conventional hydrogels CLs (Ocufilcon D), and silicone hydrogel CLs (Delefilcon A, Soaking in 1 mL of drug solutions for 18 h PBS (1 mL, completely replaced every hour) 25 ºC (Rotation not specified) Until 16 hours Qin et al., 2017 1. Introduction 110 Comfilcon A, Lotrafilcon B, Narafilcon A) (Thickness – not specified) Ciprofloxacin Commercially available silicone hydrogel CLs (Lotrafilcon B, Comfilcon A, Senofilcon A) (Thickness – 0.07/ 0.08 mm) Soaking in 2 mL of ciprofloxacin solution for 24 h Artificial tear solution (2 mL, completely replaced at predetermined timepoints) 37 ºC (Rotation not specified) 960 hours Rad et al., 2017 Ciprofloxacin HEMA, MEHQ, and PEGDM (thickness not specified) Ciprofloxacin was encapsulated into nanospheres and mixed with the monomer’s solution Buffer (2 mL, completed replaced every 24 h for 3 days, then at 2 to 4 days for 14 days) 37 ºC (Rotation not specified) 14 days Garhwal et al., 2012 Ciprofloxacin Commercially available silicone (Balafilcon A, Lotrafilcon B, Lotrafilcon A, Senfilcon A, Galyfilcon A, Comfilcon A), and conventional hydrogel CLs (Polymacon, Alphafilcon A, Etafilcon A) (Thickness – mean of 0.08 mm) Soaking in 2 mL of 0.3% of ciprofloxacin solution for 24 h at 34 ºC Unisol4 pH 7.4 (2 mL, 5 or 10 µl was taken at predetermined timepoints) 34 ºC, gently shaking 24 hours Hui et al., 2008 Ciprofloxacin, betamethasone Commercially available silicone hydrogel CLs (Lotrafilcon B, Comfilcon A, Senfilcon Soaking in 2 mL of 2 mM of ciprofloxacinbetamethasone solution at 25 ºC until achieve the equilibrium STF pH 8 (2 mL, completely replaced at predetermined timepoints) 37 ºC (Rotation not specified) About 38 days Rad et al., 2016b ANA FILIPA PEREIRA DA MOTA 111 A) (Thickness – 0.08 mm) Ciprofloxacin, dexamethasone HEMA, DMAA, TRIS, hyaluronic acid, and EGDMA (Thickness – 0.93 - 1.84 mm) Soaking in 3 mg/mL of ciprofloxacin, or 1 mg/mL of dexamethasone for 24 h at 25 ºC PBS pH 7.4 (Volume and replacement not specified) 34 ºC, 100 rpm Until 7 days Nguyen et al., 2012 Cisplatin HEMA, APS, MAAc, NN - MBA, and TEMED (Thickness – not specified) Cisplatin was mixed with the monomer’s solution Not specified 37 ºC (Rotation not specified) 96 hours Singh et al., 2011 Cromolyn sodium, ketotifen fumarate, ketorolac tromethamine,dex amethasone sodium phosphate Commercially available silicone (Balafilcon A, Lotrafilcon A), and conventional hydrogel CLs (Polymacon, Omafilcon A, Etafilcon A, Vifilcon A, Alphafilcon A) (Thickness – not specified) Soaking in 2 mL of 20 mg/mL of cromolyn, 0.2 mg/mL of ketotifen, 0.3 mg/mL of ketorolac, and 0.845 mg/mL of dexamethasone for 50 h Saline solution (2 mL, without replacement) 25 ºC (Rotation not specified) 24 hours Karlgard et al., 2003 Cyclosporine A HEMA, and EGDMA (Thickness – not specified) Cyclosporine cholesterol - hyaluronate micelles were mixed with the monomer’s mixture before polymerization PBS (1 mL, replaced completely at predetermined timepoints) 37 ºC (Rotation not specified) 288 hours Mun et al., 2019 Cyclosporine A Commercially available silicone hydrogel CLs (Lotrafilcon A, Lotrafilcon B, Senofilcon A, Balafilcon A), a conventional hydrogel CL (Etafilcon A) Soaking in 10 mL of 15/17 µg/mL of cyclosporine A solution PBS (1.75 mL, completely replaced at predetermined timepoints) Not specified Until 170 hours Peng et al., 2011 1. Introduction 112 (Thickness – not specified) Cyclosporine A HEMA, and EGDMA (Thickness – 0.2 mm) Cyclosporine A was mixed with the monomer’s solution PBS (3.5 mL, completely replaced every 24 h) 25 ºC (Rotation not specified) Until 30 days Kapoor et al., 2017 Cyclosporine A HEMA, and EGDMA (Thickness – 0.1 or 0.2 mm) Cyclosporine A microemulsions were mixed with monomer solution PBS (3.5 mL, completely replaced every 24 h or without replacement) 25 ºC (Rotation not specified) Until 25 days Kapoor et al., 2008a Cyclosporine A, Brij 98 HEMA, and EGDMA (Thickness – not specified) Surfactant solution was added to the monomer’s mixture before polymerization PBS (3.5 mL, without, and completely replacement every 24 h) Not specified About 42 days Kapoor et al., 2008b Cyclosporine A, dexamethasone, dexamethasone acetate HEMA, and EGDMA (Thickness – 0.1/ 0.2/ 0.4/ 0.8 mm) Drug solutions were mixed with the monomer solution PBS (3.5 mL, without replacement, or completely replaced every 24 h) Not specified Until 55 days Kapoor et al., 2009 Cysteamine Commercially available silicone hydrogel CLs (Senofilcon A, Narafilcon A, Lotrafilcon B, Balafilcon A), and conventional hydrogel CLs (Etafilcon A) (Thickness – mean of 0.04 mm) Soaking in 3 mL of 50 mg/mL of cysteamine for 24 h in the refrigerator PBS (2 mL, without replacement) 24 ºC (Rotation not specified) Until 6 hours Hsu et al., 2013 Dexamethasone HEMA, and EGDMA (Thickness – 0.1 or 0.2 mm) Soaking in 3 mL of dexamethasone solution for 3 h Deionized water or PBS (3 mL, completely replacement at the end of the release) 25 ºC (Rotation not specified) 35 hours Kim et al., 2008a Dexamethasone γCD, HPMC,and EGDE (Thickness – Soaking in 10 mL of 1 mg/mL of aqueous suspension of Water (5 mL, without replacement) 25 ºC (Rotation 72 hours MoyaOrtega et al., 2010 ANA FILIPA PEREIRA DA MOTA 113 cylindrical tube wit h 5 mm) dexamethasone for 1 week at 25 ºC not specified) Dexamethasone HEMA, DMA, TRIS(OH), ACR, mPDMS, and EGDMA (Thickness – 0.5 mm) Soaking in 1.5 mL of 1 mg/mL of dexamethasone for 48 h PBS pH 7.4 (1.5 mL, completely replaced at predetermined timepoints) 37 ºC (Rotation not specified) About 16 days Guidi et al., 2014a Dexamethasone Commercially available silicone hydrogel CLs (Galyfilcon A, Senofilcon A, Lotrafilcon A, Lotrafilcon B, Balafilcon A) (Thickness – 0.08-0.1 mm) Soaking in 2 mL of dexamethasone in PBS for 1 or 7 days and drug-ethanol for 3 h PBS (2 mL, without replacement) Not specified Until 7 days Kim et al., 2010 Dexamethasone 21-disodium phosphate HEMA, and EGDMA (Thickness – 0.050.015 mm) Soaking in 0.1 mg/mL of dexamethasone 21-disodium phosphate solution (140 mL solution per gram of CL) until equilibrium PBS (6 mL, without replacement) Not specified About 41 days Bengani and Chauhan, 2013 Dexamethasone acetate HEMA, MA, and EGDMA (Thickness – 0.2 mm) Soaking in 1.35 mg/mL of dexamethasone acetateethanol solution for 15 h PBS pH 7.4 (4.5 mL, aliquots removed and replaced with equivalent volume at predetermined timepoints) 37 ºC, constant stirring 30 days Lu et al., 2013 Dexamethasone sodium phosphate HEMA, and EGDMA (Thickness – 0.05 mm) Dexamethasone nanoparticles were mixed with the monomer’s solution PBS pH 7.4 (5 mL, 100 µl replaced at predetermined timepoints) 37 ºC, 50 rpm About 20 days Behl et al., 2016 Dexamethasone, cyclosporine, timolol, levofloxacin Commercially available silicone hydrogel CL (Delefilcon A) (Thickness – not specified) Soaking in 3, 4 or 10 mL of 1.5, 5.0, 0.05, 0.015 mg/mL of timolol, levofloxacin, dexamethasone, cyclosporine, respectively for 24, 48 or 72 h at 4 or 35 ºC PBS pH 7.4 (2 mL, without replacement) 25 ºC (Rotation not specified) Until 30 hours Dixon and Chauhan, 2017 1. Introduction 114 Dexamethasone, ketotifen fumarate, timolol maleate, and diclofenac sodium HEMA, AA, AM, NVP, DEAEM, MAA, PDMS, TRIS, DMA, EGDMA and ethanol (Thickness – 0.145 mm) Soaking in 5 mL of 0.30 mg/mL ketotifen, 0.20 mg/mL diclofenac, 0.08 mg/mL dexamethasone, and 0.20 mg/mL timolol maleate Deionized water pH 6.4 (2, 5, 30, 200, 300, 400 mL, replaced completely every 24 hours for small volumes) 20 ºC or 34 ºC, 30 or 160 rpm 96 hours Tieppo et al., 2014 Diclofenac HEMA, NVP, TRIS, and EGDMA (Thickness – 0.25 mm) Soaking in 3 mL of 1 mg/mL of diclofenac solution for 38 h at 4 ºC Saline solution 130 mM NaCl pH 6.9 (3 mL, 200 µl replaced at redetermined timepoints) 36 ºC (Rotation not specified) 24 hours Silva et al., 2020 Diclofenac sodium HEMA, and EGDMA (Thickness – 0.12 mm) Soaking in 3 mL of 0.2 mg/mL of diclofenac sodium for 7 days PBS pH 7.4 (3 mL, 1 mL replaced at predetermined timepoints) 25 ºC (Rotation not specified) About 100 hours TorresLuna et al., 2019a Diclofenac sodium HEMA, glycerol, deionized water, and EGDMA (Thickness – 01 mm) Soaking in 1 mL of 2 mg/mL of diclofenac sodium for 2 days at 37 ºC PSB (10 mL, 2 mL replaced at predetermined timepoints) 37 ºC (Rotation not specified) About 70 hours Li et al., 2020 Diclofenac sodium HEMA, GMA, and EGDMA (Thickness – 0.9 mm) Soaking in 10 mL of 80 mg/l of diclofenac sodium solution for 4 days at 25 ºC STF (10-20 mL, without replacement) 25 ºC (Rotation not specified) 20 days Dos Santos et al., 2009 Diclofenac sodium, betaxolol HEMA, and EGDMA (Thickness – 0.1 mm) Soaking in 1 mL of drug solutions for 3 days at 25 ºC STF (10 mL, 2 mL replaced at predetermined timepoints) 35 ºC, 100 rpm About 10 hours Zhu and Mao, 2019 Diclofenac sodium, flurbiprofen sodium, naproxen sodium HEMA, and EGDMA (Thickness – 0.12 mm) Drug–surfactant solutions were mixed with monomer’s solution PBS pH 7.4 (3 mL, 1 mL replaced at predetermined timepoints) 25 ºC (Rotation not specified) About 10 days TorresLuna et al., 2019b Diclofenac, and cefepime HEMA, DMAEMA, DMAA, and EGDMA (Thickness – not specified) Soaking in 5.0% of cefepime, and 2.5% of diclofenac solutions Water (Volume and replacement not specified) Not specified Not specifie d Primachen ko et al., 2014 ANA FILIPA PEREIRA DA MOTA 115 Diclofenac, ketorolac, chlorohexidine, and moxifloxacin 1.8 M of HEMA, 0.8 M of TRIS, 3.9 M of NVP, and 30 mM of EGDMA (Thickness – 0.3 mm) Soaking in 3 mL of 1 mg/mL of diclofenac and ketorolac for 38 h, and 2.5 mg/mL of chlorohexidine, 5 mg/mL of moxiflo xacin for 72 h at 4 ºC Saline solution 130 mM NaCl (3 mL, 200 µl replaced at predetermined timepoints) 36 ºC, 150 rpm 10 days Silva et al., 2016 Dorzolamide HEMA, MAA, and EGDMA (Thickness – 0.4 mm) Soaking in 50 mL of 0.005 - 0.1 mM of dorzolamide solutions for 48 h at 25 ºC 0.9% NaCl solution or STF pH 8 (10 mL, without replacement) 37 ºC (Rotation not specified) 48 hours Malaekeh - Nikouei et al., 2013 Econazole Econazole - PLGA films coated with pHEMA (Thickness – 450 μm) Econazole films were created by solvent casting, adding 200 mg to 100 mg of PLGA dissolved in 4 mL of ethyl acetate PBS (1 mL, replaced completely every 24 hours) 37 ºC, 100 rpm 10 days Ciolino et al., 2011 Epalrestat HEMA, MCS - MC12, APMA, and EGDMA (Thickness – 0.5 mm) Soaki ng in 50 mL of 6.14 µg/mL of epalrestat for 6 days at 25 ºC 0.9% NaCl solution (45 mL, without replacement) 25 ºC, 200 rpm 7 days Alvarez - Rivera et al., 2018 Epidermal growth factor Commercially available silicone hydrogel CLs (Vasurfilcon A and Lotrafilcon A) (Thickness – not specified) Soaking in 25 mL of 0.4 ppm of epidermal growth factor for 7 h PBS (25 mL, 1 mL was taken at predetermined timepoints but replacement is not specified) Not specified 7 hours Schultz and Morck, 2010 Ethoxzolamide HEMA, APMA, and EGDMA (Thickness – 0.9 mm) Soaking in 2 mL of ethoxzolamide solution for 24 h STF pH 8.0 (2 mL, and progressively increased up to 10 mL in the first 48 h) 25 ºC (Rotation not specified) About 7 days García - Fernández et al., 2013 Fe 3 O 4 - PEG - Dy 2 O 3 nanocomposites and Dy(OH)3 nanorods Commercially available silicone hydrogel CL (Etafilcon A) (Thickness – not specified) Soaking in the nanaoparticules solutions for 24 h at 37 ºC PBS pH 7.4 (Volume and replacement not specified) Not specified 24 hours Kusrini et al., 2021 Fluorescent glucose (2-(N-(7nitrobenz - 2 - oxa - PEG, DMA, and EGDMA (Thickness – 0.79 mm) Soaking in 1x10 - 3 , 5x10 - 4 , 1x10-4, and 1x10-5 mg/mL of 2 - NBDG Deionized water (50 mL, completely replaced at predetermined timepoints) 25 ºC, 25 rpm About 20 days Byrne et al., 2008 1. Introduction 116 1,3-diazol-4yl)amino)-2deoxy-glucose) or 2-NBDG Flurbiprofen Commercially available conventional hydrogel CLs (Hilafilcon B) (Thickness – not specified) Soaking in 20 mL of 3.3x10-5 M of flurbiprofen for 14 h, or supercritical solvent impregnation at 40 ºC, 12 MPa for 2.5 h Purified water (80 mL, without replacement) 37 ºC, 100 rpm 8 hours Yañez et al., 2011a Flurbiprofen and timolol maleate Commercially available conventional hydrogels CL (Methafilcon A) (Thickness – not specified) CLs were impregnated using supercritical solvent impregnation methodology at 9 MPa, 40 ºC, 120 min, 0.1/0.2 MPa/min of depressurization rate Physiological serum (6 or 10 mL, 300 or 400 µl replaced at predetermined timepoints) 37 ºC (Rotation not specified) 8 hours Costa et al., 2010b G2-C Peptide Commercially available conventional hydrogel CL (Ocufilcon D) (Thickness – not specified) Soaking in 100 µL of 1o mg/mL of G2-C peptide for 5 days at 37 ºC PBS (1 mL, completely replaced at predetermined timepoints) 37 ºC (Rotation not specified) 8 days Jaishankar et al., 2016 Gatifloxacin and moxifloxacin HEMA, MPTS, MAA, EGDMA, and 1,9-NDA. Commercially available silicone and conventional hydrogel CLs (Etafilcon A, Polymacon) (Thickness – not specified) Soaking in 3 mL of drug solutions and sterilized by moist heat (121 ºC, 30 min) PBS (2 mL, replaced completely at predetermined timepoints) 37 ºC (Rotation not specified) 72 hours Kakisu et al., 2013 Hyaluronic Acid Commercially available silicone hydrogel CL (Nelfilcon A) and hydrogels prepared with MA, Hyaluronic acid was mixed with the monomer’s solution Artificial tear solution pH 8 (20 mL, replaced completely at predetermined timepoints) 35 ºC, 30 rpm 120 hours Ali and Byrne, 2009 ANA FILIPA PEREIRA DA MOTA 117 NVP, and DEAEM (Thickness – 0.127 mm) Hyaluronic acid HEMA, DMAA TRIS, and EGDMA (Thickness – not specified) Hyaluronic acid was mixed with the monomer solution PBS pH 7.4 (1 mL, samples were taken and replenished at predetermined timepoints) 37 ºC, circular shaking Until 41 days Weeks et al., 2013 Hydrocortisone, acetazolamide HEMA, and EGDMA (Thickness – 0.9 mm) Soaking in 10 or 5 mL of 100 mg/l of hydrocortisone, and acetazolamide, respectively followed by autoclaving STF (10 - 20 mL, without replacement) 25 ºC (Rotation not specified) 1 month Dos Santos et al., 2008 Hydroxypropyl methylcellulose HEMA, TRIM, NVP, MAA, MMA, and EGDMA (Thickness – not specified) Soaking in 1% of hydroxypropyl methylcellulose PBS pH 5.80 - 8.03 (200 µl) (Replacement not specified) 25 ºC (Rotation not specified) 12 hours Kim et al., 2019 Ibuprofen sodium, diclofenac sodium HEMA, APMA, VP, and EGDMA (Thickness – 0.9 mm) Soaking in 10 - 30 mL of different concentrations of ibuprofen and diclofenac solutions for 1 week at 25 ºC PBS pH 5.8 or 8.0, or NaCl solutions (10, 50, 100, or 150mM) (6 or 30 mL, without replacement) 37 ºC (Rotation not specified) Until 26 hours Andrade - Vivero et al., 2007 Ketorolac tromethamine, flurbiprofen sodium, diclofenac sodium Commercially available silicone hydrogel CLs (Senofilcon A, Narafilcon A) (Thickness – not specified) Soakin g in 3 mL of 1.2 mg/mL of ketorolac, 0.2 mg/mL of diclofenac, and flurbiprofen for 24 h at 25 ºC Dulbecco PBS pH 7.4 (3 mL, 1 mL replaced at predetermined timepoints) 25 ºC (Rotation not specified) Until 200 hours Torres - Luna et al., 2019c Ketotifen fumarate AA, AM, HEMA, and NVP (Thickness – 0.40.7 mm) Soaking in concentrated solutions of ketotifen fumarate until achieve the equilibrium Artificial lachrymal fluid, deionized water, and lysozyme in artificial lachrymal fluid (30 mL, without replacement) (Temperatur e not specified) 120 rpm 7 days Venkatesh et al., 2007 Ketotifen fumarate Commercially available conventional hydrogel CLs (etafilcon A, Soaking in 6 mL of 0.25 mg/mL of ketotifen fumarate for 24 h at 34 ºC Commercially available saline solution (Unisol 4) (6 mL, 30 µl were taken at 34 ºC (Rotation not specified) 24 hours Soluri et al., 2012 1. Introduction 124 (Thickness – not specified) Prednisolone HEMA, MAA, and EGDMA (Thickness – not specified) Soaking in 0.4 g of prednisolone nanoparticles for 24 h PBS (10 mL, 0.5 mL replaced at predetermined timepoints) 35.5 ºC, 80 rpm 24 hours ElShaer et al., 2016 Prednisolone Acetate HEMA, MAA, and EGDMA (Thickness – 0.4 mm) Soaking in 10 mL of 0.01, 0.025, and 0.05 mM of prednisolone acetate at 25 ºC for 48 h 0.9% NaCl or artificial lachrymal solution pH 8 (10 mL, without replacement) 37 ºC (Rotation not specified) 48 hours MalaekehNikouei et al., 2012 Rhodamine B Commercially available silicone (Balafilcon A, Lotrafilcon B, Lotrafilcon A, Senfilcon A, Galyfilcon A), and conventional hydrogel CLs (Filcon 1 a, Vifilcon A, Etafilcon A, Filcon 3 a, Nelfilcon A) (Thickness – not specified) Soaking in 2 mL of 10 or 100 µg/mL of Rhodamine B until achieve the equilibrium PBS pH 7.4 (5 mL, completed replaced every hour) 37 ºC, 200 rpm 24 hours Mahomed and Tighe, 2014 Salicylic acid Commercially available conventional hydrogel CL (Hilafilcon B) (Thickness – not specified) CLs were impregnated using supercritical CO2 impregnation methodology at 11 MPa, 40 ºC, 0.1 MPa/min of depressurization rate PBS (10 mL, 0.3 mL replaced at predetermined timepoints) 37 ºC (Rotation not specified) 8 hours Yokozaki et al., 2015 Tetracaine, diclofenac sodium, flurbiprofen sodium, bupivacaine, Commercially available silicone hydrogel CLs (Senofilcon A, Narafilcon A), and conventional hydrogel CLs (Etafilcon A) Soaking in 5 mL of 0.2 mg/mL of tetracaine, diclofenac sodium, flurbiprofen sodium, and 1.0 mg/mL of bupivacaine, and 0.3 mg/mL of ketotifen fumarate for 24 h at 25 ºC Dulbecco PBS pH 7.4 (3 mL, 1 mL replaced at predetermined timepoints) 25 ºC, without rotation Until 62 days TorresLuna et al., 2020 ANA FILIPA PEREIRA DA MOTA 125 ketotifen fumarate (Thickness – 0.07 - 0.09 mm) Thiosemicarbazon e HEMA, 2,3 - di - O - methacrylated-6-Omethacrylated-βcyclodextrin monomer, and EGDMA (Thickness – 0.4 mm) Soaking in 10 mL of thiosemicarbazone suspensions for 1 day at 25 ºC STF (10 mL, 2 - 3 mL replaced at predetermined timepoints) 25 ºC (Rotation not specified) About 350 hours Glisoni et al., 2013 Timolol HEMA, MAA, and EGDMA (Thickness – not specified) Soaking in 5 mL of 0.1 mM of timolol solution at 25 ºC until achieve the equilibrium STF pH 8.0 (5 mL, without replacement) Not specified 12 hours Deng et al., 2018 Timolol maleate DMAA, TRIS, MAA, and EGDMA (Thickness – 0.3 mm) Soaking in 10 mL of 0.00 1 M of timolol for 3 days at 25 ºC 0.9% NaCl solution (10 mL, without replacement) 25 ºC (Rotation not specified) Until 72 hours Hiratani et al., 2005b Timolol maleate Commercially available silicone hydrogel CL (Balafilcon A) (Thickness – not specified) Exterior surface of the CLs was coated by electrospinning technique PBS pH 7.4 (10 mL, 1 mL replaced at predetermined timepoints) 37 ºC, 90 rpm About 23 hours Mehta et al., 2017a Timolol maleate Commercially available silicone hydrogel CL (Balafilcon A) (Thickness – not specified) Coating the CL surface using a electrohydrodynamic process based on ejected droplet charge and timolol PBS pH 7.4 (10 mL, completely replaced at predetermined timepoints) 37 ºC, constantly stirring About 24 hours Mehta et al., 2017b Timolol maleate TRIS, DMAA, DBE - U12, and EGDMA (Thickness – 0.13 mm) Soaking in 2 mL of 5 mg/mL of timolol for 12 h at 25 ºC. Vitamin E was loaded by supercritical impregnation in high - pressure CO 2 PBS (5 mL, without replacement) 37 ºC, gently stirring About 8 hours Yokozaki and Shimoyam a, 2018 Timolol maleate HEMA, DMA, TRIS, and EGDMA (Thickness – 1 mm) Soaking in 1 mL of 0.2 or 1 mg/mL of timolol for 6 days PBS pH 7.4 (1 mL, completely replaced at predetermined timepoints) 37 ºC, 90 rpm 8 days Guidi et al., 2014b 1. Introduction 126 Timolol maleate HEMA, and EGDMA (Thickness – 0.2 mm) Timolol was mixed with the monomer solution Deionized water (3 mL, completely replaced every 24 h) 37 ºC, minimal stirring About 40 days Li et al., 2007 Timolol maleate DEAA, MAA, and EGDMA (Thickness – 0.3 mm) Soaking in 10 mL of 1mM of timolol solution for 3 days at 37 ºC 0.9% NaCl solution (10 mL, 0.6 mL replaced at predetermined timepoints) 37 ºC (Rotation not specified) 1 week Hiratani and AlvarezLorenzo, 2002 Timolol maleate HEMA, AAc, and EGDMA (Thickness – 0.2/ 0.9 mm) Soaking in 2 or 8 mL of 0.04, 0.06, 0.08, and 0.100 mM of timolol solutions 0.9% NaCl solution (2 or 8 mL, without replacement) 25 ºC, gently shaking Until 2 weeks Yañez et al., 2011b Timolol maleate HEMA, MAA, and EGDMA (Thickness – 0.1 mm) Soaking in 0.02 µM of timolol maleate solution for 72-120 h STF pH 7.2 (2 mL, without replacement) 37 ºC (Rotation not specified) 100 hours Anirudhan et al., 2016 Timolol maleate HEMA, MAA, MMA, and EGDMA (Thickness – 0.7 mm) Soaking in 5 mL of 0.5 mg/mL of timolol solution adjusted to pH 1.5, 3.5, 5.5, 7.5, 10 0.9% NaCl solution pH 5.5, or PBS pH 7.4, or STF pH 8 (15 mL, without replacement) 37 ºC (Rotation not specified) 48 hours AlvarezLorenzo et al., 2002 Timolol maleate HEMA, DEAA, MMA, SiMA, DMAA, MAA, and EGDMA (Thickness – 0.3 mm) Soaking in 10 mL of 1 mM of timolol solution for 3 days at 37 ºC 0.9% NaCl solution (10 mL, 0.6 mL replaced at predetermined timepoints) 37 ºC (Rotation not specified) 1 week Hiratani and AlvarezLorenzo, 2004 Timolol maleate HEMA, TRIS, DMA, and EGDMA (Thickness – 1 mm) Timolol was mixed with the monomer solution PBS pH 7.4 (1 mL, completed replaced at predetermined timepoints) 37 ºC, 90 rpm About 14 days Korogiann aki et al., 2015 Timolol maleate and brimonidine tartrate Commercially available hydrogel CLs (vasurfilcon A) (Thickness – not specified) Soaking in 25 mL of timolol maleate (0.65 mg/mL) or brimonidine tartrate (0.2 mg/mL) prepared in PBS PBS (25 mL, without replacement) Not specified 420 minutes Schultz et al., 2009 Timolol maleate, dexamethasone 21-disodium Commercially available silicone hydrogel CLs (Galyfilcon A, Soaking in 3 mL of drug solutions in PBS and drugethanol for 3 h PBS (2 mL, without replacement) Not specified Until 7 days Peng et al., 2010 ANA FILIPA PEREIRA DA MOTA 127 phosphate, fluconazole Senofilcon A, Lotrafilcon A, Lotrafilcon B, Balafilcon A) (Thickness – 0.08-0.1 mm) Timolol maleate, dexamethasone, and dexamethasone 21 - acetate TRIS, DMA, NVP, and EGDMA (Thickness – 0.1/ 0.2/ 0.4 mm) Soaking in 2 or 2.5 mL of drug solutions for 3 h PBS (Volume and replacement not specified) Not specified About 150 days Kim et al., 2008b Tobramycin and Amikacin TRIS, PEG (M n of 500 or 300), EGDMA, and isopropyl alcohol (Thickness – not specified) Soaking in 1 mL of 5 mg/mL tobramycin or amikacin in PBS for 24 h PBS pH 7.4 (0.5 mL, replaced at predetermined timepoints) 37 ºC, 100 rpm 24 hours Postic and Sheardow n, 2019 Transf erulic acid HEMA, GMA, EGPEM, and EGDMA (Thickness – 0.45 mm) Soaking in 5 mL of 0.01 mg/mL of transferulic acid for 48 h at 25 ºC SLF (5 mL, without replacement) 35 ºC, 300 rpm 24 hours Varela - Garcia et al., 2020b Triamcinolone acetonide HEMA, NVP, MA, and EGDMA (Thickness – 1 mm) Soaking in 10 mL of 8 - 19 µg/mL of triamcinolone acetonide solution for 4 days at 25 ºC Artificial lachrymal fluid pH 7.4 (5 mL, 1 mL replaced at predetermined timepoints) 37 ºC, 100 rpm 3 days García - Millán et al., 2015 Triamcinolone acetonide HEMA, MA, and EGDMA (Thickness – 1 mm) Commercially available conventional hydrogel CLs (Hilafilcon B) (Thickness – mean of 0.257 mm) Soaking in 10 mL of triamcinolone acetonide saturated solutions or 5 mL of triamcinolone acetonide nanosuspension for 2 days at 25 ºC STF pH 8 (5 mL, 1 mL replaced at predetermined timepoints) 37 ºC, 100 rpm 4 days García - Millán et al., 2017 Vancomycin, gentamicin Commercially available CLs from Acuity Contact Lenses, Hoddesdon, UK Filled the concave surface of the contact lens with 25 mg/mL of vancomycin or 40 mg/mL of gentamicin PBS (200 µl, completely replaced at predetermined timepoints) 37 ºC (Rotation not specified) 72 hours Hyatt et al., 2012 1. Introduction ANA FILIPA PEREIRA DA MOTA 128 1.6 REFERENCES Adrianto, M.F.; Annuryanti, F.; Wilson, C.G.; Sheshala, R.; Thakur, R.R.S. (2022). In vitro dissolution testing models of ocular implants for posterior segment drug delivery. Drug Deliv. and Transl. Res. 12(6), 1355-1375. Ali, M.; Byrne, M.E. (2009). Controlled release of high molecular weight hyaluronic acid from molecularly imprinted hydrogel contact lenses. Pharm. Res. 26, 714–726. Alvarez-Lorenzo, C.; Anguiano-Igea, S.; Varela-Garcia, A.; ViveroLopez, M.; Concheiro, A. (2019). Bioinspired hydrogels for drugeluting contact lenses. Acta Biomater. 84, 49–62. Alvarez-Lorenzo, C.; Hiratani, H.; Gómez-Amoza, J.L.; MartínezPacheco, R.; Souto, C.; Concheiro, A. (2002). Soft contact lenses capable of sustained delivery of timolol. J. Pharm. Sci. 91(10), 2182– 2192. Alvarez-Lorenzo, C.; Yañez, F.; Barreiro-Iglesias, R.; Concheiro, A. (2006). Imprinted soft contact lenses as norfloxacin delivery systems. J. Control. Release. 113(3), 236–244. Alvarez-Rivera, F., Concheiro, A.; Alvarez-Lorenzo, C. (2018). Epalrestat-loaded silicone hydrogels as contact lenses to address diabetic eye complications. Eur. J. Pharm. Biopharm. 122, 126–136. Alvarez-Rivera, F.; Serro, A.P; Silva, D.; Concheiro, A.; AlvarezLorenzo, C. (2019). Hydrogels for diabetic eyes: Naltrexone loading, release profiles and cornea penetration. Mater. Sci. Eng. C. 105, 110092. 1. Introduction 129 Andrade-Vivero, P.; Fernandez-Gabriel, E.; Alvarez-Lorenzo, C.; Concheiro, A. (2007). Improving the loading and release of NSAIDs from pHEMA hydrogels by copolymerization with functionalized monomers. J. Pharm. Sci. 96(4), 802–813. Anirudhan, T.S.; Nair, A.S.; Parvathy, J. (2016). Extended wear therapeutic contact lens fabricated from timolol imprinted carboxymethyl chitosan-g-hydroxy ethyl methacrylate-gpolyacrylamide as a onetime medication for glaucoma. Eur. J. Pharm. Biopharm. 109, 61–71. Ashtiani, M.K.; Zandi, M.; Shokrollahi, P.; Ehsani, M.; Baharvand, H. (2020). Chitosan surface modified hydrogel as a therapeutic contact lens. Polym. Adv. Technol. 31, 741–748. Aveyard, J.; Deller, R.C.; Lace, R.; Williams, R.L.; Kaye, S.B.; Kolegraff, K.N.; Curran, J.M.; D'Sa, R.A. (2019). Antimicrobial Nitric Oxide Releasing Contact Lens Gels for the Treatment of Microbial Keratitis. ACS Appl Mater Interfaces. 11(41), 37491-37501. Bajgrowicz, M.; Phan, C.M.; Subbaraman, L.N.; Jones, L. (2015). Release of ciprofloxacin and moxifloxacin from daily disposable contact lenses from an in vitro eye model. Invest. Ophthalmol. Vis. Sci. 56(4), 2234–2242. Bao, Q.; Burgess, D.J. (2018). Perspectives on physicochemical and in vitro profiling of ophthalmic ointments. Pharm. Res. 35(12), 234. Bao, Q.; Jog, R.; Shen, J.; Newman, B.; Wang, Y.; Choi, S.; Burgess, D.J. (2017). Physicochemical attributes and dissolution testing of ophthalmic ointments. Int. J. Pharm. 523(1), 310–319. ANA FILIPA PEREIRA DA MOTA 130 Bao, Q.; Newman, B.; Wang, Y.; Choi, S.; Burgess, D.J. (2018). In vitro and ex vivo correlation of drug release from ophthalmic ointments. J. Control. Release. 276, 93–101. Behl, G.; Iqbal, J.; O’Reilly, N.J.; McLoughlin, P.; Fitzhenry, L. (2016). Synthesis and characterization of poly(2-hydroxyethylmethacrylate) contact lenses containing chitosan nanoparticles as an ocular delivery system for dexamethasone sodium phosphate. Pharm. Res. 33(7), 1638–1648. Bengani, L.C.; Chauhan, A. (2013). Extended delivery of an anionic drug by contact lens loaded with a cationic surfactant. Biomaterials. 34(11), 2814–2821. Bengani, L.C.; Kobashi, H.; Ross, A.E.; Zhai, H.; Salvador-Culla, B.; Tulsan, R.; Kolovou, P.E.; Mittal, S.K.; Chauhan, S.K.; Kohane, D.S.; Ciolino, J.B. (2020). Steroid-eluting contact lenses for corneal and intraocular inflammation. Acta Biomater. 116, 149-161. Bertens, C.J.F.; Gijs, M.; van den Biggelaar, F.J.H.M.; Nuijts, R.M.M.A. (2018). Topical drug delivery devices: A review. Exp. Eye Res. 168, 149-160. Bin Sahadan, M.Y.; Tong, W.Y.; Tan, W.N.; Leong, C.R.; Bin Misri, M.N.; Chan, M.; Cheng, S.Y.; Shaharuddin, S. (2019). Phomopsidione nanoparticles coated contact lenses reduce microbial keratitis causing pathogens. Exp. Eye Res. 178, 10-14. Braga, M.E.; Costa, V.P.; Pereira, M.J.; Fiadeiro, P.T.; Gomes, A.P.; Duarte, C.M.; de Sousa, H. C. (2010). Effects of operational conditions on the supercritical solvent impregnation of acetazolamide in Balafilcon A commercial contact lenses. Int. J. Pharm. 420(2), 231-43. 1. Introduction 131 BusinessWire (2021). MediPrint™ Ophthalmics Announces Successful Completion of Its SIGHT-1 Phase 2a Clinical Study. https://www.businesswire.com/news/home/20210316005609/en/ (accessed November 2021). Byrne, M.E.; Hilt, J.Z.; Peppas, N.A. (2008). Recognitive biomimetic networks with moiety imprinting for intelligent drug delivery. J. Biomed. Mater. Res. A. 84(1), 137–147. Cavas-Martínez, F.; Fernandez-Pacheco, D.G.; De la Cruz-Sanchez, E.; Nieto Martinez, J.; Fernandez Cañavate, F.J.; Vega-Estrada, A.; PlazaPuche, A.B.; Alio, J.L. (2014). Geometrical custom modeling of human cornea in vivo and its use for the diagnosis of corneal ectasia. PLoS One. 9, e110249. Choi, S.H.; Lionberger, R.A. (2016). Clinical, pharmacokinetic, and in vitro studies to support bioequivalence of ophthalmic drug products. AAPS J. 18(4), 1032–1038. Choi, J.H.; Li, Y.; Jin, R.; Shrestha, T.; Choi, J.S.; Lee, W.J.; Moon, M.J.; Ju, H.T.; Choi, W.; Yoon, K.C. (2019). The Efficiency of Cyclosporine A-Eluting Contact Lenses for the Treatment of Dry Eye. Curr. Eye Res. 44(5), 486-496. Chrai, S.S.; Patton, T.F.; Metha, A.; Robinson, J.R. (1973). Lacrimal and instilled fluid dynamics in rabbit eyes. J. Pharm. Sci. 62(7), 1112– 1121. Ciolino, J.B.; Hoare, T.R.; Iwata, N.G.; Behlau, I.; Dohlman, C.H.; Langer, R.; Kohane, D.S. (2009). A drug-eluting contact lens. Invest. Ophthalmol. Vis. Sci. 50(7), 3346-3352. ANA FILIPA PEREIRA DA MOTA 132 Ciolino, J.B.; Hudson, S.P.; Mobbs, A.N.; Hoare, T.R.; Iwata, N.G.; Fink, G.R.; Kohane, D.S. (2011). A prototype antifungal contact lens. Invest. Ophthalmol. Vis. Sci. 52(9), 6286–6291. Ciolino, J.B.; Ross, A.E.; Tulsan, R.; Watts, A.C.; Wang, R.F.; Zurakowski, D.; Serle, J.B.; Kohane, D.S. (2016). Latanoprost-Eluting Contact Lenses in Glaucomatous Monkeys. Ophthalmol. 123(10), 2085-2092. Ciolino, J.B.; Stefanescu, C.F.; Ross, A.E.; Salvador-Culla, B.; Cortez, P.; Ford, E.M.; Wymbs, K.A.; Sprague, S.L.; Mascoop, D.R.; Rudina, S.S.; Trauger, S.A.; Cade, F.; Kohane, D.S. (2014). In vivo performance of a drug-eluting contact lens to treat glaucoma for a month. Biomaterials. 35(1), 432-439. Costa, V.P.; Braga, M.E.M.; Duarte, C.M.M.; Alvarez-Lorenzo, C.; Concheiro, A.; Gil, M.H.; de Sousa, H.C. (2010a). Anti-glaucoma drugloaded contact lenses prepared using supercritical solvent Impregnation. J. Supercrit. Fluids. 53, 165–173. Costa, V.P.; Braga, M.E.M.; Guerra, J.P.; Duarte, A.R.C.; Duarte, C.M.M.; Leite, E.O.B.; Gil, M.H.; de Sousa, H.C. (2010b). Development of therapeutic contact lenses using a supercritical solvent impregnation method. J. Supercrit. Fluids. 52, 306–316. Danion, A.; Arsenault, I.; Vermette, P. (2007). Antibacterial activity of contact lenses bearing surface-immobilized layers of intact liposomes loaded with levofloxacin. J. Pharm. Sci. 96(9), 2350–2363. Del Amo, E.M.; Urtti, A. (2008). Current and future ophthalmic drug delivery systems. A shift to the posterior segment. Drug Discov. Today. 13(3-4), 135–143. 1. Introduction 133 Deng, J.; Chen, S.; Chen, J.; Ding, H.; Deng, D.; Xie, Z. (2018). Selfreporting colorimetric analysis of drug release by molecular imprinted structural color contact lens. ACS Appl. Mater. Interfaces. 10(40), 34611–34617. Desai, A.R.; Maulvi, F.A.; Desai, D.M.; Shukla, M.R.; Ranch, K.M.; Vyas, B.A.; Shah, S.A.; Sandeman, S.; Shah, D.O. (2020). Multiple drug delivery from the drug-implants-laden silicone contact lens: Addressing the issue of burst drug release. Mater. Sci. Eng. C. 112, 110885. Desai, A.R.; Maulvi, F.A.; Pandya, M.M.; Ranch, K.M.; Vyas, B.A.; Shah, S.A.; Shah, D.O. (2018). Co-delivery of timolol and hyaluronic acid from semi-circular ring implanted contact lenses for the treatment of glaucoma: in vitro and in vivo evaluation. Biomater. Sci. 6, 1580. Dixon, P.; Chauhan, A. (2017). Effect of the surface layer on drug release from delefilcon-A (Dailies Total 1®) contact lenses. Int. J. Pharm. 529(1-2), 89–101. Dixon, P.; Ghosh, T.; Mondal, K.; Konar, A.; Chauhan, A.; Hazra, S. (2018). Controlled delivery of pirfenidone through vitamin E-loaded contact lens ameliorates corneal inflammation. Drug Deliv. Transl. Res. 8(5), 1114–1126. Doane, M.G. (1989). An instrument for in vivo tear film interferometry. Optom. Vis. Sci. 66(6), 383–388. Dominguez-Godinez, C.; Carracedo, G.; Pintor, J. (2018). Diquafosol delivery from silicone hydrogel contact lenses: improved effect on tear secretion. J. Ocular Pharm. Ther. 34(1-2), 170–176. ANA FILIPA PEREIRA DA MOTA 236 diabetic retinopathy [Gupta et al., 2004; Ozkiris et al., 2007; Chung et al., 2017; Nielsen and Nordestgaard, 2014]. Diabetes mellitus is one of the world's greatest health challenges and is reaching epidemic proportions [International Diabetes Federation, 2021]. The lack of glycaemia control causes a variety of damages in multiple structures of the anterior and posterior segments of the eye, triggering the development of diabetic keratopathy, dry eye syndrome, cataracts, glaucoma, diabetic retinopathy and macular edema [Sayin, 2015; Misra et al., 2016]. Dysregulation of multiple molecular pathways contributes to upregulation of growth factors and inflammatory cytokines [Gologorsky et al., 2012; Wong et al. 2016]. Increased levels of plasma cytokines (tumor necrosis factor-α and interleukin-6) have been found in ocular fluid of diabetic patients [Esposito et al., 2002]. Other inflammatory mediators as prostaglandin E2 may also have a pathogenic role in diabetic retinopathy [Schoenberger et al., 2012]. Inhibition of mevalonate pathway by statins reduces oxidative stress, endothelial dysfunction, inflammation, and angiogenesis; thereby improving retina health [Zhang and McGwin, 2007]. Also, statins improve the integrity of endothelial cells, hence preserving the blood–retinal barrier [Ioannidou et al., 2017]. However, high oral doses may cause adverse systemic collateral effects, mainly on liver and muscle tissue [Murphy et al., 2020]. Therefore, the development of statin ophthalmic formulations that can provide sufficiently high levels of statins in the anterior and posterior segment of the eye, avoiding systemic absorption, may notably improve the ocular outcomes and the overall safety of the treatment. Statin ophthalmic formulations would also be useful for people who do not have hypercholesterolemia. 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 237 Pilot studies have shown that topical ocular administration of atorvastatin eye drops is well tolerated in prolonged treatments of dry eye and blepharitis, but frequent instillation is required due to low ocular bioavailability [Ooi et al., 2015]. Sustained drug release from contact lenses (CLs) may provide more prolonged levels in tear fluid and with fewer fluctuations compared to eye drops, which in turn favours drug penetration into anterior eye segment [Li and Chauhan, 2006; Dixon et al., 2018; Wu et al., 2021]. Indeed, the first medicated CL has been approved in 2021 for the management of allergic conjunctivitis [Johnson & Johnson News, 2021]. Moreover, it has been recently demonstrated in animal models that such an increase in the amount permeated may make even possible that some specific drugs may reach the posterior segment as recorded for, for example, epalrestat [Zhu and Sheng, 2020], ofloxacin [Shikamura et al., 2016] and timolol [Maulvi et al., 2019]. Nevertheless, the development of medicated CLs has to face up to the problem of that the affinity of commercially available CLs for most used ophthalmic drugs is low. Therefore, CLs have to be ad hoc designed for each target drug in order to uptake therapeutic doses and to provide sustained release on the eye surface [Alvarez-Lorenzo et al., 2019]. A variety of approaches, which include film coatings, molecular imprinting and drug nanoencapsulation, among others, are under investigation to overcome the loading and release limitations shown by CLs [Kakisu et al., 2013; Minami et al., 2019; Sekar and Chauhan, 2019; DiPasquale et al., 2021; Toffoletto et al., 2021]. Among these approaches, the recreation of biomimetic receptors that resemble the composition and spatial ordering of the physiological receptor of the drug without compromising relevant optical properties of CLs is pointed out as a successful strategy [Alvarez-Lorenzo et al., 2019]. Another barrier to be solved for the development and clinical translation ANA FILIPA PEREIRA DA MOTA 238 of drug-loaded CLs is that in vitro release methods that can predict in vivo release behaviour are still to be identified because there are too many variables involved [Tieppo et al., 2014; Mahomed et al., 2016; Paradiso et al., 2017; Phan et al., 2021]. In most reports on CLs the in vitro release tests are used for optimization of drug release kinetics although with limited basis of how the in vitro release profiles may correlate with the in vivo release kinetics [Hui and Wilcox, 2016; Pereira-da-Mota et al., 2022]. In a previous study, CLs soaked in atorvastatin calcium solution demonstrated improved drug accumulation ex vivo in cornea and sclera, but the drug did not penetrate further probably because its large molecular weight (1155.3 g/mol) and hydrophobicity (Log P = 6.36) caused the ocular permeability to be very low [Pereira-da-Mota et al., 2021]. Thus, the present work relied on the hypothesis of that a smaller and more hydrophilic statin, like pravastatin sodium (446.5 g/mol; Log P = - 0.23 [Murphy et al., 2020]; Figure 4.1) adequately formulated in a CL may get access to deeper eye tissues. Accordingly, the aim of this work was to design CLs bioinspired in the natural receptor of pravastatin, i.e., the HMG-CoA reductase. This natural receptor contains amino acids with a wide variety of hydroxyl groups, hydrophobic groups, and high density of protonated amino groups [Jain et al., 2007], which could be resembled using 2-hydroxyethyl methacrylate (HEMA), ethylene glycol phenyl ether methacrylate (EGPEM), and N-(3-aminopropyl) methacrylamide hydrochloride (APMA), respectively (Figure 4.1). All hydrogels were characterized in terms of solvent uptake, light transmission, mechanical properties and pravastatin loading and release capacity. The effects of high hydrostatic pressure (HHP) and autoclave sterilization on pravastatin solutions and pravastatin-loaded hydrogels were also evaluated. Preliminary screening of cytocompatibility and ocular tolerance was 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 239 carried out with Balb/3T3 fibroblasts and HET-CAM assay. Pravastatin permeability was assessed ex vivo using porcine cornea and sclera tissues. Anti-inflammatory activity of pravastatin-loaded hydrogels was also investigated. Finally, in vivo experiments with the most promising CLs were performed in New Zealand white rabbits to investigate the in vivo release profiles of pravastatin-loaded CLs and to quantify pravastatin levels in anterior and posterior segment ocular tissues. An additional aim of the work was to identify in vitro release conditions that may provide in vitro-in vivo correlations (IVIVC). To the best of our knowledge, this is the first time that the effect of the volume and protein composition of the release medium on Levy plots is evaluated. Figure 4.1. Molecular structure of pravastatin sodium (A) and monomers used to synthesize the hydrogels: (B) 2-hydroxyethyl methacrylate (HEMA); (C) N-(3aminopropyl) methacrylamide hydrochloride (APMA), and (D) ethylene glycol phenyl ether methacrylate (EGPEM). ANA FILIPA PEREIRA DA MOTA 240 4.2 MATERIALS AND METHODS 4.2.1 Materials Pravastatin sodium was supplied by Biocon Limited (Bengaluru, Karnataka, India). 2-Hydroxyethyl methacrylate (HEMA) was from Merck (Darmstadt, Germany), and N-(3-aminopropyl) ethacrylamide hydrochloride (APMA) from PolySciences Inc. (Warrington, PA, USA). Ethylene glycol dimethacrylate (EGDMA), ethylene glycol phenyl ether methacrylate (EGPEM), 2,2′ -azobis(isobutyronitrile) (AIBN), and dichlorodimethylsilane were from Sigma-Aldrich (Steinheim, Germany). Simulated lachrymal fluid (SLF), pH 7.4, was prepared in 1 liter of ultrapure water as follows: 2.18 g NaHCO3 from Probus (Barcelona, Spain), 6.78 g NaCl from Scharlab (Barcelona, Spain), 1.38 g KCl, and 0.084 g CaCl2⋅2H2O from Merck (Darmstadt, Germany). Bicarbonate Ringer's solution, pH 7.2, was prepared by mixing 100 mL of buffer solution A (0.071 g KCl, 1.24 g NaCl, 0.49 g NaHCO3, and 0.02 g NaH2PO4, from Merck (Darmstadt, Germany)) and buffer solution B (0.031 g MgCl2, and 0.023 g CaCl2 from Panreac (Barcelona, Spain)). Ultrapure water (resistivity >18.2 MΩ cm; MilliQ®, Millipore Ibérica, Madrid, Spain) was obtained by reverse osmosis. Methanol 99.9% for LC-MS grade was from Fisher Scientific (Loughborough, UK). Acetonitrile for HPLC LC-MS grade and NaOH were from VWR Chemicals (Fontenary-Sous-Bois, France). Balb/3T3 fibroblasts (ATCC CCL-163™) and THP-1 cells (ATCC TIB-202™) were provided by American Type Culture Collection (ATCC, Manassas, VA, USA). Dulbecco's modified Eagle medium (DMEM), Dulbecco's phosphate buffered solution (DPBS), and RPMI 1640 was from Fisher Scientific (Newington, NH, USA). Fetal bovine serum, antibiotic solution (penicillin and streptomycin), lipopolysaccharides (LPS) from Escherichia coli O111:B4, phorbol 12-myristate 13-acetate 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 241 (PMA), and TrypLE® were acquired from Sigma-Aldrich (St. Louis, MO, USA). Cell counting kit-8 (CCK-8) was from Dojindo Molecular Technologies (Rockville, MD, USA). Schirmer test strips were from Contacare Ophthalmics and Diagnostics (Gujarat, India). Pentobarbital (400 mg/mL) was from Richter Pharma AG, Wels, Austria. 4.2.2 Synthesis of Hydrogel Discs and CLs Four different hydrogels were synthesized combining HEMA, EGPEM, and APMA, as summarized in Table 4.1. Briefly, the monomers were mixed into glass vials and magnetically stirred (200 rpm) for 120 min at room temperature. Then, EGDMA (crosslinker) and AIBN (initiator) were added, and the solutions were magnetically stirred again (100 rpm) for 10 min. The resulting solutions were immediately injected into moulds composed of two glass plates (10 × 10 cm) separated by a silicone frame with 0.3 mm of thickness, and heated at 50 °C for 12 h, and then to 70 °C for further 24 h for thermal polymerization [Alvarez-Lorenzo et al., 2002]. After polymerization, each hydrogel sheet was boiled (1 L water) for 15 min to remove unreacted substances and facilitate cutting in the form of 10 mm discs. The discs were alternatively washed in water and 0.9% NaCl twice a day (200 rpm magnetic stirring, room temperature), until complete removal of the unreacted monomers as monitored spectrophotometrically (UV–Vis spectrophotometer, Agilent 8453, Waldbronn, Germany). Finally, the discs were dried at 70 °C for 24 h before being stored and used in further experiments. CLs were prepared with the same composition as E200A40 hydrogel, but adding more AIBN (14.79 mg, initiator) for a complete ANA FILIPA PEREIRA DA MOTA 242 polymerization of the thinner hydrogels. The monomer solution was pipetted (60 μL) in curved polypropylene moulds (similar to those used for industrial production of CLs) to have final CL dimensions in the hydrated state (phosphate buffer pH 7.4) of approx. 12 mm diameter, 7.8 mm curvature, and 0.1 mm thickness. After polymerization at 50 °C for 12 h and then at 70 °C for 24 h, all CLs were washed as described above. Finally, the CLs were dried at 70 °C for 24 h. Table 4.1. Composition of the monomer mixtures used to synthesize the hydrogels. To each mixture, 4.93 mg AIBN were added. Hydrogel Code HEMA (mL) EGDMA (µL) EGPEM (µL) APMA (mg) E0A0 3 12.10 - - E200A0 3 12.10 112.50 - E200A40 3 12.10 112.50 21.45 E0A40 3 12.10 - 21.45 4.2.3 Hydrogel Characterization Solvent uptake capacity was evaluated, in triplicate, from the mass increase of previously weighed dried discs (W0) after being immersed in 10 mL of SLF at room temperature without agitation until equilibrium (Wt), as follows: 𝑆𝑜𝑙𝑣𝑒𝑛𝑡 𝑢𝑝𝑡𝑎𝑘𝑒 =  × 100 (Eq. 4.1) Water contact angles on dried and wet hydrogels (10 mm in diameter) were measured using the sessile drop method with a Phoenix300 goniometer (Surface Electro Optics Co., Suwon, Korea). The measurements were performed over 2 discs of each formulation, at room temperature, and with distilled water. A total of 10 images were 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 243 captured in 10 s, and the best-adjusted pictures were selected to obtain the contact angle value. Ion permeability was recorded in triplicate by fixing individual wet hydrogel discs between two flat Teflon washers (8 mm open diameter), which were fixed in a screw cap (open top) and screwed in a 5-mL HPLC vial. The bottom of the vials was previously removed. The vials were placed downward in beakers containing 15 mL distilled water at 35 °C and magnetically stirred (130 rpm) (Figure 4.2). The surface available for diffusion was 50.27 mm2. Then 4 mL NaCl 0.1 M solution were added to the vial and the conductivity was recorded as a function of time using an InLab® 741 conductivity probe fitted to a Mettler Toledo T70 titrator (Greifensee, Switzerland). Figure 4.2. Setup used to measure the ion permeability of the hydrogels. ANA FILIPA PEREIRA DA MOTA 244 The transmittance of the hydrogels swollen in SLF was recorded in an UV–Vis spectrophotometer (Agilent 8453, Waldbronn, Germany), in the 200–800 nm range (n = 3). Mechanical properties of dog bone-shaped hydrogels (18 × 5 mm, in the center 6 × 2.5 mm) were tested using a TA.XT Express Texture Analyzer (Stable Micro Systems, Godalming, UK) fitted with a 50 N load cell. The hydrogel probes (n = 5) were fixed to the upper and lower clamps, and the tensile test was carried out applying a test speed of 0.25 mm/s and 0.005 N trigger force. The elastic modulus was calculated from the initial slope of the engineered stress-strain curve [Kim et al., 2018]. 4.2.4 Pravastatin Loading and Release Pravastatin was loaded by soaking the dried discs in pravastatin sodium (0.1 mg/mL; 10 mL) solution. The experiment was repeated several times and in each run 4 discs of each composition were tested. The vials were kept under oscillatory movement (180 osc/min) at room temperature and protected from light. After 48 h, the absorbance of the loading solution was measured at 238 nm (UV–Vis spectrophotometer, Agilent 8453, Waldbronn, Germany). The amount loaded was calculated as the difference between the initial and final amount of pravastatin in the solutions using previously prepared calibration curves. The network/water partition coefficient was estimated as follows [Kim et al., 1992]: 𝐾 = ×   (Eq. 4.2) 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 245 In this equation, Vs is the solvent volume sorbed by the hydrogel, Wp the dried hydrogel weight, and C0 the initial concentration of the loading solution. In vitro drug release patterns were recorded by immersing loaded discs (n = 4) in vials containing 10 mL of SLF (pH = 7.4) at 37 °C. The discs were previously rinsed with SLF to remove the excess of pravastatin from the hydrogel surface. The experiments were performed for 120 h under oscillatory movement (180 osc/min) and protected from light. At predetermined time points, 3 mL of the release medium were collected, absorbance measured at 238 nm and the sample returned to the vial immediately. After 8 h of experiment, 2 mL of the release medium were collected and replaced by the same volume of fresh SLF to avoid false plateaus. The test was carried out under sink conditions. 4.2.4.1 Pravastatin Release from Sterile CLs The synthetized CLs (E200A40 composition) were immersed in a pravastatin solution (0.1 mg/mL) for 48 h and then sterilized in closed packages by high hydrostatic pressure (HHP) (Section 4.2.8). The CLs were stored at room temperature until release experiments were performed. The CLs were immersed in 2 or 10 mL of SLF without or with lysozyme or bovine serum albumin (2.68 mg/mL). The experiments were performed at 37 °C, under oscillatory movement (180 osc/min) and at predefined timepoints, 150 μL were removed and replaced by the same volume of fresh SLF, or SLF plus lysozyme or BSA. The amount of pravastatin released from the CLs was quantified by HPLC (as described in section 4.2.7) after denaturation of the proteins (section 4.2.9.2). The amount of protein deposited on blank (non-drug loaded) CLs was monitored at 280 nm (UV–Vis ANA FILIPA PEREIRA DA MOTA 252 discarded. No adverse events were detected. On the first day, two rabbits of 8 h wear CLs, and one rabbit of eye drop groups were assayed. On the second day, one rabbit of 8h wear CLs and three rabbits of the eye drop groups were assayed. On the third day, the three rabbits of 10 h wear CLs group were assayed. For the sample size calculation the 3R's principles were followed minimizing the number of rabbits used and trying to obtain the most reliable results. A total of 10 animals were used in this two-treatment parallel-design study. This design allowed the study detected a treatment difference at 0.05 significance level with a probability of 80% if the difference between treatments was 2.03 times the standard deviation. Figure 4.3. Animals distribution in in vivo experiments. In vivo experiments (n= 10) CLs group (n= 6) 8 h using CLs (n= 3) (Euthanized after 8 h) 10 h using CLs (n = 3) (Euthanized after 24 h) Eye drops group (n= 4) (Euthanized after 8 h) 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 253 4.2.9.1 In vivo Release The ocular surface of each rabbit was carefully inspected with a VX75 slit lamp (Luneau Technology, Chartres, France) before and after 8 h of the experiment. The CLs were removed from the sterilized bags, rinsed with sterile saline solution for CLs (to remove excess drug on the surface of the CL), and carefully placed on the rabbit right eye without local anaesthesia. The CLs were placed in the cornea and bulbar conjunctiva below the nictitating membrane. No nictitating membranectomy or tarsorrhaphy was needed. To prevent drying of CL on the surface of the eye, every 15 min the rabbits' eyes were closed for 1 min. The left eye was kept as control without treatment. In the eye drop group, a single drop (50 μL) of 2 mg/mL of pravastatin solution, which contained the same amount of pravastatin as that released from the CLs after 8–10 h, was gently instilled in the lower conjunctival sac of the right eye using a micropipette. Before and after administration (t = 5 min, 15 min, 30 min, and every hour until 8 or 10 h) tear fluid samples were collected using Schirmer test strips which were placed in the tarsal conjunctiva of the lower lid for 10 s with closed eyes [van Bijsterveld, 1969]. The volume of tears collected from the Schirmer strips was recorded as millimetres of moistened strip. The mean residence time (MRT) was calculated as the area under the curve of concentration versus time (AUMC) referred to the area under the curve of concentration versus time (AUC). Both AUMC and AUC were calculated from the experimental data applying the trapezoidal rule [Hiratani et al., 2005]. After wearing, CLs were extracted with SLF (0.8 mL under stirring at 37 °C for 24 h) and the remnant pravastatin was quantified by HPLC. ANA FILIPA PEREIRA DA MOTA 254 4.2.9.2 Quantification of Pravastatin in Tear Fluid The Schirmer test strips were placed in 2 mL Eppendorf tubes with 200 μL of SLF for 12 h at 4 °C. For protein denaturation, the samples were vortexed for 1 min, and the strips were removed from the tubes. Then, SLF solution was heated at 98 °C for 2 min, cooled in ice for 10 min, and centrifuged at 13,000 rpm for 10 min at 25 °C [Lazarowski et al., 2003]. Finally, the supernatants were collected and stored at − 20 °C until HPLC analysis (as described in Section 4.2.7). In preliminary tests, this protein denaturation method was shown to reproducibly recover >95% pravastatin present in the samples. In vitro-in vivo correlations (IVIVC) were attempted through Levy plot analysis, where the X-axis was reported as the percentage of drug released in vitro at a certain time, and the Y-axis the percentage of drug released in the tear fluid at the same time estimated as follows [Xu et al., 2019]. 𝑃𝑟𝑎𝑣𝑎𝑠𝑡𝑎𝑡𝑖𝑛 𝑟𝑒𝑙𝑒𝑎𝑠𝑒𝑑 𝑖𝑛 𝑣𝑖𝑣𝑜 (%)=  ×100 (Eq. 4.4) Regression analysis was carried out using SigmaPlot for Windows v.14 software (Systat Sofware Inc., Germany). 4.2.9.3 Quantification of Pravastatin in Ocular Tissues All the rabbits were euthanized by intravenous administration of 0.75 mL/Kg of propofol and 0.5 mL/Kg of pentobarbital sodium (approx. 2 mL, Exagon 400 mg/mL). Following euthanization, aqueous humour was directly extracted from the anterior chamber using a needle and stored at 4 °C until being processed for protein denaturation. Then, the eyes were enucleated and immediately stored at − 80 ◦C until tissue 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 255 dissection, when cornea, sclera, crystalline lens, vitreous humour, and retina were weighed and separately placed in Eppendorf tubes. SLF was added to cornea (500 μL), crystalline (500 μL), sclera (800 μL) and retina (200 μL). The tissues remained immersed in SLF for 12 h at 4 °C, and then the protein denaturation process was performed as previously described. The supernatant was collected and stored at − 20 °C until UPLC analysis. Before UPLC analysis all the samples were centrifuged at 5,000 rpm, 4 °C for 10 min. UPLC analysis was performed on a Waters Acquity UPLC H-Class coupled with a Xevo TQD MS System. The chromatographic separation was performed on a HypersilGOLD C18 (Thermo-Fisher, 1.9 μm, 2.1 × 50 mm) column with a column temperature of 35 °C, a flow rate of 0.6 mL/min, and a mobile phase containing 0.1% formic acid/water (A) and 0.1% formic acid/acetonitrile (B). The gradient was programmed as follows: 0–0.1 min 5% B, 0.1–1.0 min 5–100% B, 1.0–2.0 min 100% B, 2.0–2.1 min 100–95% B, and 2.1–2.5 min 5% B. Electrospray ionization (ESI) was run in positive mode with a source temperature of 150 °C and a desolvation temperature of 500 °C. Capillary voltage was set to 3 kV and the cone voltage was set to 45 V. Desolvation gas flow was 900 L/h and cone gas flow was set to 50 L/h. Pravastatin was monitored in selected ion recording (SIR) mode using the m/z of 447.101. 4.2.10 Statistical Analysis Statistical analysis was performed using Statgraphics Centurion 18 v. 18.1.13 (Statgraphics Technologies, Inc., Warrenton, VA, USA). The descriptive data were presented as mean ± standard deviation. One-way analysis of variance (ANOVA) followed by Multiple Range Test was carried out. The level of significance was 0.05. ANA FILIPA PEREIRA DA MOTA 256 4.3 RESULTS AND DISCUSSION 4.3.1 Hydrogels Synthesis and Characterization Hydrogels were synthesized combining HEMA with EGPEM and APMA searching for a good balance between solvent uptake, light transmission and mechanical properties, and binding affinity for statins. APMA was chosen as functional monomer able to mimic cationic amino acids involved in the binding site of HMG-CoA reductase [Istvan and Deisenhofer, 2001]. Incorporation of EGPEM was proposed for hydrophobic interactions with the benzyl group of pravastatin sodium enhancing drug binding through π–π stacking [Pereira-da-Mota et al., 2021]. EGPEM and APMA dissolved quickly in HEMA solution at concentrations of 200 and 40 mM, respectively. The hydrogels were coded as ExAy, where x referred to the concentration of EGPEM in mM, and y referred to APMA concentration in mM in the monomers mixture. Dried hydrogels swelled quite fast after being immersed in SLF, achieving the solvent uptake plateau in 1 h (Figure 4.4A). Hydrogels prepared with the hydrophobic monomer EGPEM showed lower water uptake (44.05 ± 0.40% for E200A0) compared to those bearing only APMA (52.91 ± 1.97% for E0A0) (p < 0.05). APMA did not alter the swelling of HEMA hydrogels. EGPEM-bearing hydrogels also showed greater water contact angles when tested both in the dried and wet states (Table 4.2); nevertheless, in all cases the contact angles were below 90° and the hydrogels can be considered to have hydrophilic surface. NaCl permeability, calculated as recorded in Figure 4.4B ranked in the order E0A40 (8.3⋅10 cm/s) > E0A0 (7.8⋅10 cm/s) > E200A0 (5.7⋅10 cm/s) > E200A40 (3.5⋅10 cm/s). The higher values corresponded to the most hydrophilic hydrogels in good agreement with previous reports 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 257 and in the range typical of CLs [Peng and Chauhan, 2012; Pozuelo et al., 2014]. Figure 4.4. (A) Solvent uptake profiles in SLF at room temperature and (B) data recorded as increase in NaCl concentration in the receptor as a function of time. Codes as in Table 4.1. Table 4.2. Water contact angles of the hydrogels before and after being swollen in water using the sessile drop method. Hydrogel codes as in Table 1 (n = 4; mean value ± standard deviations). Hydrogel Code Dry hydrogels Hydrated hydrogels E0A0 56.39 ± 4.40 61.68 ± 1.89 E200A0 72.48 ± 0.11 64.78 ± 1.60 E200A40 72.57 ± 2.46 71.05 ± 2.34 E0A40 49.50 ± 1.99 58.08 ± 0.41 For all compositions, light transmission of wet hydrogels was close to 90% in the visible range (Figure 4.5A) fulfilling the requirements for CLs. Representative optical images of discs and CLs are shown in Time (hours) 0 1 2 3 4 5 6 24 Solvent uptake (%) 0 10 20 30 40 50 60 E0A0 E200A0 E200A40 E0A40 Wavelength (nm) 200 300 400 500 600 700 Transmittance (%) 0 20 40 60 80 100 E0A0 E200A0 E200A40 E0A40 90% A B Time (min) 0 30 60 90 120 150 180 210 NaCl concentration (M) 0 5.0x10 -5 10 -4 1.5x10 -4 2.0x10 -4 2.5x10 -4 3.0x10 -4 3.5x10 -4 E0A40 E0A0 E200A0 E200A40 B ANA FILIPA PEREIRA DA MOTA 258 Figure 4.6. The elastic modulus of the hydrogels ranged between 0.65 and 0.79 MPa (Figure 4.5B); E0A40 hydrogel presented a significantly lower value compared to the other hydrogels (p < 0.05). Nevertheless, the obtained values were for all hydrogels in the range of those typical for HEMA-based CLs (0.2–2.0 MPa) [Musgrave and Fangg, 2019]. Figure 4.5. (A) Light transmittance (%) of the hydrogels after being swollen in SLF. Dashed line indicates 90% transmittance; (B) elastic modulus (MPa) obtained for hydrogels hydrated in 0.9% NaCl. Codes as in Table 4.1 E0A0 E200A0 E200A40 E0A40 Elastic Modulus (MPa) 0.0 0.2 0.4 0.6 0.8 1.0 Wavelength (nm) 200 300 400 500 600 700 Transmittance (%) 0 20 40 60 80 100 E0A0 E200A0 E200A40 E0A40 90% A B 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 259 Figure 4.6. Pictures of (A) hydrated hydrogel discs and (B1 and B2) E200A40 CLs. Codes as in Table 4.1. 4.3.2 Pravastatin Loading and In vitro Release Pravastatin loading was carried out by immersing the dried discs in pravastatin solution (0.1 mg/mL in water) for 48 h at room temperature. Loading of pravastatin during CL synthesis was discarded due to the poor solubility of the drug in the monomers mixture and the potential loss of a fraction of the incorporated drug during the required washing. Control E0A0 hydrogel loaded a small amount of pravastatin (0.42 ± 0.22 mg/g, Figure 4.7A). Incorporation of EGPEM monomer (E200A0 hydrogel) did not significantly improve the affinity for pravastatin (0.32 ± 0.17 mg/g; p > 0.05). Differently, the copolymerization of HEMA with APMA monomer (E200A40 and E0A40 hydrogels) remarkably increased the amount of pravastatin loaded (7.26 ± 0.59 and 6.70 ± 0.20 mg/g, respectively; p < 0.05). The KN/W values recorded for APMAcontaining hydrogels (66.44 ± 2.30 for E0A40 and 72.17 ± 6.78 for E200A40) were about twenty-fold higher than those recorded for control hydrogels (3.70 ± 2.58 for E0A0) and hydrogels prepared with ANA FILIPA PEREIRA DA MOTA 260 EGPEM solely (2.75 ± 1.99 for E200A0) (p < 0.05). Compared to previous reports on atorvastatin [Pereira-da-Mota et al., 2021], KN/W values were lower for pravastatin, which may be related to the higher hydrophilicity of the latter. When transferred to SLF, APMA hydrogels released higher amounts of pravastatin than non-functionalized (E0A0) and EGPEM (E200A0) hydrogels (Figure 4.7B). Comparing both APMA hydrogels, E200A40 presented slower release rate in the first hours, but the total amount of pravastatin released after 120 h was higher (7.84 ± 0.66 mg/g) compared to E0A40 hydrogel (7.07 ± 0.45 mg/g). The slower release rate recorded for E200A40 in the first hours may be related to the higher KN/W and pointed out to the role of the drug-EGPEM hydrophobic interactions in sustaining the release of the drug. These findings also suggest that both APMA and EGPEM are needed for resembling the multiple-point redundant interactions of the physiological receptor [Ribeiro et al., 2011; Tieppo et al., 2012]. Subsequent studies were carried out with both types of hydrogels containing APMA (E0A40 and E200A40). 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 261 Figure 4.7. (A) Pravastatin loaded by the hydrogel discs (0.3 mm thickness) after 48 h of soaking in the drug solution (0.1 mg/mL in water) at room temperature under oscillatory movement (180 osc/min); (B) release profiles from pravastatinloaded discs in SLF at 37 ◦C under oscillatory movement (180 osc/min). Codes as in Table 4.1. (n = 4; mean values and standard deviations). 4.3.3 Cytocompatibility and HET-CAM Cell viability tests were performed using sterilized non-loaded and pravastatin-loaded hydrogels against Balb/3T3 fibroblasts, which is one of the most sensitive cell lines for biomaterials testing [Wataha et al., 1994]. Two solutions with different concentration in pravastatin were also evaluated. After 24 h of incubation, cell viability was 88.5 ± 8.2% and 89.6 ± 6.0% for non-loaded E200A40 and E0A40 hydrogels, and 80.2 ± 10.6% and 76.1 ± 4.4% for pravastatin-loaded hydrogels, respectively (Figure 4.8). After 48 h of incubation cell viability increased for non-loaded hydrogels and slightly decreased for pravastatin-loaded ones (72.5 ± 8.4% and 71.3 ± 7.7% for E200A40 Time (hours) 0 6 12 18 24 120 Pravastatin released (mg/g) 0 2 4 6 8 E0A0 E200A0 E200A40 E0A40 E0A0 E200A0 E200A40 E0A40 Pravastatin loaded (mg/g) 0 2 4 6 8 A B ANA FILIPA PEREIRA DA MOTA 268 Pravastatin release in 2 mL of SLF was slightly slower than in 10 mL SLF, but the profiles run in parallel and no statistical differences could be detected due to the data variability. For 2 mL experiment, the incorporation of lysozyme and BSA proteins accelerated drug release, and the percentage released in the 8 to 10 h was statistically higher than in the absence of proteins (p < 0.05). Differently, lysozyme and BSA decreased release rate in 10 mL volume in the first 4 h but once again the percentage released in the 8 to 10 h was statistically higher than in the absence of proteins (p < 0.05). As evidenced in a previous comprehensive report, the volume of the in vitro release medium may significantly affect drug release from CLs [Tieppo et al., 2014]. Since the CL is not inert against the drug, i.e., there exists an affinity, the volume of the release medium determines the concentration of drug outside the hydrogel, which affects to the release-rebinding equilibrium. When the loaded hydrogel is exposed to the release medium, the release is trigged by the drug concentration gradient between inside and outside the hydrogel. As more drug molecules accumulate in the release medium, hydrogels with high affinity for the drug, i.e., high KN/W, may stop the release [Ribeiro et al., 2011; Blanco-Fernández et al., 2011]. The drug binding interactions counteract the release, and an equilibrium is observed when the number of drug molecules that abandon the network is similar to the number that is rebound by the hydrogel. Only under infinite dilution rebinding does not occur. Therefore, in any finite volume the equilibrium may be reached, and false plateaus or delays in the release may be observed. Nevertheless, this affinity-driven controlled release phenomenon gains more relevance as the volume of the release medium decreases and the KN/W increases. The high solubility of pravastatin and intermediate values of KN/W indicated that pravastatin affinity for water 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 269 is still quite high, which may explain the small effect of the volume on the percentage of pravastatin release. Also, drug release from CLs could be affected by the presence of proteins, lipids, and mucin present in the tear film [Mahomed et al., 2016]. Interestingly, in this study, the incorporation of lysozyme and BSA in 2 mL volume slightly increased pravastatin release at all time points, while for 10 mL volume a change in the release profile, first slower and then faster, was observed (Figure 4.13). Lysozyme (pI 11.0) and BSA (pI 4.7) are positively and negatively charged, respectively, in SLF. Therefore, the likelihood of that their effect is related to ionic interactions with the drug or the hydrogel is low. In previous studies with a hydrophobic drug model (Rose Bengal dye), these same proteins were shown to accelerate the release, which could be related to hydrophobic interactions between the drug and the protein, promoting the extraction from the hydrogel [Phan et al., 2021]. Pravastatin sodium is a hydrophilic drug and the fraction unbound in plasma has been reported to be 0.485, which means that pravastatin is not a highly bound protein drug [Mao et al., 2018]. Also, the experimental protocol applied to denature the protein before drug quantification in the HPLC was demonstrated to recover >95% drug from the samples, which ensured that the decrease in the release was not due to losses of drug due to protein binding. Slower release rate in artificial tear solution containing proteins and lipids was observed for moxifloxacin from a variety of CLs, compared to simple PBS [Phan et al., 2016]. As previously reported for other HEMA hydrogels [Moradi et al., 2004], deposition of albumin and lysozyme on the tested CLs was very low and not measurable. Thus, although the formation of a thin layer of proteins on the CL surface that may act as an additional barrier for drug diffusion could not be ANA FILIPA PEREIRA DA MOTA 270 discarded, its effect might be more relevant in the first hours after immersion of the discs in 10 mL medium (more total amount of proteins) due to the Voomer effect [Vilaseca et al., 2013]. This might explain the initial slower release and also the subsequent faster release as the adsorbed proteins detach and may drag some drug molecules with them. 4.3.6 Ex vivo Corneal and Scleral Permeability Pravastatin diffusion tests from E200A40 and E0A40 hydrogels through cornea and sclera were evaluated ex vivo using porcine tissues. The vertical Franz cell may simulate the drug flux from the exterior of the eye to the aqueous or vitreous humour. As a control, the donor chamber was filled with a pravastatin solution (128 μg/mL) that provided an amount of pravastatin similar to the maximum amount of drug released from the hydrogels in 6 h. After 6 h of experiment the concentration of pravastatin in the donor chamber in contact with cornea and sclera was significantly lower for both hydrogels compared to the drug solution (Figure 4.14A; p < 0.05). No differences were found between the hydrogels. The amount of pravastatin accumulated in cornea was significantly lower than in sclera (Figure 4.14B; p < 0.05), but for a given tissue no differences were found among the hydrogels and the solution. This means that pravastatin released from the hydrogels during 6 h could diffuse into the tissues in a similar amount than when supplied as a solution, in spite of that in the case of the solution the entire dose was placed into contact with the cornea or sclera tissue since the very beginning of the experiment. 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 271 Pravastatin permeated through sclera was detected in the receptor chamber after two hours of contact with pravastatin solution and after three hours for E200A40 and E0A40 hydrogels (Figure 4.14C). For porcine cornea tissue, the amount of pravastatin in the receptor chamber was below the HPLC quantification limit. This means that pravastatin mainly accumulated into the cornea in the time frame of the experiment. Generally, a poor penetration across the cornea was observed for more hydrophilic and larger molecules [Subrizi et al., 2019]. Pravastatin presents greater hydrophilicity than other statins (log P = − 0.23) which could explain the higher sclera permeability observed. Very recently, ex vivo cornea and sclera permeability studies revealed that atorvastatin mainly accumulated into cornea and sclera tissues but did not progress further [Pereira-da-Mota et al., 2021]. Due to its lipophilicity (log P = 5.39), atorvastatin could present higher cornea permeability, but this effect may have been thwarted by the higher molecular weight (558.64 g/mol). In the case of sclera, flux (J) and apparent permeability coefficient (Papp) of pravastatin ranked in the order: E200A40 < E0A40 < pravastatin solution (Table 4.3). These values agreed with those previously reported for other drugs such as transferulic acid (0.035 to 0.115 μg/ cm2⋅h; 5.4 to 11.5⋅10 cm/s) and valaciclovir (0.42 to 4.23 μg/cm2⋅h; 9.75 to 11.37⋅10 cm/s) when delivered from drug-eluting CLs [Varela-Garcia et al., 2020; Varela-Garcia et al., 2018]. ANA FILIPA PEREIRA DA MOTA 272 Figure 4.14. (A) Amounts of pravastatin in the donor chamber, (B) amounts accumulated in cornea and sclera tissues, and (C) amounts permeated through porcine sclera after 6 h of contact with drug-loaded hydrogels (E200A40 and E0A40) and pravastatin solution (128 μg/mL, 1 mL). Codes as in Table 4.1. (n = 3; mean values and standard deviations). 1 2 3 4 5 6 Amount of pravastatin permeated through sclera (µg/cm2) 0 2 4 6 8 10 12 14 E200A40 E0A40 Prav. sol Amount of pravastatin accumulated after 6 h (µg/cm2) 0 2 4 6 8 10 12 14 Cornea Sclera E200A40 E0A40 Prav. sol. Amount of pravastatin in the donnor chamber after 6 h (µg/mL) 0 20 40 60 80 100 120 140 Cornea Sclera A B Time (hours) E200A40 E0A40 Prav. sol. C 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 273 Table 4.3. Flux (J) and apparent permeability coefficient (Papp) of pravastatin applied as a solution or as pravastatin-loaded E200A40 and E0A40 hydrogels onto to sclera tissue (n = 3; mean values and standard deviations). Hydrogel Code J (μg/(cm 2 ⋅ h)) P app (×10 6 ) (cm/s) E200A40 0.47 ± 0.11 3.16 ± 1.85 E0A40 1.02 ± 0.05 4.92 ± 2.93 Prav. Sol. 2.46 ± 0.09 6.57 ± 2.72 4.3.7 Anti-inflammatory Activity Once pravastatin was demonstrated to be able to penetrate in the eye structures ex vivo, the next sept was to elucidate the antiinflammatory activity of drug itself and of the drug-CL combination product. Cytokines and other mediators, such as prostaglandins, play important roles in eye inflammation. Some previous studies have been focused on the anti-inflammatory effects of statins, but results were not homogeneous [Bessler et al., 2005; Loppnow et al., 2011; McFarland et al., 2017]. McFarland et al. [McFarland et al., 2017] compared the effects of the six statins (including pravastatin) on PMA differentiated THP-1 cells and using LPS to induce inflammatory conditions. All statins significantly reduced LPS induced IL-1β and TNF-α release and also decreased prostaglandin E2 (PGE2) levels. Loppnow et al. [Loppnow et al., 2011] showed that pravastatin reduced the IL-6 production by 60% in human vascular smooth muscle cells (SMC) and human mononuclear cells (MNC). Conversely, Bessler et al. [Bessler et al., 2005] found that the IL-6 production in human peripheral blood mononuclear cells was not affected by pravastatin. In the present study, the secretion of tumor necrosis factor alpha (TNF-α) and interleukin-6 (IL-6), two pro-inflammatory cytokines, was examined by stimulation of macrophages with LPS. The macrophages ANA FILIPA PEREIRA DA MOTA 274 were previously treated with non-loaded and pravastatin-loaded E200A40 hydrogels or pravastatin solutions (Figure 4.15). A statistically significant reduction in TNF-α secretion (p < 0.05) was observed for hydrogels (Figure 4.15B). Overall, the obtained results point out the effectiveness of pravastatin-loaded hydrogels to decrease the secretion of TNF-α showing promising anti-inflammatory activity. 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 275 Figure 4.15. Effect of non-loaded and pravastatin-loaded E200A40 discs and pravastatin sodium solutions (0.1, 1, 10 μM) on the secretion levels of (A) TNF-α and (B) IL-6 from macrophages that were previous treated with the hydrogels or pravastatin solutions followed by addition of LPS. Positive control refers to cells only stimulated with LPS; negative control refers to unstimulated cells (without LPS); (n = 4; mean values and standard deviations). Control + Control - Non-loaded Loaded 0.1 µM 1 µM 10 µM TNF-α secreted (pg) 0 100 200 300 400 500 Control + Control - Non-loaded Loaded 0.1 µM 1 µM 10 µM IL-6 secreted (pg) 0 50 100 150 200 A B ANA FILIPA PEREIRA DA MOTA 276 4.3.8 In vivo Studies After in vitro, in ovo, in cell and ex vivo tests were fulfilled, an in vivo study was planned with the double aim of (i) assess the advantages of using pravastatin-eluting CLs compared to eye drops in terms of drug levels in tear fluid and biodistribution to eye tissues (for a same drug dose), and (ii) elucidate whether in vitro-in vivo correlations (IVIVC) could be obtained and which in vitro release conditions may provide better correlations. New Zealand rabbit eyes have been extensively used in ophthalmology research due to their similarity with human eyes [Vézina, 2012; Zernii et al., 2016]. The in vivo release profiles of pravastatin in tear fluid (concentration versus time) after application of E200A40 CLs and eye drops (2 mg/mL of pravastatin) are shown in Figure 4.16. In the eye drops group, the maximum pravastatin concentration (Cmax) in the rabbit tear fluid was recorded in the first 5 min (975.21 ± 456.40 μg/mL), followed by an exponential decrease, with a low drug concentration being detected after 1 h (14.88 ± 10.99 μg/mL to 9.30 ± 13.13 μg/mL at 8 h), which was found to be consistent with previous studies that demonstrated a rapid decline [van Bijsterveld, 1969; Xu et al., 2019; Maulvi et al., 2016]. Pravastatin-loaded CLs led to a smoother concentration curve and a sustained release over a period of 10 h, which demonstrated much less variability in the concentration of pravastatin in tear fluid and a significant improvement in mean residence time of pravastatin (Table 4.4). The Cmax was found to be 177.5 ± 116.8 μg/mL after 30 min of CL wearing. No burst release was observed. Only in the first 5 min, pravastatin concentration in tear fluid was higher for the eye drop group than for the CL group (p < 0.05). No statistically significant differences were recorded at 15 and 30 min. After that, the concentrations recorded 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 277 in the CL group were higher than those provided by the eye drops in the first 7 h of wearing (with the exception of time point 6 h, which were not significantly different). When wearing drug-loaded CLs, pravastatin diffused through the CL matrix and gradually entered the post-lens tear film prolonging the retention in the precorneal area. Previous studies also evidenced that drug-eluting CLs can improve drug retention in comparison with eye drops. For example, in New Zealand white rabbits, the MRT of ketotifen delivered from non-imprinted CLs wore for an entire day was found to be 3.4 h compared to 0.25 h for eye drops [Tieppo et al., 2012]. Also, CLs loaded with timolol during synthesis increased the mean residence time from 0.33 h for eye drops to 26.27 h for CLs wore for 5 days [Xu et al., 2019]. In our case, pravastatin-loaded CLs were intended to be daily disposable and, therefore, the release was aimed to be completed in the common half a day time frame of CL wearing. Indeed, the amounts of pravastatin remnant inside the CLs after 8 h and 10 h wearing were 7.5 ± 4.2 μg and 2.8 ± 1.6 μg, respectively, which indicated that >90% drug was released in the first 8 h of wearing. Thus, when the CLs were taken out, environmental impact due to waste of drug was not expected. Pravastatin-loaded CLs provided >2-fold increase in AUC0-8h and almost triplicate the MRT compared to the eye drops, for the same dose administered of 100 μg. MRT values of 2.82 h recorded for pravastatin loaded CLs were significantly larger than those recorded for timolol (~0.3 h) [Hiratani et al., 2005] and puerarin (1.3 h) [Xu et al., 2010] using HEMA-based CLs. The left eye of the rabbits served as a control, and no pravastatin was detected in tear fluid of left eye during treatment of the right eye either with CLs or eye drops. Also, no traces of pravastatin were detected in the tissues of the left eye. Therefore, systemic absorption could be considered as unlikely. ANA FILIPA PEREIRA DA MOTA 284 4.4 CONCLUSIONS CLs prepared with EGPEM and APMA as monomers that can resemble the hydrophobic and amine-based binding points of the natural receptor have been shown able to load remarkably high pravastatin amounts. Pravastatin-loaded CLs exhibited swelling capacity, light transmission and mechanical properties typical of soft CLs (thus, they may be suitable for refractive errors correction) and also sustained in vitro release for 10 h. Preliminary biocompatibility tests in cell cultures, HET-CAM and then in vivo confirmed the safety of this combination product. Regarding efficacy, pravastatin released from the CLs exhibited anti-inflammatory properties and capability to penetrate and accumulate in anterior and posterior eye segments. In vivo tests evidenced the suitability of the developed CLs to control pravastatin release on the ocular surface, notably prolonging the permanence time in the tear fluid compared to eye drops and also facilitating drug access to aqueous humour and vitreous humour. Preliminary assessment of IVIVC revealed that, despite pravastatin is not a highly bound protein drug and the CLs have low affinity for proteins, addition of BSA or lysozyme to the release medium may affect to the in vitro release profiles. Indeed, more linear Levy plots and with slopes closer to 1 were recorded for in vitro release media containing proteins. The information gathered in the present chapter may serve as a first step towards the search of how a change in the performance of the CLs to regulate drug release in vitro may be translated to a change in the in vivo drug profiles. 4.5 REFERENCES Al-Ghabeish, M.; Xu, X.; Krishnaiah, Y.S.; Rahman, Z.; Yang, Y.; Khan, M.A. (2015). Influence of drug loading and type of ointment base 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 285 on the in vitro performance of acyclovir ophthalmic ointment. Int. J. Pharm. 495(2), 783-791. Ahmad, A.; Panda, B. P.; Mujeeb, M. (2011). A validated stabilityindicating method for simultaneous analysis of mevastatin and pravastatin in fermentation broth during bioconversion by Actinomadura macra. Acta Chromatogr. 23(1), 121-131. Alvarez-Lorenzo, C.; Hiratani, H.; Gómez-Amoza, J.L.; MartínezPacheco, R.; Souto, C.; Concheiro, A. (2002). Soft contact lenses capable of sustained delivery of timolol. J. Pharm. Sci. 91(10), 21822192. Alvarez-Rivera, F.; Serro, A.P.; Silva, D.; Concheiro, A.; AlvarezLorenzo, C. (2019). Hydrogels for diabetic eyes: Naltrexone loading, release profiles and cornea penetration. Mater. Sci. Eng. C 105, 110092. Bedi, O.; Dhawan, V.; Sharma, P.L.; Kumar, P. (2016). Pleiotropic effects of statins: new therapeutic targets in drug design. Naunyn Schmiedebergs Arch. Pharmacol. 389(7), 695-712. Bessler, H.; Salman, H.; Bergman, M.; Straussberg, R.; Djaldetti M. (2005). In vitro effect of statins on cytokine production and mitogen response of human peripheral blood mononuclear cells. Clin. Immunol. 117(1), 73-77. Blanco-Fernández, B.; López-Viota, M.; Concheiro, A.; AlvarezLorenzo, C. (2011). Synergistic performance of cyclodextrin-agar hydrogels for ciprofloxacin delivery and antimicrobial effect. Carbohydr. Polym. 85(4), 765–774. ANA FILIPA PEREIRA DA MOTA 286 Brain-Isasi, S.; Requena, C.; Álvarez-Lueje, A. (2008). Stability study of pravastatin under hydrolytic conditions assessed by HPLC. J. Chil. Chem. Soc. 53(4), 1684-1688. Chanput, W.; Mes, J.J.; Wichers, H.J. (2014). THP-1 cell line: an in vitro cell model for immune modulation approach. Int. Immunopharmacol. 23(1), 37-45. Chung, Y.R.; Park, S.W.; Choi, S.Y.; Kim, S.W.; Moon, K.Y.; Kim, J.H.; Lee, K. (2017). Association of statin use and hypertriglyceridemia with diabetic macular edema in patients with type 2 diabetes and diabetic retinopathy. Cardiovasc. Diabetol. 16(4). DiPasquale, S.A.; Uricoli, B.; DiCerbo, M.C.; Brown, T.L.; Byrne, ME. (2021). Controlled release of multiple therapeutics from silicone hydrogel contact lenses for post-cataract/post-refractive surgery and uveitis treatment. Transl. Vis. Sci. Technol. 10(14), 5. DiPasquale, S. A.; Wuchte, L. D.; Mosley, R. J.; Demarest, R. M.; Voyles, M. L.; Byrne, M. E. (2022). One week sustained in vivo therapeutic release and safety of novel extended-wear silicone hydrogel contact lenses. Adv. Healthcare Mater. 11, 2101263. Dixon, P.; Ghosh, T.; Mondal, K.; Konar, A.; Chauhan, A.; Hazra, S. (2018). Controlled delivery of pirfenidone through vitamin E-loaded contact lens ameliorates corneal inflammation. Drug Deliv. Transl. Res. 8(5), 1114-1126. Esposito, K.; Nappo, F.; Marfella, R.; Giugliano, G.; Giugliano, F.; Ciotola, M.; Quagliaro, L.; Ceriello, A.; Giugliano, D. (2002). Inflammatory cytokine concentrations are acutely increased by hyperglycemia in humans: role of oxidative stress. Circulation. 106(16), 2067-2072. 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 287 Galante, R.; Paradiso, P.; Moutinho, M.G.; Fernandes, A.I.; Mata, J.L.; Matos, A.P.; Colaço, R.; Saramago, B.; Serro, A.P. (2015). About the effect of eye blinking on drug release from pHEMA-based hydrogels: an in vitro study. J. Biomater. Sci. Polym. Ed. 26(4), 235-251. Gologorsky, D.; Thanos, A.; Vavvas, D. (2012). Therapeutic interventions against inflammatory and angiogenic mediators in proliferative diabetic retinopathy. Mediators Inflamm. 2012, 629452. Gupta, A.; Gupta, V.; Thapar, S.; Bhansali, A. (2004). Lipid-lowering drug atorvastatin as an adjunct in the management of diabetic macular edema. Am. J. Ophthalmol. 137(4), 675-82. Hiratani, H.; Fujiwara, A.; Tamiya, Y.; Mizutani, Y.; Alvarez-Lorenzo, C. (2005). Ocular release of timolol from molecularly imprinted soft contact lenses. Biomaterials. 26(11), 1293-1298. Hui, A.; Willcox, M. (2016). In vivo studies evaluating the use of contact lenses for drug delivery. Optom. Vis. Sci. 93(4), 367-376. International Diabetes Federation (2021). IDF Diabetes Atlas; https://www.diabetesatlas.org/> (accessed on February 2022). Ioannidou, E.; Tseriotis, V.S.; Tziomalos, K. (2017). Role of lipidlowering agents in the management of diabetic retinopathy. World J. Diabetes. 8(1), 1-6. ISO 10993-5:2009(en). (2009). Biological evaluation of medical devices—Part 5: Tests for in vitro cytotoxicity; https://www.iso.org/obp/ui/#iso:std:iso:10993:-5:ed-3:v1:en (accessed on February 2022). Istvan, E.S.; Deisenhofer, J. (2001). Structural mechanism for statin inhibition of HMG-CoA reductase. Science. 292(5519), 1160–1164. ANA FILIPA PEREIRA DA MOTA 288 Itagaki, M.; Takaguri, A.; Kano, S.; Kaneta, S.; Ichihara, K.; Satoh, K. (2009). Possible mechanisms underlying statin-induced skeletal muscle toxicity in L6 fibroblasts and in rats. J. Pharmacol. Sci. 109(1), 94-101. Jain, K.S.; Kathiravan, M.K.; Somania, R.S.; Shishoo, C.J. (2007). The biology and chemistry of hyperlipidemia. Bioorg. Med. Chem. 15, 4674–4699. Johnson & Johnson News (2021). Johnson & Johnson Vision Receives Approval in Canada for World's First and Only Drug-Releasing Contact Lens for Vision Correction and Allergic Eye Itch: ACUVUE® Theravision™ with Ketotifen; https://www.jjvision.com/pressrelease/johnson-johnson-vision-receives-approval-canada-worldsfirst-and-only-drug-releasing (accessed February 2021). Kakisu, K.; Matsunaga, T.; Kobayakawa, S.; Sato, T.; Tochikubo, T. (2013). Development and efficacy of a drug-releasing soft contact lens. Invest. Ophthalmol. Vis. Sci. 54(4), 2551-2561. Kim, E.; Saha, M.; Ehrmann, K. (2018). Mechanical properties of contact lens materials. Eye Contact Lens. 44, S148–S156. Kim, S.W.; Bae, Y.H.; Okano, T. (1992). Hydrogels: Swelling, drug loading, and release. Pharm. Res. 9(3), 283–290. Kishore, A.S.; Surekha, P.A.; Sekhar, P.V.; Srinivas, A.; Murthy, P.B. (2008). Hen egg chorioallantoic membrane bioassay: an in vitro alternative to draize eye irritation test for pesticide screening. Int. J. Toxicol. 27(6), 449-453. Lazarowski, E.R.; Boucher, R.C.; Harden, T.K. (2003). Mechanisms of release of nucleotides and integration of their action as P2Xand P2Yreceptor activating molecules. Mol. Pharmacol. 64(4), 785-795. 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 289 Liao, J.K.; Laufs, U. (2005). Pleiotropic effects of statins. Annu. Rev. Pharmacol. Toxicol. 45, 89-118. Li, C.C.; Chauhan, A. (2006). Modeling ophthalmic drug delivery by soaked contact lenses. Ind. Eng. Chem. Res. 45, 3718–3734. Loppnow, H.; Zhang, L.; Buerke, M.; Lautenschläger, M.; Chen, L.; Frister, A.; Schlitt, A.; Luther, T.; Song, N.; Hofmann, B.; Rose-John, S.; Silber, R.E.; Müller-Werdan, U.; Werdan, K. (2011). Statins potently reduce the cytokine-mediated IL-6 release in SMC/MNC cocultures. J. Cell. Mol. Med. 15(4), 994-1004. Mahomed, A.; Wolffsohn, J.S.; Tighe, B.J. (2016). Structural design of contact lens-based drug delivery systems; in vitro and in vivo studies of ocular triggering mechanisms. Cont. Lens Anterior Eye. 39(2), 97-105. Mao, J.; Doshi, U.; Wright, M.; Hop, C.E.C.A.; Li, A.P.; Chen, Y. (2018). Prediction of the Pharmacokinetics of Pravastatin as an OATP Substrate Using Plateable Human Hepatocytes With Human Plasma Data and PBPK Modeling. CPT Pharmacometrics Syst Pharmacol. 7(4), 251-258. Marques, R.C.; Loebenberg, M.; Almukainzi, M. (2011). Simulated biological fluids with possible application in dissolution testing. Dissol. Technol. 18, 15–28. Maulvi, F.A.; Lakdawala, D.H.; Shaikh, A.A.; Desai, A.R.; Choksi, H.H.; Vaidya, R.J.; Ranch, K.M.; Koli, A.R.; Vyas, B.A.; Shah, D.O. (2016). In vitro and in vivo evaluation of novel implantation technology in hydrogel contact lenses for controlled drug delivery. J. Control. Release. 226, 47-56. ANA FILIPA PEREIRA DA MOTA 290 Maulvi, F.A.; Patil, R.J.; Desai, A.R.; Shukla, M.R.; Vaidya, R.J.; Ranch, K.M.; Vyas, B.A.; Shah, S.A.; Shah, D.O. (2019). Effect of gold nanoparticles on timolol uptake and its release kinetics from contact lenses: In vitro and in vivo evaluation. Acta Biomater. 86, 350-362. Maurice, D. (1995). The effect of the low blink rate in rabbits on topical drug penetration. J. Ocul. Pharmacol. Ther. 11(3), 297-304. McFarland, A.J.; Davey, A.K.; Anoopkumar-Dukie, S. (2017). Statins reduce lipopolysaccharide-induced cytokine and inflammatory mediator release in an in vitro model of microglial-like cells. Mediat. Inflamm. 2017, 2582745. Minami, T.; Ishida, W.; Kishimoto, T.; Nakajima, I.; Hino, S.; Arai, R.; Matsunaga, T.; Fukushima, A.; Yamagami, S. (2019). In vitro and in vivo performance of epinastine hydrochloride-releasing contact lenses. PLoS One. 14(1), e0210362. Misra, S.L.; Braatvedt, G.D.; Patel, D.V. (2016). Impact of diabetes mellitus on the ocular surface: a review. Clin. Exp. Ophthalmol. 44(4), 278-288. Moradi, O.; Modarress, H.; Noroozi, M. (2004). Experimental study of albumin and lysozyme adsorption onto acrylic acid (AA) and 2hydroxyethyl methacrylate (HEMA) surfaces. J. Colloid Interface Sci. 271(1), 16-19. Murphy, C.; Deplazes, E.; Cranfield, C.G.; Garcia, A. (2020). The role of structure and biophysical properties in the pleiotropic effects of statins. Int. J. Mol. Sci. 21(22), 8745. Musgrave, C.S.A.; Fang, F. (2019). Contact lens materials: A materials science perspective. Materials. 12, 261. 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 291 Nielsen, S.F.; Nordestgaard, B.G. (2014). Statin use before diabetes diagnosis and risk of microvascular disease: a nationwide nested matched study. Lancet Diabetes Endocrinol. 2(11), 894-900. Onal, A.; Sagirli, O. (2006). Development of a selective LC method for the determination of pravastatin sodium. Chromatogr. 64, 157–162. Ooi, K.G.; Khoo, P.; Vaclavik, V.; Watson, S.L. (2019). Statins in ophthalmology. Surv. Ophthalmol. 64(3), 401-432. Ooi, K.G.; Wakefield, D.; Billson, F.A.; Watson, S.L. (2015). Efficacy and safety of topical atorvastatin for the treatment of dry eye associated with blepharitis: A pilot study. Ophthalmic Res. 54(1), 26-33. Ozkiris, A.; Erkiliç, K.; Koç, A.; Mistik, S. (2007). Effect of atorvastatin on ocular blood flow velocities in patients with diabetic retinopathy. Br. J. Ophthalmol. 91(1), 69-73. Paliwal, S.K.; Chauhan, R.; Sharma, V.; Majumdar, D.K.; Paliwal, S. (2009). Entrapment of ketorolac tromethamine in polymeric vehicle for controlled drug delivery. Indian J. Pharm. Sci. 71(6), 687-691. Paradiso, P.; Colaço, R.; Mata, J.L.G.; Krastev, R.; Saramago, B.; Serro, A.P. (2017). Drug release from liposome coated hydrogels for soft contact lenses: the blinking and temperature effect. J. Biomed. Mater. Res. B Appl. Biomater. 105(7), 1799-1807. Peng, C.C.; Chauhan, A. (2012). Ion transport in silicone hydrogel contact lenses, J. Membr. Sci. 399-400, 95–105. Pereira-da-Mota, A.F.; Phan, C.M.; Concheiro, A.; Jones, L.; AlvarezLorenzo, C. (2022). Testing drug release from medicated contact lenses: The missing link to predict in vivo performance. J. Control. Release. 343, 672-702. ANA FILIPA PEREIRA DA MOTA 292 Pereira-da-Mota, A.F.; Vivero-Lopez, M.; Topete, A.; Serro, A.P.; Concheiro, A.; Alvarez-Lorenzo, C. (2021). Atorvastatin-eluting contact lenses: Effects of molecular imprinting and sterilization on drug loading and release. Pharmaceutics. 13(5), 606. Phan, C.M.; Bajgrowicz-Cieslak, M.; Subbaraman, L.N.; Jones, L. (2016). Release of moxifloxacin from contact lenses using an in vitro eye model: Impact of artificial tear fluid composition and mechanical rubbing. Transl. Vis. Sci. Technol. 5(6), 3. Phan, C.M.; Shukla, M.; Walther, H.; Heynen, M.; Suh, D.; Jones, L. (2021). Development of an in vitro blink model for ophthalmic drug delivery. Pharmaceutics. 13(3), 300. Pozuelo, J.; Compañ, V.; González-Méijome, J.M.; González, M.; Mollá, S. (2014). Oxygen and ionic transport in hydrogel and siliconehydrogel contact lens materials: An experimental and theoretical study. J. Membr. Sci. 452, 62-72. Ribeiro, A.; Veiga, F.; Santos, D.; Torres-Labandeira, J.J.; Concheiro, A.; Alvarez-Lorenzo, C. (2011). Bioinspired imprinted PHEMAhydrogels for ocular delivery of carbonic anhydrase inhibitor drugs. Biomacromolecules. 12(3), 701-709. Sayin, N.; Kara, N.; Pekel, G. (2015). Ocular complications of diabetes mellitus. World J. Diabetes. 6(1), 92-108. Schoenberger, S.D.; Kim, S.J.; Sheng, J.; Rezaei, K.A.; Lalezary, M.; Cherney, E. (2012). Increased prostaglandin E2 (PGE2) levels in proliferative diabetic retinopathy, and correlation with VEGF and inflammatory cytokines. Invest. Ophthalmol. Vis. Sci. 53(9), 59065911. 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations 293 Sekar, P.; Chauhan, A. (2019). Effect of vitamin-E integration on delivery of prostaglandin analogs from therapeutic lenses. J. Colloid Interface Sci. 539, 457-467. Shikamura, Y.; Yamazaki, Y.; Matsunaga, T.; Sato, T.; Ohtori, A.; Tojo, K. (2016). Hydrogel ring for topical drug delivery to the ocular posterior segment. Curr. Eye Res. 41(5), 653-661. Subrizi, A.; Del Amo, E.M.; Korzhikov-Vlakh, V.; Tennikova, T.; Ruponen, M.; Urtti, A. (2019). Design principles of ocular drug delivery systems: importance of drug payload, release rate, and material properties. Drug Discov. Today. 24(8), 1446-1457. Tieppo, A.; Boggs, A.C.; Pourjavad, P.; Byrne, M.E. (2014). Analysis of release kinetics of ocular therapeutics from drug releasing contact lenses: Best methods and practices to advance the field. Contact Lens Ant. Eye. 37(4), 305–313. Tieppo, A.; White, C.J.; Paine, A.C.; Voyles, M.L.; McBride, M.K.; Byrne, M.E. (2012). Sustained in vivo release from imprinted therapeutic contact lenses. J. Control. Release. 157(3), 391-397. Toffoletto, N.; Saramago, B.; Serro, A.P. (2020). Therapeutic ophthalmic lenses: A review. Pharmaceutics. 13(1), 36. van Bijsterveld, O.P. (1969). Diagnostic tests in the Sicca syndrome. Arch. Ophthalmol. 82(1), 10-14. Varela-Garcia, A.; Concheiro, A.; Alvarez-Lorenzo, C. (2020). Cytosine-functionalized bioinspired hydrogels for ocular delivery of antioxidant transferulic acid. Biomater. Sci. 8(4), 1171–1180. ANA FILIPA PEREIRA DA MOTA 300 this issue are under development [Minami et al., 2019; Sekar and Chauhan, 2019; Alvarez-Lorenzo et al., 2019; DiPasquale et al., 2022]. Additionally, there are no standardized methods for testing in vitro the drug release profiles from CLs. In vitro methods for mimicking the composition and dynamics of tear fluid, the frequency and pressure of blinking are still a challenge. Thus, in most reports focused on CLs, the in vitro drug release profiles are recorded in small beakers, using a variety of medium composition, volume, stirring, and replacement conditions [Hui and Wilcox, 2016; Pereira-da-Mota et al., 2022a]. As a consequence, most in vitro results are not predictive of in vivo performance [Minami et al. 2019]. This means that in vivo testing in animal models and human preclinical studies are still needed, even to evaluate early-stage drug-CL combination products, which makes the development very costly in time and resources. In vitro models that mimic the in vivo scenario and key ocular parameters are highly explored. Microfluidic devices have been designed to regulate the flow and volume where CLs are immersed, but other physiological conditions were not reproduced by these devices, including factors such as corneal and eyelid shape and format, tear film thickness, or blinking [Tieppo et al., 2012; Pimenta et al., 2016]. 3D printed in vitro eye models to evaluate the in vitro performance of CLs have recently been undertaken to overcome some challenges faced by using microfluidic devices and more appropriately simulate the effects of tear flow rate, tear volume, air exposure, and eyelid blinking frequency [Phan et al., 2019; Phan et al., 2021]. Not surprisingly, under dynamic conditions of low tear fluid flow, CLs showed slower drug release profiles compared to static release in a beaker, prolonging the release for days or weeks for some specific compounds and drugs such as a red food dye [Phan et al., 2021], polyethylene glycol, hydroxypropyl methylcellulose [Phan et al., 2018], moxifloxacin [Phan 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 301 et al., 2016a] and fluconazole [Phan et al., 2016b]. Moreover, in vitro models can provide relevant insights in the development process of drug-loaded CLs and prioritize successful materials that may go forward to in vivo testing in animal preclinical models or human clinical studies. There is still a paucity of information on how the in vitro release profiles recorded in the 3D printed eye models correlate with in vivo profiles. Comparison of the behaviour of the same drug-loaded CLs in both the in vitro model and the common rabbit eye model is, therefore, required for the validation of the information gathered in vitro. To gain an insight into the in vitro-in vivo correlations, the aim of this work was to analyze the release profiles of drug-loaded CLs recorded in a 3D printed in vitro eye blink model and compare the obtained results with the release in a small beaker and the tear levels previously obtained in vivo. For the sake of robustness, CLs loaded with drugs differing in physicochemical properties were tested; namely, CLs designed to uptake pravastatin (a hydrophilic statin, Log P = -0.23 [Murphy et al., 2020]) and resveratrol (a highly hydrophobic antioxidant, Log P = 3.09 [Yang et al., 2017]) were prepared and evaluated [Pereira-da-Mota et al., 2022b; Vivero-Lopez et al., 2021]. Pravastatin sodium and resveratrol may be useful for the treatment of a wide range of anterior and posterior ocular diseases. Prolonged oral therapy for hypercholesterolemia with statins has been shown to promote corneal healing, prevent cataract formation, reduce glaucoma severity, and reduce the appearance of hard exudates and microaneurysms in patients diagnosed with diabetic macular edema; ; topical ocular treatment has the advantage of avoiding systemic adverse reactions [Ooi et al., 2019; Gupta et al., 2004; Ozkiris et al., 2007; Nielsen and Nordestgaard, 2014]. Resveratrol is an antioxidant agent ANA FILIPA PEREIRA DA MOTA 302 that aids the management of oxidative-stress-related eye diseases and improves the healing of corneal epithelial cells [Tsai et al., 2015]. In previous studies, both drugs were incorporated in model CLs, and the in vivo performance evaluated in New Zealand white rabbits [Pereirada-Mota et al., 2022b; Vivero-Lopez et al., 2022]. Both drug loadedCLs provided significantly higher, and more prolonged drug levels in the rabbits' tear fluid compared to eye drops with the same dose, which favoured ocular biodistribution in the anterior and posterior structures of the eye, including cornea, sclera, lens, aqueous and vitreous humour, and retina. To carry out the present work, HEMA-based CLs were copolymerized with specific functional monomers that enhance drug affinity. In the case of pravastatin, HEMA was copolymerized with ethylene glycol phenyl ether methacrylate (EGPEM) and N-(3aminopropyl) methacrylamide hydrochloride (APMA) (Figure 5.1). For resveratrol, methacryloyloxyethyl phosphorylcholine (MPC) was added as an antifouling comonomer. The developed CLs, coded as AECLs and MCLs, respectively, demonstrated adequate solvent uptake, light transmission, mechanical properties, and ocular safety. In vitro release experiments were carried out with sterile CLs loaded under specific conditions for each drug, depending on their physicochemical properties. In vitro release in the 3D eye blink model was tested at two different tears fluid flow rates (5 and 10 µL/min of fluid) and a blink speed of 1 blink/10 s. The amount of drug released from the CLs was collected and quantified by high-performance liquid chromatography (HPLC). To the best of our knowledge, this is the first time that drugloaded CLs release profiles have been evaluated in an in vitro 3D eye model and in vitro-in vivo correlations (IVIVC) are attempted. 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 303 Figure 5.1. Chemical structures of the drugs and monomers. (A) Pravastatin sodium; (B) N-(3-aminopropyl) methacrylamide hydrochloride (APMA); (C) ethylene glycol phenyl ether methacrylate (EGPEM); (D) trans-resveratrol, and (E) 2methacryloyloxyethyl phosphorylcholine (MPC). 5.2 MATERIALS AND METHODS 5.2.1 Materials Pravastatin sodium was supplied by Biocon Limited (Bengaluru, Karnataka, India). Resveratrol was from ChemCruz, Santa Cruz Biotechnology Inc. (Dallas, TX, USA). 2-Hydroxyethyl methacrylate (HEMA) and di-sodium hydrogen phosphate anhydrous (NaH2PO4) were from Merck (Darmstadt, Germany). N-(3-aminopropyl) methacrylamide hydrochloride (APMA) was from PolySciences Inc. (Warrington, PA, USA). Ethylene glycol dimethacrylate (EGDMA), ethylene glycol phenyl ether methacrylate (EGPEM), 2,2′ - ANA FILIPA PEREIRA DA MOTA 304 azobis(isobutyronitrile) (AIBN), 2-methacryloyloxyethyl phosphorylcholine (MPC), polyvinyl alcohol (PVA, 89-98 kDa, 99% hydrolyzed), and dimethyl sulfoxide (DMSO) were from SigmaAldrich (Steinheim, Germany). Sodium chloride (NaCl) was from Labkem (Barcelona, Spain), and sodium hydroxide (NaOH) was from VWR Chemicals (Leuven, Belgium). 3D printing UV-sensitive resin was from Anycubic Technology Co. (Shenzhen, Guangdong, China). Methanol 99.9% for LC-MS grade was from Fisher Scientific (Loughborough, UK). Simulated lachrymal fluid (SLF) was prepared as previously reported [Pereira-da-Mota et al., 2022b]. Ultrapure water (resistivity >18.2 MΩ cm; Milli-Q®, Millipore Ibérica, Madrid, Spain) was obtained by reverse osmosis. 5.2.2 Contact Lens Preparation Two different types of HEMA-based CLs were prepared, as previously described [Pereira-da-Mota et al., 2022b; Vivero-Lopez et al., 2022]. Briefly, AECLs for pravastatin were prepared by mixing HEMA (3 mL) with APMA (21.45 mg), EGPEM (112.50 µL) and EGDMA (12.10 µL). The monomer solutions were magnetically stirred (200 rpm, at room temperature) for 120 min, and the initiator (AIBN; 14.79 mg) was then added and solubilized by magnetic stirring for a further 30 min. To prepare MCLs for resveratrol, HEMA (3 mL) was mixed with MPC (337.5 mg) and EGDMA (12.10 µL) under magnetic stirring (150 rpm, at room temperature) for 60 min. The mixture was kept under magnetic stirring for 30 min more to ensure the complete dissolution of the AIBN (32.85 mg). 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 305 Both AECLs and MCLs were synthesized by adding 60 µL of monomers solution into curved polypropylene moulds typically used for daily disposable CLs preparation. The moulds were kept at 50 °C for 12 h and then at 70 °C for 24 h to complete thermal polymerization. Then, the moulds were immersed in MilliQ® water to facilitate CL separation. The obtained CLs were washed under magnetic stirring (200 rpm) in 1 L of MilliQ® water and NaCl 0.9% until complete removal of the unreacted monomers occurred; the solvent was replaced at least 3 times per day. The absence of unreacted monomers was monitored by UV-Vis spectrophotometry (Agilent 8453, Waldbronn, Germany). The final dimensions of hydrated CLs (immersed in phosphate buffer pH 7.4) were approx. 12 mm diameter, 7.8 mm curvature, and 0.1 mm thickness for AECLs and approx. 14 mm diameter, 8.8 mm curvature, and 0.1 mm thickness for MCLs. 5.2.3 Drug Loading 5.2.3.1 Pravastatin Sodium Dried AECLs (average mass 16.91 ± 1.28 mg) were packaged and sealed in polyamide/polyethylene vacuum bags filled with 10 mL of an aqueous pravastatin solution (0.1 mg/mL) for at least 48 h and sterilized by high hydrostatic pressure (HHP, 70 °C and 600 MPa for 10 min) [Pereira-da-Mota et al., 2021]. The CLs were stored in sealed bags at room temperature and protected from the light until release experiments were performed. All experiments were carried out in quadruplicate. The amount of pravastatin loaded was quantified by HPLC, as explained in Section 5.2.6. ANA FILIPA PEREIRA DA MOTA 306 5.2.3.2 Resveratrol Sterile MCLs (average mass 16.8 ± 1.86 mg, sterilized by steam heat at 121 °C, 20 min) were placed in tubes containing 7 mL of a resveratrol solution (0.1 mg/mL in ethanol:water 10:90 v/v) previously filtered (Filter-Lab® Polyethersulphone (PES) syringe filter 0.22 μm; Barcelona, Spain). The loading solution was added to the tubes under sterile conditions in a biological safety cabinet. The tubes were maintained protected from light to avoid resveratrol degradation at 37 °C, 180 rpm for 72 h, after which the release tests were performed (n = 4). The amount of resveratrol loaded was quantified by HPLC, as described in Section 5.2.6. 5.2.4 Eye Blink Model The fabrication and assembly of the 3D eye model were similar to previous publications by some authors of this paper [Phan et al., 2019; Phan et al., 2021], with some minor changes (Figure 5.2). 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 307 Figure 5.2. In vitro eye blink model (Ocublink) setup. The eyelid movement spreads the tear solution, which is supplied through the tubing that is attached to the eyelid support, over the eyeball, and the contact lens (fitted on the eyeball, shown in the close-up image). The out-flow solution is collected in the collection unit located below the eyeball. 5.2.4.1 Eyeball and Collection Unit The eyeball, lower eyelid and collection unit were fabricated using a combination of 3D printing and moulding techniques [Phan et al., 2016c]. The components were printed using a hydrophobic UVpolymerizable resin on an SLA (stereolithography) 3D printer (Photon S, Anycubic, Shenzhen, China) and an FDM (fused deposition modelling) 3D printer (Prusa i3 MK3S+, Prusa Prague, Czech Republic) to ensure water-sealed parts. All printing parameters were set to the manufacturer’s default settings. The eyeball was composed of two curvatures - 11.25 mm for the larger globe and 8.6 mm for the smaller globe containing the cornea. The 8.6 mm was chosen to match the most common base curve of CLs. Compared to previous models ANA FILIPA PEREIRA DA MOTA 308 [Phan et al., 2021], the model used in the current study did not have any coatings for the front corneal surface but instead had a 300 µm groove at the centre to allow for a CL to be mounted. In preliminary tests, it was found that the resin materials for the eyeball did not absorb pravastatin and resveratrol. The collection unit of the model was designed to allow the tear film to flow from the eyeball into the wells via gravity. 5.2.4.2 Eyelid The eyelids were designed to have a curvature of 8.8 mm, which leaves a gap of approximately 200 µm between the eyeball and eyelid. Once a contact lens was applied on the eyeball, this gap was reduced by the thickness of the lens (75-150 µm). The eyelid was designed to rotate around the smaller globe, and flexes over the larger globe. PVA eyelids were prepared by dissolving PVA (89-98 kDa, 20% w/v) in dimethyl sulfoxide: ultrapure water 80:20 v/v mixtures, following a previously described protocol [Phan et al., 2021]. Briefly, the mixture was gently stirred for 5 min and heated at 120 °C for 2 h. After heating, the mixture was stirred again to ensure proper mixing of PVA. The obtained viscous solution was cast in 3D printed moulds (allowing the preparation of 4 eyelids in the same mould) and then frozen at -30 °C for 12 h. The resulting gels were thawed at room temperature for 1 h, removed from the moulds and immersed in ultrapure water for 3 days, replacing the medium daily to remove the dimethyl sulfoxide used to prepare the PVA solution. After the washing process, the eyelids were immersed in ultrapure water to maintain the hydration of the eyelid until being used in the release experiments. 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 309 5.2.4.3 Flow Rate and Blinking A commercial syringe pump (PHD ULTRA, Harvard Apparatus, Holliston, MA) was used to simulate the dynamic tear flow in the eye blink model. Simulated lachrymal fluid (SLF) or NaCl 0.9% was delivered through a hole at the top of the eyelid and spread over the eyeball surface/CL through blinking. In this study, the blink speed was set to 1 blink/10 s, and the flow rate was adjusted to 5 and 10 µL/min to ensure enough volume for sample collection at the predetermined time points. The blink velocity used in this model was 50 rpm or 45 mm/s for both closing and opening speeds; the average physiological speeds for closing and opening of the eyelid has been reported to be approximately 134 ± 4 mm/s and 26 ± 2 mm/s, respectively [Kwon et al., 2013]. 5.2.4.4 Temperature and Humidity The entire system was covered with an acrylic chamber to maintain a stable humidity and temperature (20 ± 1.5 °C) during the experiment. Humidity levels were maintained close to approximately 80% using a humidifier and controlled through a hygrometer. 5.2.4.5 Release Sampling At predetermined time points (5, 15, 30 min and every hour until 10 h), the out-flow solution was pipetted from the collection unit, stored in 300 or 600 µL Eppendorf® tubes and frozen at -30 °C until HPLC analysis. The amount of resveratrol and pravastatin released from the CLs was quantified by HPLC previous dilution of the samples in ethanol:water 50:50 v/v and SLF, respectively. All the experiments were carried out in quadruplicate. ANA FILIPA PEREIRA DA MOTA 316 approximately 13.20 ± 0.90 mg of resveratrol per g of dried hydrogel, after being immersed in drug solution for 72 h. In the in vitro vial conditions, the MCLs released 54.43 ± 7.29% resveratrol in the first 10 h (Figure 5.4A). As happened with pravastatin, the amount of resveratrol released in the eye blink model was significantly lower: 0.37 ± 0.22% at 5 µL/min and 0.47 ± 0.26% at 10 µL/min (ANOVA, p < 0.001) than the amount released in the vial. The difference between the vial results and eye blink model might be related to the diffusion resistance associated with the hydrophobic nature of resveratrol [Salehi et al., 2018] since resveratrol solubility in NaCl 0.9% was quantified to be approx. 27.4 μg/mL [Vivero-Lopez et al., 2021]. In the case of the vial tests, the volume of the release medium was increased up to 12 mL to avoid medium saturation and false plateaus. The amount of resveratrol absorbed by the PVA eyelid was about 10-fold higher compared to pravastatin, showing a higher affinity between resveratrol and the PVA eyelid (Figure 5.4B). This higher affinity could also contribute to the decrease of resveratrol detected in the fluid collected from the eye blink model. A higher fluid flow rate (10 µL/min) induced a higher amount of resveratrol absorbed by the eyelid (ANOVA, p < 0.001). As a consequence of the slow release from the MCLs, the amount of resveratrol remaining in the CLs after 10 h of the release tests was approx. 40% of the drug loaded (91.29 ± 13.85 µg for 5 µL/min and 71.37 ± 8.77 µg for 10 µL/min). No statistical differences were detected between both flow rates (ANOVA, p = 0.10). 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 317 Figure 5.4. (A) Resveratrol release profiles from MCLs experimentally recorded in a vial filled with 6 mL of NaCl 0.9% over 10 h and on the eye blink model (flow rate of 5 and 10 µL/min); (B) amount of resveratrol retained in the PVA eyelid and CLs after 10 h on the eye blink model (n = 4; mean values and standard deviations; * Statistically different between the amount of resveratrol retained in the model eyelid with the flow rate of 5 and 10 µL/min, p < 0.05. 5.3.2 In vivo Release – Eye Blink Model Comparisons The release profiles from the eye blink models for both pravastatin and resveratrol (concentration versus time) obtained were also compared to the in vivo rabbit data already reported for these same CLs [Pereira-da-Mota et al., 2022b; Vivero-Lopez et al., 2022]. Briefly, six healthy male New Zealand white rabbits were selected for the in vivo release studies wearing drug-loaded CLs. Drug-loaded CLs were removed from the loading solutions, rinsed with sterile saline solution for CLs, and carefully placed on the rabbits' right eye below the nictitating membrane and without local anesthesia. Samples of the tear Time (hours) 0 2 4 6 8 10 Cumulative release of resveratrol (%) 0.0 0.5 1.0 20.0 40.0 60.0 80.0 100.0 Eye blink model 5 Eye blink model 10 In vitro Eyelid (eye blink model - 5) Eyelid (eye blink model - 10) CLs (eye blink model - 5) CLs (eye blink model - 10) Amount of resveratrol retained in the CLs and eyelid (µg) 0 10 20 30 40 70 80 90 100 110 a b * ANA FILIPA PEREIRA DA MOTA 318 fluid were collected using Schirmer test strips before and after CLs wearing (t = 5 min, 15 min, 30 min, and every hour until 8 or 10 h). The drug concentration in tear fluid was quantified by immersing the Schirmer strips in SLF or ethanol: water (50:50 v/v), and the resulting solutions were quantified by HPLC [Pereira-da-Mota et al., 2022b; Vivero-Lopez et al., 2022]. 5.3.2.1 Pravastatin Pravastatin release profiles from AECLs during the in vivo experiment and in the eye blink model for both flow rates are compared in Figure 5.5. The maximum concentration for both experiments was obtained after 30 min of CL application. Pravastatin maximum levels were 177.5 ± 116.8 μg/mL for in vivo, and 28.39 ± 3.00 μg/mL and 39.13 ± 20.48 μg/mL for the eye blink model with a flow rate of 5 μL/min and 10 μL/min, respectively. The peak of maximum concentration was followed by a smooth decrease of drug concentration in the tear fluid and in the fluid collected from the eye blink model. No burst release was observed. 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 319 Figure 5.5. (A) In vivo tear fluid levels of pravastatin were recorded during wear of pravastatin-loaded AECLs for 8 and 10 h (n = 6 for 8 h and n = 3 for 9 and 10 h), data taken from Pereira-da-Mota et al. [Pereira-da-Mota et al., 2022b]; (B) Normalized pravastatin released concentration in SLF from pravastatin-loaded AECLs over 10 h on the eye model (flow rate of 5 and 10 µL/min) (n = 4; mean values and standard deviations). Despite the similar release patterns, the amount of pravastatin released from the CLs in the eye blink model was about 5-fold lower than that recorded in the in vivo studies (ANOVA, p < 0.05). This finding might be related to several factors: (i) the drug absorption into the PVA eyelid in the model, (ii) the composition of the release medium, (iii) the temperature, and (iv) complexity of the eyeball piece. Firstly, the PVA eyelid absorbed approximately 2.42 ± 0.26 µg pravastatin when tested under 5 µL/min and 0.89 ± 0.49 µg for 10 µL/min. Secondly, the physiological tear fluid contains proteins, lipids and mucin that can promote drug release from the CLs [Mahomed et al., 2016]. In previous studies, the influence of BSA and lysozyme on pravastatin release rate from CLs was evaluated, and an increase in the amount of pravastatin released was observed with the incorporation of both proteins in the 2 mL of SLF [Pereira-da-Mota et al., 2022b]. The effect of proteins, lipids, and other components of the tear fluid on the Time (hours) 0 2 4 6 8 10 Pravastatin concentration in tears medium (µg/mL) 0 10 20 30 40 60 Eye blink model 5 Eye blink model 10 Time (hours) 0 2 4 6 8 10 Pravastatin concentration in tear fluid (µg/mL) 0 25 50 75 100 125 150 175 200 300 In vivo ba ANA FILIPA PEREIRA DA MOTA 320 release kinetics from CLs is an important aspect to take into account in further studies. Thirdly, in this work, a stable room temperature of 20 ± 1.5 °C was maintained during the release experiments in the eye model, but an increase in temperature from 20°C to 34 °C was previously shown to enhance 20% of the fractional mass released from pHEMA hydrogels after 48 h in vitro in vials [Tieppo et al., 2014; Miyazaki et al., 2001]. This phenomenon could be related to higher kinetic energy of the drug molecules when the temperature increases, leading to a faster diffusional transport [Ferreira et al., 2020]. The effect of temperature on pravastatin solubility could also contribute to a higher amount of drug released from the CLs in vivo [Jia et al., 2008]. Fourthly, another limitation that could compromise the release from CLs in the eye blink model pertains to the composition of the eyeball piece. In the present study, the eyeball piece was printed with a hydrophobic UVpolymerizable resin which does not represent corneal surface properties. 90 per cent of the cornea consists of the stroma, which contains a high percentage of water and consists of collagen fibril lamellae oriented parallel to each other, which is more comparable to a hydrophilic hydrogel. In in vivo conditions, corneal osmosis adds water to the post-lens tear film (the tear film between the back surface of the CL and the corneal epithelium), diluting the concentration of the drug in the post-lens tear film, which could increase the drug release from the drug-loaded CL. Comparing the release profiles from the different flow rates in the eye blink model, a slightly faster release was observed for a flow rate of 10 µL/min, achieving a higher maximum concentration at 30 min, followed by a decrease in the concentration of drug present in fluid collected (ANOVA, p > 0.05). 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 321 Correlations for in vivo-in vitro in the eye blink model or in vitroin vitro (in a vial versus eye blink model) were investigated using Levy Plots (Figure 5.6). Release tests in the eye blink model with a flow rate of 10 µL/min led to a correlation coefficient (r2) closer to 1 (0.993) compared to the Levy plot obtained for 5 µL/min (Figure 5.6A). The intercepts at the origin were +6.25 and +1.93; the slopes were 1.83 and 1.62 for a flow rate of 5 and 10 µL/min, respectively. Thus, a stronger correlation was observed with 10 µL/min flow rate that favoured the in vivo-in vitro correlations. Also, a higher correlation coefficient was obtained for the Levy plot comparing the in vitro tests in a vial (2 and 10 mL) and eye blink model with a flow rate of 10 µL/min (Figure 5.6B). ANA FILIPA PEREIRA DA MOTA 322 Figure 5.6. Levy plots for in vivo (A) or in vitro (B) vs. eye blink model percentage of pravastatin released. The eye blink model experiments were carried out with 5 and 10 µL/min of flow rate (eye blink model 5/Model 5 and eye blink model 10/Model 10, respectively). The in vitro tests in a vial were carried out in 2 and 10 mL of SLF. Pravastatin released - eye blink model (%) 0 20 40 60 80 100 Pravastatin released in vivo (%) 0 20 40 60 80 100 Eye blink model 5 Eye blink model 10 r 2 = 0.976 r 2 = 0.993 a Pravastatin released - eye blink model (%) 0 20 40 60 80 Pravastatin released in vitro (%) 0 20 40 60 80 Model 5 - In vitro 10 Model 10 - In vitro 10 Model 5 - In vitro 2 Model 10 - In vitro 2 r 2 = 0.976 r 2 = 0.989 r 2 = 0.961 r 2 = 0.985 b 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 323 5.3.2.2 Resveratrol Resveratrol-loaded CLs presented a sustained release in the eye blink model as occurred in vivo (Figure 5.7). However, the concentration of resveratrol detected in the fluid collected in the eye blink model was approximately 150-fold lower than in vivo (ANOVA, p < 0.001), likely associated with several factors that also conditioned the release of the hydrophilic statin in the eye blink model, plus the low solubility of resveratrol in an aqueous medium, as previously mentioned. No statistical differences were detected between the two different flow rates (ANOVA, p > 0.05). Figure 5.7. (A) In vivo tear fluid levels of resveratrol were recorded during wear of resveratrol-loaded MCLs for 8 and 10 h (n = 6 for 8 h and n = 3 for 9 and 10 h), data taken from Vivero-Lopez et al. [Vivero-Lopez et al., 2022]; (B) Normalized resveratrol released concentration in NaCl 0.9% from resveratrol-loaded MCLs over 10 h on the eye model (flow rate of 5 and 10 µL/min) (n = 4; mean values and standard deviations). For resveratrol experiments, plots of in vivo release versus in vitro release in the eye blink model presented slopes that were remarkably high: 333.96 and 245.16 for a flow rate of 5 and 10 µL/min, respectively (Figure 5.8), which supported that the percentage of resveratrol Time (hours) 0 2 4 6 8 10 Resveratrol concentration in NaCl 0.9% (µg/mL) 0.0 0.5 1.0 1.5 2.0 Eye blink model 5 Eye blink model 10 Time (hours) 0 2 4 6 8 10 Resveratrol concentration in tear fluid (µg/mL) 0 50 100 150 200 250 300 350 400 450 In vivo ba ANA FILIPA PEREIRA DA MOTA 324 released in vivo was significantly higher than that recorded in the eye blink model. In comparison, no differences were obtained for in vitro vial release and the eye blink model. This finding highlights the difficulties in developing in vitro release models that can predict the in vivo performance of CLs loaded with hydrophobic drugs, which represent about 40% of current pharmaceutical treatments [Torres-Luna et al., 2020]. 5. In vitro – in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model 325 Figure 5.8. Correlations for in vivo (A) or in vitro (B) vs. eye blink model percentage of resveratrol released. The eye blink model experiments were carried out with 5 and 10 µL/min of flow rate (eye blink model 5; eye blink model 10, respectively). Resveratrol released - eye blink model (%) 0.0 0.2 0.4 0.6 0.8 1.0 Resveratrol released in vivo (%) 0 20 40 60 80 100 Eye blink model 5 Eye blink model 10 r 2 = 0.968 r 2 = 0.990 a Resveratrol released - eye blink model (%) 0.0 0.2 0.4 0.6 0.8 1.0 Resveratrol released in vitro (%) 0 10 20 30 40 50 Eye blink model 5 Eye blink model 10 r 2 = 0.910 r 2 = 0.957 b ANA FILIPA PEREIRA DA MOTA 332 Torres-Luna, C.; Fan, X.; Domszy, R.; Hu, N.; Wang, N.S.; Yang, A. (2020). Hydrogel-based ocular drug delivery systems for hydrophobic drugs. Eur. J. Pharm. Sci. 154, 105503. Tsai, T.Y.; Chen, T.C.; Wang, I.J.; Yeh, C.Y.; Su, M.J.; Chen, R.H.; Tsai, T.H.; Hu, F.R. (2015). The effect of resveratrol on protecting corneal epithelial cells from cytotoxicity caused by moxifloxacin and benzalkonium chloride. Invest. Ophthalmol. Vis Sci. 56(3), 1575-1584. Vivero-Lopez, M.; Muras, A.; Silva, D.; Serro, A.P.; Otero, A.; Concheiro, A.; Alvarez-Lorenzo, C. (2021). Resveratrol-loaded hydrogel contact lenses with antioxidant and antibiofilm performance. Pharmaceutics. 13(4), 532. Vivero-Lopez, M.; Pereira-da-Mota, A.F.; Carracedo, G.; Huete-Toral, F.; Parga, A.; Otero, A.; Concheiro, A.; Alvarez-Lorenzo, C. (2022). Phosphorylcholine-based contact lenses for sustained release of resveratrol: design, antioxidant and antimicrobial performances, and in vivo behavior. ACS Appl. Mater. Interf.; in press (DOI: 10.1021/acsami.2c18217) . Wichterle O. (1961). U.S. Patents. 3,660,545; 3,408,429; 3,496,254; 3,499,862. Yang, S.C.; Tseng, C.H.; Wang, P.W.; Lu, P.L.; Weng, Y.H.; Yen, F.L.; Fang, J.Y. (2017). Pterostilbene, a methoxylated resveratrol derivative, efficiently eradicates planktonic, biofilm, and intracellular MRSA by topical application. Front. Microbiol. 8, 1103. Zupančič, Š.; Lavrič, Z.; Kristl, J. (2015). Stability and solubility of trans-resveratrol are strongly influenced by pH and temperature. Eur. J. Pharm. Biopharm. 93, 196-204. Chapter 6 The work described in this chapter has not been published yet. 335 6. BIOINSPIRED MELANIN CONTACT LENSES WITH ANTIBACTERIAL PROPERTIES 6.1 INTRODUCTION The natural pigment melanin is one of the most enigmatic compounds found in living organisms from different species. Its presence in microorganisms, animals and plants leads to a variety of functionalities, including pigmentation, radical scavenging, radiation protection, and thermal regulation [Simon et al., 2010]. The exact physical and chemical properties of melanin are still unclear and has captured the attention of several scientists in the last century. Generally, melanin is classified into five groups: eumelanin, pheomelanin, neuromelanin, allomelanin and pyomelanin, based on its specific origin and location [Cao et al., 2021]. Allomelanin and pyomelanin are mostly found in microorganisms (fungi and bacteria) and plants, respectively [d’Ischia et al., 2015]. In human beings, melanin is synthesized in melanocytes and exists in three main forms, eumelanin, pheomelanin and neuromelanin (the last one mainly associated with the nervous system) [Xie et al., 2019]. Eumelanin and pheomelanin present different molecular structures and color, eumelanin is a black-to-brown insoluble polymer existing in human skin and eyes and pheomelanin is a yellow-to-reddish, alkali-soluble pigment of red-haired individuals ANA FILIPA PEREIRA DA MOTA 336 with high propensity to sunburn and skin cancer [d’Ischia et al., 2009; Napolitano et al., 2013]. Natural eumelanin biosynthesis is initiated involving the tyrosinase-catalyzed oxidation of tyrosine to the common precursors Ldihydroxyphenylalanine (L-DOPA) and DOPAquinone (Figure 6.1). Then, DOPAquinone forms the 5,6-dihydroxylindole (DHI) and 5,6dihydroxyindole-2-carboxylic (DHICA) through the intermediate molecule DOPAchrome [d’Ischia et al., 2014]. DOPAchrome is decomposed into DHI and DHICA depending on the presence of tyrosinase-related protein 2 (Tyrp-2). In the presence of the protein (in vivo), the reaction induces isomerization to DHICA, whereas in vitro the reaction is spontaneously induced with decarboxylation to DHI [Xie et al., 2019]. Numerous methods have been proposed to initiate the formation of synthetic eumelanin from L-DOPA, through autoxidation, chemical oxidants and tyrosinase assisted oxidation [d’Ischia et al., 2015]. In vitro, under ambient oxygen conditions, conversion of LDOPA to DHI/DHICA monomers is thermodynamically spontaneous and kinetically facile from the dopaquinone oxidation state. In the current work, we are investigating the pathway of synthetic DOPAmelanin preparation from L-DOPA via tyrosinase-assisted oxidation under oxygen conditions. 6. Bioinspired melanin contact lenses with antibacterial properties 337 Figure 6.1. Biosynthesis of natural eumelanin. Adapted from [Hida et al., 2020] (Open access Creative Common CC BY license). In the eye, melanin is primarily present in the iris, ciliary body, choroid and retinal pigment epithelium (RPE) [Rimpelä, et al., 2017]. In general, eumelanin presents a much larger amount in pigmented ocular tissues acting as an antioxidant and protecting the eye against ANA FILIPA PEREIRA DA MOTA 338 oxidative stress caused by reactive oxygen species (ROS) through its metal ion chelation capacity [Hu et al., 2008; Bustamante et al., 1993]. Ocular melanin function is mainly focused on light absorption including near-infrared, visible light and ultraviolet (UV) radiation [Sarna, 1992]. In the anterior segment of the eye, the melanin present in iris melanocytes reduces the amount of visible light and UV radiation entering in the eye. However, even a high amount of light escapes to the posterior segment being absorbed by the RPE protecting the neural retina and minimizing light reflection and scattering [Sarna, 1992]. In addition to these functionalities, melanin has been proposed to act as a chemical binding reservoir by retaining drugs inside the pigmented cells and influencing drug biodistribution [Larsson, 1993]. Comparative studies performed in albino and pigmented animals revealed that the accumulation of a variety of compounds in melanin containing tissues causes a slow-release kinetics and increases the ocular residence time [Salazar et al., 1993]. Indeed, melanin binding has been suggested as a strategy for reaching a sustained release of therapeutic drugs to eye tissues. Diabetes mellitus is reaching pandemic proportions and is the leading cause of blindness in adults. It has been reported that diabetic patients have an increased risk of developing conjunctival bacterial infections and corneal infections (such as fungal keratitis) due to weakened corneal epithelial barrier function [Skarbez et al., 2010]. Ciprofloxacin (Figure 6.2) is a fluoroquinolone antibiotic frequently used in ophthalmology as eyedrops or ointments to treat bacterial infections/blepharitis or ulcers of the eye [Smith et al., 2001]. The recommended dosage regimen for the treatment of bacterial infections/blepharitis is: one drop every two hours while awake for two days and one drop every four hours while awake for the next five days. For the treatment of corneal ulcers, the recommended dosage regimen 6. Bioinspired melanin contact lenses with antibacterial properties 339 is: two drops in the affected eye every 15 minutes for the first 6 h, then two drops every 30 min for the remainder of the first day. Melanin has high binding affinity for a wide range of drugs from different therapeutic classes. Pelkonen et al. classified drugs as low (e.g. diclofenac and prednisolone), intermediate (e.g. atropine and betaxolol) and high binders to melanin (e.g. ciprofloxacin and propranolol) [Pelkonen et al., 2017]. Among others, ciprofloxacin demonstrated a high binding percentage to melanin (about 95%). Figure 6.2. Molecular structure of ciprofloxacin hydrochloride. Melanin-like nanoparticles, and in particular polydopamine (PDA), have been investigated for photothermal therapy due to its capability to convert absorbed light into heat [Xiao, 2021]. Fast and local increase in temperature has been shown to be a promising method to kill bacteria without risk of bacterial resistance or side effects. Near-infrared (NIR) light irradiation of materials that can absorb light energy and transfer into heat could lead to irreparable damage of bacterial cells and appears as a suitable tool for prophylaxis and treatment of medical devicerelated infections [Cabana et al., 2017]. To effectively kill bacteria through temperature increase, high temperature of 70 °C is generally ANA FILIPA PEREIRA DA MOTA 340 necessary. However, such high temperature could cause inevitable damage to nearby host cells or tissues. Combination of photothermal therapy and chemotherapy with antimicrobial agents has been demonstrated to provide synergic effects on bacteria eradication with lower temperature [Fan et al, 2019]. Polydopamine nanoparticles modified with magainin I, an antimicrobial peptide, were shown to effectively kill Escherichia coli reaching 45 °C under NIR light irradiation [Fan et al, 2019]. In this part of the Thesis, melanin was incorporated into contact lenses with the double aim of (i) evaluating an alternative procedure of incorporating drugs into contact lenses through drug-melanin affinity, and (ii) endowing contact lenses with photothermal capacity to increase temperature in localized areas. The work was focused on the incorporation of ciprofloxacin and the possible use of antimicrobial therapy combined with photothermal therapy in the treatment of bacterial infections associated with the use of contact lenses in diabetic patients. 6.2 MATERIALS AND METHODS 6.2.1 Materials Ciprofloxacin hydrochloride, L-dihydroxyphenylalanine (LDOPA), tyrosinase from mushroom, ethylene glycol dimethacrylate (EGDMA), 2,2′-azobis(isobutyronitrile) (AIBN), and Nvinylpyrrolidone (NVP) were supplied by Sigma-Aldrich (Steinheim, Germany). 2-Hydroxyethyl methacrylate (HEMA) and di-sodium hydrogen phosphate anhydrous (NaH2PO4) were from Merck (Darmstadt, Germany). 3-(Methacryloyloxy)propyl 6. Bioinspired melanin contact lenses with antibacterial properties 341 tris(trimethylsiloxy)silane (TRIS) was from Alfa Aesar (Thermo Fisher, Kandel, Germany). Sodium chloride (NaCl) was from Labkem (Barcelona, Spain). Tryptic Soy Broth (TSB-1) was from Oxoid (Hampshire, UK). Simulated lachrymal fluid (SLF) was prepared as previously reported [Pereira-da-Mota et al., 2022]. Ultrapure water (resistivity >18.2 MΩ cm; Milli-Q®, Millipore Ibérica, Madrid, Spain) was obtained by reverse osmosis. Clarity® 1-day CLs (Somofilcon A, CooperVision®, Lake Forest, CA, USA), power - 3.50 D, water content 56%, Dk/t 86; and Air Optix® Aqua (Lotrafilcon B, Alcon®, Geneve, Switzerland), power –3.50 D, water content 33%, Dk/t 138, were gently offered by a local optical store. 6.2.2 Silicone Hydrogel Discs Preparation Silicone hydrogel (SiHy) discs were prepared by mixing 0.92 mL of HEMA, 1.58 mL of TRIS, 2.50 mL of NVP and 32 µL of EGDMA. The monomers mixture was magnetically stirred (180 rpm) for 120 min at room temperature. AIBN (12.50 mg) was added, and the mixture was kept under magnetic stirring for 30 min more to ensure the complete dissolution of the initiator. The resulting solution was injected into moulds made of two glass plates (10 x 10 cm) separated by Teflon frame (0.2 mm of thickness) and pre-treated with dichlorodimethylsilane. The hydrogels were polymerized at 50 °C for 12 h and then at 70 °C for 24 h. Following the polymerization procedure, the hydrogel sheets were boiled in 1 L of distilled water to remove the unreacted monomers and facilitate cutting in 10-mm discs. The discs were washed in water and 0.9% NaCl replacing the medium twice a day, until complete removal ANA FILIPA PEREIRA DA MOTA 348 Figure 6.4. Raman results for DOPA-melanin, SiHy discs, and melanin-incorporated SiHy discs (5, 8 and 16h). Raman shift (cm -1 ) 0 1000 2000 3000 Intensity (A.U.) 0 200 400 1500 2000 DOPA-melanin SiHy discs SiHy discs 5h SiHy discs 8h SiHy discs 16h C-H C-H -Si-O3 -Si-(CH3)x 6. Bioinspired melanin contact lenses with antibacterial properties 349 Figure 6.5. Raman results for DOPA-melanin, commercially available CLs (Clarity® and Air Optix®), and melanin-incorporated CLs. 6.3.2 Light Transmittance and Photothermal Activity Due to the relevant photo-protectant properties of melanin (absorbs harmful ultraviolet and visible radiation) [Madkhali et al., 2019], the light transmittance of blank and melanin-incorporated discs and CLs was evaluated (Figure 6.6). Blank SiHy discs and CLs showed transmittance values close to 80% and 90% in the visible range (600 nm), respectively. As expected, a transmittance decrease was verified in the UV and visible region after melanin incubation for all hydrogels. Air Optix® CLs exhibited the slightest decrease at 600 nm, 86% to 72% Raman shift (cm -1 ) 0 1000 2000 3000 Intensity (A.U.) 0 200 400 600 800 1000 1200 1400 1600 DOPA-melanin Clarity Clarity 5h Air Optix Air Optix 16h C-H C-H -Si-O 3 -Si-(CH 3 ) x ANA FILIPA PEREIRA DA MOTA 350 (after 16 h of incubation) and melanin incorporation gave a significant UV protection to the lens, especially in the UV-B and UV-A regions (280-400 nm). Clarity® CL already presented a UV protection filter and a decrease in the visible range was verified, 90% to 60%. Regarding to the SiHy discs, the light transmittance decrease was more accentuated with longer incubation times. UV-visible light transmittance results were in agreement with the measured light intensity that passed through the discs and CLs when irradiated with NIR laser (Table 6.1). Figure 6.6. Light transmittance of the silicone hydrogel discs (SiHy discs) and commercially available contact lenses (Air Optix® and Clarity®) without and with DOPA-melanin incorporation (n = 3). Wavelength (nm) 200 300 400 500 600 700 Transmittance (%) 0 20 40 60 80 100 SiHy discs SiHy discs 5h SiHy discs 8h SiHy discs 16h Air Optix Air Optix 16h Clarity Clarity 5h 6. Bioinspired melanin contact lenses with antibacterial properties 351 Table 6.1. Light intensity that passed through the discs or CLs without and with melanin incorporation when irradiated with an 808 nm laser. Reference NIR laser intensity: 1284 lux (without any lens). Without melanin (lux) With melanin (lux) - 5h 8h 16h SiHy discs 1288 958.3 830.4 820.0 Air Optix ® 1250 - - 1090 Clarity ® 1267 880.1 - - Photothermal activity of melanin-incorporated discs and CLs was evaluated. The blank and melanin-incorporated materials were irradiated with a NIR laser (808 nm) for 10 s and the temperature change of the materials was monitored (Figure 6.7 and 6.8). Minimal temperature changes were verified on the blank SiHy discs and CLs. On the other hand, a notably and very fast (in 10 s) temperature increase was recorded on discs incorporating melanin, this temperature change was more accentuated for discs that were incubated in melanin solution for longer incubation times. Similar results were obtained for commercially available contact lenses, melanin-incorporated Air Optix® CLs revealed a lower temperature increase as the amount of pigment incorporated was lower than Clarity® CLs, as verified with the light transmittance values. Other authors observed a significantly improved antibacterial activity of melanin-modified hydrogels following NIR radiation. Recently, NIR-responsive chitosan/silk fibroin cryogels incorporating polydopamine have been developed and when irradiated with NIR radiation for 10 min an increase from 25 to 35 °C was verified showing a significant antibacterial efficacy [Han et al., 2019]. The combination of PDA-mediated photothermal treatments with antibiotic drugs has ANA FILIPA PEREIRA DA MOTA 352 also been explored. PDA hydrogels incorporated with doxycycline have demonstrated a photothermal-mediated antimicrobial activity and sustained antibiotic release capacity [Liang et al., 2022]. 6. Bioinspired melanin contact lenses with antibacterial properties 353 Figure 6.7. Infrared thermal camera images of temperature changes in the irradiated area of the SiHy discs after 10 s of laser radiation. ANA FILIPA PEREIRA DA MOTA 354 Figure 6.8. Infrared thermal camera images of temperature changes in the irradiated area of the commercial CLs after 10 s of laser radiation. 6.3.3 Ciprofloxacin Loading and Release SiHy discs and commercial CLs without and with melanin incorporation were immersed in a prepared ciprofloxacin solution (2 or 3 mg/mL in ultrapure water) and incubated for 48 h under mild agitation (180 rpm) and protected from the light. The amount of drug loaded was 6. Bioinspired melanin contact lenses with antibacterial properties 355 not possible to quantify due to the high absorbance of the loading solution, the difference between the initial and final absorbance was not quantifiable. Thus, it was assumed that the amount of ciprofloxacin loaded was totally released. After this time, ciprofloxacin-loaded discs and CLs were incubated in 10 mL of SLF at 37 °C for 24 h. At predetermined timepoints, the absorbance of the release medium was measured at 275 nm and the results are showed in Figures 6.9 and 6.10. For all discs and CLs, the majority of ciprofloxacin loaded was released on the first 1-2 h of release experiment. However, a slightly increase in the amount of ciprofloxacin released was verified for the melanin-incorporated discs (continuous lines, Figure 6.9). An increase of drug released was demonstrated when the discs were immersed in 3 mg/mL of ciprofloxacin, which reveal an improvement of drug loaded from the discs. The amount of drug release was higher for discs with longer melanin incubation times, evidencing the affinity of ciprofloxacin and melanin. For commercialised CLs, the scenario is different, Clarity® CLs released a significant higher amount of ciprofloxacin over 24 h compared to Air Optix® CLs. However, the amount of drug released was lower for melanin-incorporated Clarity® CLs than blank CLs. In the case of Air Optix® CLs, the lenses that were incubated in 3 mg/mL showed a higher amount of drug released. Ciprofloxacin is a bactericidal antibiotic that inhibits DNA replication by inhibiting bacterial DNA topoisomerase and DNAgyrase. The minimum inhibitory concentration (MIC90) of ciprofloxacin for Gram-positive S. aureus was reported to be 0.5 μg/mL [Liu et al., 2020]. The amount of ciprofloxacin released over 24 h from discs and CLs is above the minimum inhibitory concentration (between 5 – 11 μg/mL in 10 mL of simulated lachrymal fluid). ANA FILIPA PEREIRA DA MOTA 356 Figure 6.9. Ciprofloxacin release profiles from blank SiHy discs and melaninincorporated discs (5, 8 or 16 h) after being incubated for 48 h in a 3 or 2 mg/mL ciprofloxacin loading solution (n = 3; mean values and standard deviations). Time (hours) 0123456724 Ciprofloxacin released (mg/g of hydrogel) 0 1 2 3 4 5 6 7 SiHy discs 3mg/mL SiHy discs 2mg/mL SiHy discs 5h 3mg/mL SiHy discs 5h 2mg/mL SiHy discs 8h 3mg/mL SiHy discs 8h 2mg/mL SiHy discs 16h 3mg/mL SiHy discs 16h 2mg/mL 6. Bioinspired melanin contact lenses with antibacterial properties 357 Figure 6.10. Ciprofloxacin release profiles from blank Air Optix® and Clarity® CLs and melanin-incorporated CLs (16 or 5 h) after being incubated for 48 h in a 3 or 2 mg/mL ciprofloxacin loading solution (n = 3; mean values and standard deviations). 6.3.4 Antibacterial Performance Isothermal microcalorimetric analyses were performed as a proof of concept to gain further insight into the effects that the SiHy discs with melanin incorporation and the amount of ciprofloxacin released from the discs may have on the metabolic activity of the bacteria. This method is a highly sensitive technique that enables a real-time monitoring of bacterial viability in terms of metabolism-related heat production and detect any change in bacterial metabolic rate due to the administration of drugs or compounds [Braissant et al., 2010]. Time (hours) 0 1 2 3 4 5 6 7 24 Ciprofloxacin released (mg/g of hydrogel) 0 2 4 6 8 Air Optix 3mg/mL Air Optix 2mg/mL Air Optix 16h 3mg/mL Air Optix 16h 2mg/mL Clarity 3mg/mL Clarity 2mg/mL Clarity 5h 3mg/mL Clarity 5h 2mg/mL ANA FILIPA PEREIRA DA MOTA 364 Madkhali, N.; Alqahtani, H.; Al-Terary, S. et al. (2019). Control of optical absorption and fluorescence spectroscopies of natural melanin at different solution concentrations. Opt. Quant. Electron. 51, 227. Napolitano, A.; Panzella, L.; Leone, L.; d’Ischia, M. (2013). Red hair benzothiazines and benzothiazoles: Mutation-inspired chemistry in the quest for functionality. Acc. Chem. Res. 46(2), 519−528. Pelkonen, L.; Tengvall-Unadike, U.; Ruponen, M.; Kidron, H.; Del Amo, E.M.; Reinisalo, M.; Urtti, A. (2017). Melanin binding study of clinical drugs with cassette dosing and rapid equilibrium dialysis inserts. Eur. J. Pharm. Sci. 109, 162-168. Pereira-da-Mota, A.F.; Vivero-Lopez, M.; Serramito, M.; Diaz-Gomez, L.; Serro, A.P.; Carracedo, G.; Huete-Toral, F.; Concheiro, A.; AlvarezLorenzo, C. (2022). Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations. J. Control. Release. 348, 431-443. Saghaie, L.; Pourfarzam, M.; Fassihi, A.; Sartippour, B. (2013). Synthesis and tyrosinase inhibitory properties of some novel derivatives of kojic acid. Res. Pharm. Sci. 8(4), 233-242. Salazar, M.; Shimada, K.; Patil, P. N. Iris pigmentation and atropine mydriasis. J. Pharmacol. Exp. Ther. 1976, 197, 79−88; Nagata, A.; Mishima, H. K.; Kiuchi, Y.; Hirota, A.; Kurokawa, T.; Ishibashi, S. (1993). Binding of antiglaucomatous drugs to synthetic melanin and their hypotensive effects on pigmented and nonpigmented rabbit eyes. Jpn. J. Ophthalmol. 37, 32−38. Sarna, T. (1992). Properties and function of the ocular melanin-a photobiophysical view. J. Photochem. Photobiol. B. 12, 215–258. 6. Bioinspired melanin contact lenses with antibacterial properties 365 Skarbez, K.; Priestley, Y.; Hoepf, M.; Koevary, S.B. (2010). Comprehensive review of the effects of diabetes on ocular health. Expert Rev. Ophthalmol. 5(4), 557-577. Simon, J. D.; Peles, D. N. (2010). The red and the black. Acc. Chem. Res. 43(11), 1452−1460. Smith, A.; Pennefather, P.M.; Kaye, S.B.; Hart, C.A. (2001). Fluoroquinolones: place in ocular therapy. Drugs. 61(6), 747-761. Sono, K.; Lye, D.; Moore, C.A.; Boyd, W.C.; Gorlin, T.A.; Belitsky, J.M. (2012). Melanin-based coatings as lead-binding agents. Bioinorg. Chem. Appl. 2012, 361803. Rimpelä, A.K.; Reinisalo, M.; Hellinen, L.; Grazhdankin, E.; Kidron, H.; Urtti, A.; Del Amo, E.M. (2018). Implications of melanin binding in ocular drug delivery. Adv. Drug Deliv. Rev. 126, 23-43. Tran, N.P.; Yang, M.C. (2020). The ophthalmic performance of hydrogel contact lenses loaded with silicone nanoparticles. Polymers. 12(5), 1128. Xiao, Z. (2021). Photothermal Therapy Mediated by Nanomaterial; https://encyclopedia.pub/entry/6885 (accessed on November 2022). Xie, W.; Pakdel, E.; Liang, Y.; Kim, Y. J.; Liu, D.; Sun, L.; Wang, X. (2019). Natural eumelanin and its derivatives as multifunctional materials for bioinspired applications: A Review. Biomacromolecules. 20(12), 4312−433 CONCLUSIONS 369 7. CONCLUSIONS According to the aims of the present Thesis, the development of drug-CL combinations products was carried out, and the possibility of loading drugs useful for ophthalmic administration was evaluated. Several bioinspired approaches were explored, and the obtained combination products were characterized in detail. The work was carried out in four steps and the specific conclusions of each chapter are summarized in the following paragraphs. 1. Bioinspired contact lenses mimicking the active site of HMG– CoA reductase were successfully designed for the first time with an enhanced affinity for atorvastatin calcium. The hydrogels presented solvent uptake, light transmission, and mechanical properties values in the common range for commercially available CLs. No potential eye (CAM) irritation was observed neither cytotoxicity effects. The APMA hydrogels notably increased the amount of atorvastatin loaded and released. Hydrostatic pressure sterilization was found to be an advantageous substitute of steam heat and gamma radiation in the sterilization of atorvastatin-loaded hydrogels, maintaining atorvastatin stability and hydrogels performance. Permeability studies suggest that ANA FILIPA PEREIRA DA MOTA 370 the amount of atorvastatin accumulated in the cornea and sclera could be effective to treat ocular surface diseases. 2. Pravastatin was successfully loaded in bioinspired contact lenses mimicking the active site of HMG–CoA reductase prepared with EGPEM and APMA as monomers that can resemble the hydrophobic and amine-based binding points of the natural receptor. Pravastatinloaded CLs sustained in vitro the release for 10 h, exhibited antiinflammatory properties, and the drug released could penetrate and accumulate in anterior and posterior eye segments. The safety of this combination product was confirmed in cell studies, HET-CAM and in vivo studies. In vivo tests in a rabbit model demonstrated the suitability of the developed CLs to control pravastatin release on the ocular surface, notably prolonging the permanence time in the tear fluid compared to eye drops as well as facilitating drug access to aqueous humour and vitreous humour. Preliminary assessment of IVIVC revealed that addition of BSA or lysozyme to the release medium may affect to the in vitro release profiles demonstrating stronger IVIVC for in vitro release media containing proteins. 3. The usefulness of a 3D printed eye blink model for the evaluation of drug release profiles of medicated CLs was explored in detail in this part of the Thesis. Two drugs of different physicochemical properties, pravastatin sodium and resveratrol, were used to load the CLs in order to validate the robustness of this approach. The release profile of both drugs was more sustained and lower in the in vitro eye blink model compared to the in vitro release in vials, especially for the hydrophobic drug resveratrol. Interestingly, both drug-loaded CLs showed similar 7. Conclusions 371 release patterns in the eye blink model as in in vivo studies. However, the amount of drug released in the eye blink model was significantly lower compared to previously obtained in vivo data. Levy plots presented stronger correlations for pravastatin release in the eye model (flow rate of 10 µL/min) and in vivo data. The obtained results may serve as a guide for further improvements of the 3D printed eye blink model. 4. Melanin was successfully incorporated to SiHy discs and commercially available CLs to communicate multifunctional performances. Namely, melanin was shown to act as a light filter that can efficiently convert patient-friendly near-infrared light into moderate increases in temperature. A very fast (in 10 s) and localized increase in temperature conveyed through the contact lenses may be useful to treat a variety of surface eye diseases including tumors and infections. Moreover, melanin may endow CLs with affinity for a variety of drugs, allowing for the combination of photothermal therapy and chemotherapy. The information gathered in the present chapter may serve as a first step towards the development of strategies that combine hydrogels modification and incorporation of active substances for stimuli-responsiveness. Taken together, the bioinspired strategies developed in this PhD Thesis render drug-CL combination products for topical ocular drug delivery that provide sustained drug levels in the ocular surface and also allow for feasible photothermal therapy of surface eye diseases. Moreover, in vitro test methods that mimic drug release on the ocular ANA FILIPA PEREIRA DA MOTA 372 surface have been identified, which may pave the way towards the setup of methodologies that can provide strong in vitro-in vivo correlations. ANNEXES 380 Pharmaceutics ISSN: 1999-4923 Journal Impact Factor: 6.525 (2021) JCR Category: Pharmacology & Pharmacy Category Quartile: (Q1, Pharmacology & Pharmacy); CiteScore (Q2, Pharmaceutical Science). Reproduction permission: The article was published in Open Access. Chapter 4. Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations The work described in this chapter was published in Contact lenses for pravastatin delivery to eye segments: Design and in vitro-in vivo correlations, Journal of Controlled Release 348, 2022, 431-443, authored by Ana F. Pereira-da-Motaa, María Vivero-Lopeza, Maria Serramitob, Luis DiazGomeza, Ana Paula Serroc, Gonzalo Carracedob, Fernando Huete-Torralb, Angel Concheiroa and Carmen Alvarez-Lorenzoa 381 aDepartamento de Farmacología, Farmacia y Tecnología Farmacéutica, I+D Farma Group (GI-1645), Facultad de Farmacia, Instituto de Materiales (iMATUS) and Health Research Institute of Santiago de Compostela (IDIS), Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain. bOcupharm Research Group, Faculty of Optic and Optometry, University Complutense of Madrid, C/Arcos del Jalon 118, 28037 Madrid, Spain. cCentro de Química Estrutural, Departamento de Engenharia Química, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1049-001 Lisbon, Portugal. Author contribution: Bibliographic research, collaboration in the design of the experiments, data collection and processing, participation in the elaboration of the final text. Journal of Controlled Release ISSN: 0168-3659 Journal Impact Factor: 11.467 (2021) JCR Category: Pharmaceutical Science Category Quartile: Q1 (D1) 382 Reproduction permission: Chapter 5. In vitro-in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model The work described in this chapter was accepted for publication in In vitro-in vivo correlation of drug release profiles from medicated contact lenses using an in vitro eye blink model, Drug Delivery and Translation Research, 2022, authored by Ana F. Pereira-da-Motaa, María Vivero-Lopeza, Piyush Gargb, Chau Minh-Phanb,c, Angel Concheiroa, Lyndon Jonesb,c, and Carmen Alvarez-Lorenzoa aDepartamento de Farmacología, Farmacia y Tecnología Farmacéutica, I+D Farma Group (GI-1645), Facultad de Farmacia, Instituto de Materiales (iMATUS) and Health Research Institute of Santiago de Compostela (IDIS), Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain. bCentre for Ocular Research & Education (CORE), School of Optometry and Vision Science, University of Waterloo, Waterloo, ON, Canada. cCentre for Eye and Vision Research (CEVR), 17W, Hong Kong, Science Park, Hong Kong. 383 Author contribution: Bibliographic research, collaboration in the design of the experiments, data collection and processing, participation in the elaboration of the final text. Drug Delivery and Translation Research ISSN: 2190-3948 Journal Impact Factor: 5.671 (2021) JCR Category: Pharmaceutical Science Category Quartile: Q1 Reproduction permission: 384 CHECKLIST FOR THESIS THAT INCLUDE EXPERIMENTAL ANIMALS. EXPERIMENTAL ANIMALS ARRIVE Yes/No/NA page Title Yes Provide as accurate and concise a description of the content of the article as possible. 235 Abstract Yes Provide an accurate summary of the background, research objectives, including details of the species or strain of animal used, key methods, principal findings and conclusions of the study. 11-13 Background Yes Provide an accurate summary of the background, research objectives, including details of the species or strain of animal used, key methods, principal findings and conclusions of the study. 235-295 Yes Explain how and why the animal species and model being used can address the scientific objectives and, where appropriate, the study’s relevance to human biology. 276-280 Objectives Yes Clearly describe the primary and any secondary objectives of the study, or specific hypotheses being tested. 238-239 Methods Ethical statement Yes Indicate the nature of the ethical review permissions, relevant licenses, and national or institutional guidelines for the care and use of animals, that cover the research. 251 Study design 385 Yes Number of experimental and control groups 251-252 Yes Steps taken to minimize the effects of subjective bias when allocating animals to treatment (e.g. randomization procedure) and when assessing results (e.g. if done, describe who was blinded and when). 251-252 Yes The experimental unit (e.g. a single animal, group or cage of animals). A time-line diagram or flow chart can be useful to illustrate how complex study designs were carried out. 251-252 Experimental procedures Yes How (e.g. drug formulation and dose, site and route of administration, anesthesia and analgesia used [including monitoring], surgical procedure, method of euthanasia). Provide details of any specialist equipment used, including supplier(s). 250-255 Yes When (e.g. time of day). 251 Yes Where (e.g. home cage, laboratory, water maze). 251 Yes Why (e.g. rationale for choice of specific anesthetic, route of administration, drug dose used). 250-255 Experimental animals Yes Provide details of the animals used, including species, strain, sex, developmental stage (e.g. mean or median age plus age range) and weight (e.g. mean or median weight plus weight range). 251 Yes Provide further relevant information such as the source of animals, international strain nomenclature, genetic modification status (e.g. knock-out or transgenic), genotype, health/immune status, drug or test naïve, previous procedures, etc. 251 Housing and husbandry Yes Housing (type of facility e.g. specific pathogen free [SPF]; type of cage or housing; bedding 251 386 material; number of cage companions; tank shape and material etc. for fish). Yes Husbandry conditions (e.g. breeding program, light/dark cycle, temperature, quality of water etc for fish, type of food, access to food and water, environmental enrichment). 251 Yes Welfare-related assessments and interventions that were carried out prior to, during, or after the experiment. 250-255 Sample size Yes Specify the total number of animals used in each experiment, and the number of animals in each experimental group. 251-252 Yes Explain how the number of animals was arrived at. Provide details of any sample size calculation used. 251-252 Yes Indicate the number of independent replications of each experiment, if relevant. 251-252 Allocating animals to experimental groups Yes Indicate the number of independent replications of each experiment, if relevant. 251-252 Yes Describe the order in which the animals in the different experimental groups were treated and assessed. 251-252 Experimental outcomes Yes Clearly define the primary and secondary experimental outcomes assessed (e.g. cell death, molecular markers, behavioral changes). 253-255 Statistical methods Yes Provide details of the statistical methods used for each analysis. 255 Yes Specify the unit of analysis for each dataset (e.g. single animal, group of animals, single neuron). 255, 278, 280 387 Yes Describe any methods used to assess whether the data met the assumptions of the statistical approach. 255, 278, 280 Results and discussion Basal data Yes For each experimental group, report relevant characteristics and health status of animals (e.g. weight, microbiological status, and drug or test naïve) prior to treatment or testing (this information can often be tabulated). 251-252 Numbers analyzed Yes Report the number of animals in each group included in each analysis. Report absolute numbers (e.g. 10/20, not 50%). 251-252 Yes If any animals or data were not included in the analysis, explain why. 251-252 Outcomes and estimation Yes Report the results for each analysis carried out, with a measure of precision (e.g. standard error or confidence interval). 276-280 Adverse events NA Give details of all important adverse events in each experimental group. Yes Describe any modifications to the experimental protocols made to reduce adverse events. 251-253 Interpretation/scientific implications Yes Interpret the results, taking into account the study objectives and hypotheses, current theory and other relevant studies in the literature. 276-280 Yes Comment on the study limitations including any potential sources of bias, any limitations of the animal model, and the imprecision associated with the results. 251 NA Describe any implications of your experimental methods or findings for the replacement, 388 refinement or reduction (the 3Rs) of the use of animals in research. Generalizability/translation Yes Comment on whether, and how, the findings of this study are likely to translate to other species or systems, including any relevance to human biology. 276-280 Funding Yes List all funding sources (including grant number) and the role of the funder(s) in the study. 251 Based on The ARRIVE guidelines: Animal Research: Reporting of In Vivo Experiments. PhD Student signature COMITÉ DE EXPERIMENTACIÓN ANIMAL Solicitud de Evaluación y Memoria de Contenidos de Proyecto (RD 53/2013, Experimentación Animal) F01-PROYECTOS Servicio de Investigación GEISER: Unidad Tramitadora U01000294 OH-CEA-UCM-F01 Versión 01-15102020 VALORACIÓN DEL USO DE ANIMALES DE EXPERIMENTACIÓN EN PROYECTOS DE INVESTIGACIÓN Pág. 1 de 18 Investigador Principal Proyecto Nombre: Juan Gonzalo Carracedo Rodríguez DNI: 07505953H Centro: Facultad de Óptica y Optometría Departamento/ Sección: Departamento de Optometría y Visión Teléfono: 616513539 Correo electrónico: [email protected] Investigador Responsable Procedimientos Nombre: Juan Gonzalo Carracedo Rodríguez (IP) DNI: Centro: Departamento/ Sección: Teléfono: Correo electrónico: Título del Proyecto Título: Efecto de la luz azul y las lentes de contacto de ortoqueratología para el control de la progresión de la miopía sobre la morfología, la fisiología y la bioquímica ocular Fecha de inicio: 01/01/2022 Fecha de finalización: 03/03/2022 Finalidad del Informe Científica Docente Imposición legal Presentación de proyecto para ser financiado Organismo: Ministerio de Ciencia e Innovación Convocatoria: 2021 Proyecto financiado en ejecución Organismo: Referencia: Autorización de experimentación o actividad Otros (especificar): Fecha y Firma 06/05/2021 Vº.Bº. Vicerrectora de Investigación y Transferencia y Responsable administrativo de usuario Investigador/a Principal Dña. Margarita San Andrés Moya D./Dña. Juan Gonzalo Carracedo Rodríguez (Al firmar declaro haber leído y aceptado la cláusula de protección de datos al pie indicada) Información básica de protección de datos del tratamiento: Investigación Responsable Vicerrectorado de Investigación y Transferencia Finalidad Ayudas y acciones para desarrollo de la investigación Científica Legitimación Cumplimiento de una obligación legal; Misión en interés público Destinatarios Se prevén cesiones Derechos Acceder y rectificar los datos, así como otros derechos, explicados en la información adicional Infor. adicional Puede consultarla con detalle en: https://www.ucm.es/data/cont/docs/3-2018-05-23-Info-Adic-Tratamiento-Investigación.pdf GEISER-5740-83a2-42e5-469d-a558-3855-bbfc-93ed