Also available as PDF (figures and tables appear in the PDF).
Ophthalmic drug development remains one of the most challenging areas in pharmaceutical research due to the complex anatomy and protective barriers of the eye. Conventional ocular dosage forms, particularly topical formulations, often exhibit poor bioavailability because of rapid tear turnover, blinking, nasolacrimal drainage, and limited corneal permeability. These limitations reduce therapeutic efficiency and necessitate frequent dosing, ultimately affecting patient compliance. In recent years, significant advances have been achieved in the development of innovative ocular drug delivery systems, including ocular implants, hydro- gels, nanoparticles, nanomicelles, liposomes, microneedles, iontophoresis, in situ gels, and drug-eluting contact lenses. These technologies aim to improve drug retention, enhance tissue penetration, and provide sustained therapeutic effects for both anterior and posterior segment eye diseases. Simultaneously, artificial intelligence (AI) and digital technologies are transforming ophthalmic drug discovery and development by accelerating target identification, molecular screening, biomarker analysis, and formulation optimiza- tion. AI-driven approaches such as machine learning, deep learning, explainable AI, and large language models have shown considerable potential in reducing development timelines and improving predictive accuracy. In addition, additive manufacturing technologies, including three-dimensional printing and hot- melt extrusion, have enabled the development of personalized ophthalmic devices and customized drug delivery systems. Despite these advances, several challenges remain, including manufacturing complexity, regulatory uncertainties, scalability issues, data privacy concerns, and the absence of harmonized global guidelines for AI-assisted therapeutics. This review highlights recent developments in advanced ocular drug delivery systems, emerging AI-based technologies, regulatory perspectives, and future opportunities aimed at improving the treatment of complex ocular diseases and reducing the global burden of visual impairment.
Keywords: Ophthalmic drug delivery; Artificial intelligence; Ocular implants; Nanotechnology; 3D Printing; Ophthalmicdrugdelivery; Artificialintelligence; Ocularimplants; 3DPrinting ever; itsapplicationislimitedtothisspecificgenetic
Drugdiscoveryanddevelopmentiswidelyrecognized as one of the most complex and resource-intensive processesinmodernbiomedicalscience.Thejourney fromidentifyingatherapeutictargettoobtainingregulatoryapprovaloftenspansdecadesandrequiresextensivecollaborationbetweenacademicinstitutions, pharmaceutical companies, regulatory bodies, and public–privatepartnerships.Financially,theprocessis extremelydemanding,withestimatessuggestingthat bringing a single new drug to market may cost anywherefromapproximately$314millionto$4.4billion, dependingonthetherapeuticarea,developmentpathway,andsuccessrateacrossclinicalphases.Thefield ofophthalmologyprovidesaclearexampleofthese challenges.Theglobalmarketforophthalmicdrugsis expandingrapidlyandisprojectedtoreachnearly$66 billionby2030,largelydrivenbythegrowingburdenof vision-threateningdiseasessuchasage-relatedmaculardegeneration(AMD)anddiabeticretinopathy.Despitethiscommercialgrowth,translatingscientificdiscoveriesintoeffectiveophthalmictherapiesremains slowandtechnicallychallenging.[1]Anotableillustrationisthecaseofbevacizumab,amonoclonalantibody initiallydevelopedforcancertreatment.Thedrugrequiredmorethantwodecadesofresearchandbillions ofdollarsindevelopmentbeforecliniciansobserved, almostserendipitously,thatitcouldalsobebeneficial intreatingAMDwhenadministeredintheeye. This unexpectedobservationultimatelyledtothedevelopment of a new class of therapies targeting vascularendothelialgrowthfactor(VEGF).Anti-VEGFdrugs havesincetransformedthemanagementofretinaldiseases,includingAMD,diabeticretinopathy,andretinal veinocclusion,andhavecontributedsignificantlyto reducingblindnessratesinmanypartsoftheworld. Nevertheless,thejourneyfromidentifyingVEGFasa therapeutictargettosecuringU.S.FoodandDrugAdministration(FDA)approvalforophthalmicanti-VEGF treatmentsrequiredmorethanthreedecadesofresearch and clinical testing. Progress in ophthalmic drugdevelopmentremainsrelativelylimited. Since 2010,only17newophthalmicdrugshavereceivedFDA approval,mostofwhichfocusonasmallgroupofconditionssuchasAMD,open-angleglaucoma,anddry eyedisease. Manyseriousoculardisordersstilllack effectivetherapeuticoptions.Forexample,geographic atrophy(GA),alatestageofdryAMD,currentlyhasno approvedtreatmentcapableofrestoringvisualacuity orreversingretinaldamage.Genetherapyhasshown promiseincertaininheritedretinaldisorders; forinstance,thetherapyLuxturnacanimprovenightvision inindividualswithmutationsintheRPE65gene.Howsubgroupandthereforebenefitsonlyasmallfraction ofpatients.[2]Recognizingtheneedforinnovation,recentresearchinitiativeshaveattemptedtoidentifythe mostcriticalprioritiesinophthalmicscience.Anotable exampleistheUKClinicalEyeResearchStrategy,publishedin2024,whichhighlightedtwelvekeyresearch themesrequiringurgentattention. Thesethemesincludepreventionofcataracts, improvedtreatments formicrobialkeratitis,developmentofmoreeffective communityoptometriccarepathways, strategiesto slowtheprogressionofrefractiveerrors,earlydetectionofchildhoodvisualdisorders,andimprovedtherapeuticapproachesforglaucoma,neurodegenerative eyediseases,dryAMD,andorbitalinflammatorydisorders.[3]Inrecentyears,artificialintelligence(AI)has emergedasapowerfultoolcapableoftransforming manyaspectsofthedrugdiscoverypipeline.AIencompassesabroadcollectionofcomputationalmethods, including machine learning (ML) and deep learning (DL)algorithms,thatcanprocessextremelylargeand complexdatasets.Thesemodelsarecapableofrecognizinghiddenpatternswithinbiologicaldata,generatingpredictivemodels,andguidingdecision-making in ways that were previously impossible with traditionalexperimentalmethods.Inthecontextofpharmaceuticalresearch,AIcanassistscientistsindesigning new molecules, predicting drug–target interactions, estimatingpharmacokineticandtoxicityprofiles,and optimizingclinicaltrialdesign. SeveralrecentexamplesdemonstratethetransformativepotentialofAIassisteddrugdiscovery.Onewell-knowncaseisDSP- 1181,adrugcandidatedevelopedforthetreatment ofobsessive-compulsivedisorder.Thecompoundwas designedthroughacollaborationbetweenSumitomo DainipponPharmaandtheAI-drivenbiotechnology companyExscientia.Usingadvancedmachinelearning algorithms,researcherswereabletooptimizethecompound’schemicalstructure,pharmacologicalproperties, and safety characteristics in just 12 months, a remarkablyshorttimeframecomparedwiththetypicalfive-yearperiodoftenrequiredforearlydrugdiscoverystages.[4]Anothergroundbreakingexampleis Halicin,anantibioticidentifiedusinganAImodeldevelopedattheMassachusettsInstituteofTechnology (MIT).Byscreeningavirtuallibrarycontainingmore than100millionchemicalcompounds,theAIsystem identifiedamoleculewithacompletelynovelstructurethatdemonstratedpotentactivityagainstseveral antibiotic-resistant bacterial strains. This discovery highlightedhowAI-basedapproachescandramatically accelerate the identification of promising therapeutic molecules compared with traditional laboratory screeningtechniques.Growingevidencesuggeststhat integratingAIintoearlydrugdiscoverycouldreduce developmenttimelinesbymorethan60%,whilealso improvingtheprobabilityofsuccessinclinicaltrials. Thisimprovementisparticularlynotableinbiologics andprotein-basedtherapeutics,wherepredictivemodelingcanenhancetheevaluationofefficacy,safety,and manufacturabilitybeforecostlyexperimentaltesting begins.Beyondtraditionalmachinelearningsystems, theemergenceoflargelanguagemodels(LLMs)and vision–languagemodels(VLMs)hasfurtherexpanded thecapabilitiesofAIinbiomedicalresearch.Theseadvancedmodelsarecapableofanalyzingvastamounts ofbiomedicalliterature,generatinghypotheses,predictingdrug–targetinteractions,andevendesigning entirelynewmolecularstructures.Byintegratinginformationfromchemical,genomic,andclinicaldatasets, suchsystemscanassistresearchersinmakingmoreinformeddecisionsduringtheearliestphasesofdrugdevelopment,includingbiomarkerdiscoveryandclinical trialplanning.[5]Theneedforacceleratedinnovation inophthalmicmedicinebecomesevenmoreevident whenconsideringtheglobalburdenofvisualimpairment.AccordingtoWorldHealthOrganization(WHO) datareleasedin2023,approximately2.2billionpeople worldwideareaffectedbynearordistancevisionimpairment.Alarmingly,nearlyhalfofthesecasesremain untreatedorcouldpotentiallyhavebeenprevented withappropriateinterventions.Thesocietalandeconomicconsequencesofvisionlossareimmense,with globalproductivitylossesestimatedtoreach$411billionannually. Thesituationisalsoconcerninginthe UnitedStates,whereoculardisordersaffectasubstantialportionofthepopulation.Recentestimatesindicatethataround90millionAmericansaged40years andolderexperiencesomeformofvisionimpairment oreyedisease. Thisrepresentsmorethan60%ofindividualswithinthatagegroup.Thecausesofvision lossarediverseandmultifactorial,butseveralmajor conditionsaccountforalargeproportionofcases,includingage-relatedmaculardegeneration,glaucoma, cataracts,diabeticretinopathy,andretinalveinocclusion.Withouteffectivemanagement,theseconditions canprogressivelydamageoculartissuesandultimately leadtopermanentblindness. Althoughthepharmaceuticalindustryhasmadenotableadvancesinthe preventionandmanagementofeyediseases,current therapiesoftenfallshortofeliminatingorreversing theseconditions.Oneofthecentralchallengesinophthalmicdrugdevelopmentliesindeliveringtherapeuticallyeffectivedrugconcentrationstotheappropriatetissueswithintheeye.Theeyepossessesmultiple protectivebarriersthatlimitdrugpenetration,making efficientdrugdeliveryparticularlydifficult.[6]
Mostcurrentlyavailableophthalmicmedicationsare administered as topical formulations, including eye drops,suspensions,emulsions,andointments.While theseformulationsareconvenientandnoninvasive, they suffer from extremely low drug bioavailability. Severalphysiologicalfactorscontributetothislimitation,includingrapidtearturnover,blinking,drainage throughthenasolacrimalduct,nonproductiveabsorptionthroughconjunctivaltissues,andthelimitedpermeabilityofthecornealmembrane.Asaresult,topicallyadministereddrugstypicallyremaininthetear filmforonlyonetothreeminutes,andlessthan5% oftheadministereddoseactuallypenetratesocular tissues[7].Thesepharmacokineticlimitationssignificantlyreducetherapeuticeffectiveness,particularly whenattemptingtotreatdiseasesaffectingdeeperocularstructures.Topicalformulationsaregenerallyeffectiveforconditionsaffectingtheanteriorsegment oftheeye,suchasconjunctivitisorcornealinflammation.However,theyarelargelyineffectivefordiseases oftheposteriorsegment, whichincludestheretina, vitreoushumor,andchoroid.[8]Posteriorsegmentdiseasesrepresentoneoftheleadingcausesofvisionimpairmentindevelopedcountries.Treatingtheseconditionsisespeciallychallengingbecausedrugsmust penetratemultipleanatomicalbarrierstoreachretinaltissues. Inrecentyears,theintravitrealinjection (IVI) route has become increasingly common for deliveringdrugsdirectlyintothevitreouscavity,allowingtherapeuticagentstoreachretinaltargetsmore effectively.Severalintravitrealproducts,includinginjectablesuspensions,solutions,andsustained-release implants, havereceivedapprovalfromtheU.S.FDA [9].Despitetheseadvances,thenumberofavailable ophthalmicdrugsremainsrelativelysmallcompared withmedicationsdevelopedforotherroutesofadministration,suchasoraltherapies. Thisdisparityhighlightstheurgentneedforinnovativedrugdeliverytechnologiescapableofovercomingocularbarriersand improvingdrugresidencetimewithintheeye. Consequently,researchersareactivelyexploringarange of next-generation ocular drug delivery systems, including nanoparticle-based carriers, biodegradable implants,insituforminggels,microneedlesystems, andgene-basedtherapeutics. Theseemergingtechnologies aim to enhance drug penetration, prolong therapeuticeffects,andreducetheneedforfrequent invasiveproceduressuchasrepeatedintravitrealinjections[10]. Equallyimportantistheneedtocarefully evaluatethemanufacturing,scalability,andregulatory considerationsassociatedwiththeseadvanceddeliverysystems. Successfultranslationfromlaboratory
Perfluorohe that Miebo xyloctane stabilizesthetear
Lotilaner Xdemvy
Izervay complement
Tebentafusp Kimmtrak
Faricimab Vabysmo
Omidenepag Selectiveagonistofthe glaucoma Omlonti isopropyl prostaglandinE2receptor orocular
Satralizumab Enspryng blockinginterleukin-6 optica 2020 2021
Teprotumumab Tepezza
Brolucizumab Beovu
Cenegermin Oxervate nervegrowth
Latanoprostene Nitricoxide–donating Reductionof Vyzulta bunod prostaglandinanalog intraocularpressure
Netarsudil Rhopressa
Lifitegrast Xiidra blocksT-cellmediated Dryeyedisease 2016 Notapproved
Tafluprost Zioptan
Ocriplasmin Jetrea formofhuman
Aflibercept Eylea
Alcaftadine Lastacaft
Semifluorinatedalkane Dryeyedisease 2023 Notapproved filmlipidlayer Antiparasiticagent Demodex 2023 Notapproved targetingectoparasites blepharitis Geographic Aptamerinhibitortargeting atrophy 2023 Notapproved associated componentC5 withAMD BispecificT-cellengager linkinggp100 Uvealmelanoma 2022 2022 peptide–HLA complexwithCD3 Dual-targetmonoclonal NeovascularAMD antibodyinhibiting and 2022 2022 VEGF-Aand diabeticmacular ANG2pathways edema Open-angle 2022 Notapproved hypertension Monoclonalantibody Neuromyelitis
receptorsignaling spectrumdisorder IGF-1receptor–targeted Thyroideyedisease 2020 Notapproved monoclonalantibody Neovascular Anti-VEGFmonoclonal age-related 2019 2020 antibodyfragment macular degeneration Recombinanthuman Neurotrophic 2018 2017 keratitis factor(NGF) 2017 Notapproved Open-angle 2019 Rho-associatedprotein glaucoma 2017 (marketedas kinase(ROCK)inhibitor orocular Rhopressa) hypertension LFA-1antagonistthat
inflammation Open-angle 2008 Selectiveprostaglandin glaucoma 2012 (marketedas Freceptoragonist orocular Taflotan) hypertension Recombinanttruncated Symptomatic vitreomacular 2012 2013 plasminenzyme adhesion Recombinantfusion Neovascular proteinbinding age-related 2011 2012 VEGF-Aandplacental macular growthfactor degeneration HistamineH1 Allergic 2010 Notapproved receptorantagonist conjunctivitis researchtoclinicalapplicationwillrequiremultidisciplinarycollaborationinvolvingpharmaceuticalscientists,clinicians,engineers,andregulatoryexperts. Byaddressingthesechallenges,itmaybepossibleto significantlyexpandthetherapeuticarsenalavailable fortreatingcomplexoculardiseases,particularlythose affectingtheposteriorsegmentoftheeye.[11]
Theeyeisahighlyspecializedorganresponsiblefor vision and is considered an immune-privileged site, meaning it has protective mechanisms that limit inflammationandprotectdelicatevisualtissues.Itscomplexstructureconsistsofspecializedtissues,nerves, muscles, glands, and a well-organized vascular system. Whilethesefeaturesareessentialformaintainingnormalocularfunction,theyalsocreatesignificant challengesforeffectivedrugdelivery.Fromapharmaceuticalperspective,theanatomicalandphysiological characteristicsoftheeyeactasprotectivebarriersthat restricttheentry,distribution,andretentionoftherapeuticagents.Thesebarriersaredesignedtoprevent harmfulsubstancesfromenteringoculartissues,but theycanalsoreducetheeffectivenessofadministered medications.[12] Although the anatomy of the eye has been widely describedinscientificliterature,thissectionfocuses specifically on the barriers that interfere with ocular drug delivery and briefly highlights possible approachestoovercomethem.Understandingthesebarriersisessentialforthedevelopmentofeffectiveophthalmictreatmentsdeliveredthroughroutessuchas topicaladministration,subconjunctivalinjection,intravitrealdelivery,andsubretinaltherapy.Ingeneral, ocularbarrierscanbebroadlycategorizedintostatic barriersanddynamicbarriers.Staticbarriersinclude thephysicalstructuresoftheeyethatlimitdrugpenetration,whereasdynamicbarriersinvolvephysiologicalprocessessuchastearturnover,blinking,andblood circulationthatremovedrugsfromtheocularsurface. Together,thesebarriersstronglyinfluencedrugstability,distribution,andoveralltherapeuticeffectiveness inoculartreatments.[13]
Staticbarriersintheeyerefertothestructuralcomponentsthatphysicallylimitthepenetrationanddistributionofdrugswithinoculartissues. Thesebarriers arisefromtheinherentanatomicalfeaturesandcompositionoftheeye.Amongthem,thecorneaandthe blood–ocularbarriersareconsideredthemostimportantobstaclesthatinfluenceophthalmicdrugdelivery. Thecornea,whichformsthetransparentfrontsurface oftheeye,actsasthefirstandmostsignificantbarrier fordrugsadministeredtopically[14].Itconsistsofmultiplelayerswithbothhydrophilicandlipophiliccharacteristics,makingdrugpermeationhighlydependenton thephysicochemicalpropertiesofthedrugmolecule. In general, compounds with moderate lipophilicity (logPvaluesbetweenabout2and4)tendtopenetrate thecorneamoreeffectively.Tightjunctionsbetween epithelialcellsrestrictthepassageofhydrophilicor highlypolarmolecules,makingdrugsolubilityandformulation characteristics critical for successful delivery.BeneaththecornealepitheliumliestheBowman’s layer,athinacellularmembranemeasuringroughly 8–12 minthickness.Althoughitwasoncebelievedto µ providestructuralprotectionandregulatethemovementoflargemolecules,morerecentstudiessuggest thatitsdirectinfluenceondrugtransportisrelatively limited. Interestingly, research has shown that the thicknessofBowman’slayergraduallydecreaseswith ageandhasminimalregenerativecapacity,whichmay haveimplicationsforfutureoculardrugdeliverystrategies.[15]Thecornealstroma,locatedbelowBowman’s layer,accountsfornearly90%ofthetotalcornealthicknessandisprimarilyhydrophilicinnature. Because ofitsfluid-likestructure,itgenerallyallowsthediffusionofmanypolarmolecules,includingrelativelylarge compounds.However,hydrophobicdrugsmayexperiencegreaterresistancewhenmovingthroughthislayer. Inaddition,thestromacanactasatemporarydrug reservoir,retainingcompoundsthathavecrossedthe epithelium.Whilethiseffectcanbebeneficialfortreatingcornealconditions,itmaylimitdrugavailabilityin deeperoculartissuessuchastheretina.Betweenthe stromaandtheinnerendotheliallayerliesDescemet’s membrane, whichservesasathinsupportivestructure. This membrane usually does not significantly restrictsmalldrugmolecules,butitcanhinderthepassageoflargermacromoleculesorparticulatedrugsystems.[16]Theinnermostcorneallayeristhecorneal endothelium, which plays a key role in maintaining cornealhydrationandtransparency.Endothelialcells activelyregulatefluidbalancethroughiontransport systems,includingsodium–potassiumATPasepumps andothertransportproteins. Althoughtheendotheliumisgenerallymorepermeablethanthecornealepithelium,drugmovementacrossthislayercanstillbe influencedbymolecularsizeandotherphysicochemicalproperties.Inadditiontothecornea,theeyecontainsspecializedvascularbarriersthatfurtherrestrict drugentryfromthebloodstream[17].Theseinclude theblood–aqueousbarrierandtheblood–retinalbarrier,bothofwhichtightlyregulatetheexchangeofsubstancesbetweensystemiccirculationandoculartissues. Theblood–aqueousbarrier,formedbytheepitheliumoftheciliarybodyandbloodvesselsofthe iris,controlsthecompositionoftheaqueoushumor. Theblood–retinalbarrier,composedofretinalpigment epithelialcellsandretinalvascularendothelialcells, protectstheneuralretina.Together,thesebarrierssignificantlylimittheeffectivenessofdrugsadministered throughsystemicroutes.[18]
Inadditiontothestructuralbarriersoftheeye,several physiologicalprocessesactivelylimittheeffectiveness ofoculardrugdelivery.Thesedynamicbarriersfunctionbycontinuouslyremovingordilutingdrugsfrom theocularsurface,therebyreducingthetimeavailable fordrugabsorption[19].Importantdynamicbarriers includethetearfilm, blinkingactivity, nasolacrimal drainage,andconjunctivalbloodcirculation.Thetear filmisoneoftheprimaryfactorsinfluencingdrugretentionontheocularsurface. Thisthinprotectivelayer, approximately 6 µm thick, plays a crucial role in lubricating the eye and maintaining a smooth optical surfaceoverthecornea,whichisnecessaryforclear vision. Itiscomposedofthreelayerslipid,aqueous, andmucin,eachcontributingtotearstabilityandocularprotection. Thecompositionoftearscanvarydependingonthestimulusthattriggerstearproduction, suchasemotionalresponsesorirritation.Thisvariabilitycaninfluencehowdrugsbehavewhenappliedto theeye. Additionally,thetearfilmhasapHtypically rangingbetween6.5and7.6,whichmustbeconsideredduringformulationdevelopmenttoensuredrug stability and minimize irritation. Tears also contain protectiveproteinssuchaslysozyme,lactoferrin,and immunoglobulin A, which provide antimicrobial defense but may interact with certain drug molecules. [20]Anotherchallengeassociatedwiththetearfilmis itsrapidturnover. Thetotaltearvolumeisrelatively small,about7–9 µL,andtearsecretionoccursatarate ofroughly1 µLperminute.Thismeansthetearlayeris refreshedwithinashorttime.Blinkingfurtheracceleratesthisprocess,asthehumaneyemayblinkmultiple timesperminute.Althoughthesemechanismsareessentialformaintainingocularhealth,theysignificantly reducetheresidencetimeofdrugsappliedtotheeye, makingitdifficultformedicationstoremainonthesurfacelongenoughtobeabsorbedeffectively[21].The nasolacrimaldrainagesystemalsoplaysanimportant roleindruglossfollowingtopicaladministration.This systemremovesexcesstearfluidfromtheocularsurfaceanddrainsitintothenasalcavity.Tearsfirstenter smallopeningscalledpunctalocatedontheinnerside oftheeyelids. Fromthere,fluidpassesthroughnarrowchannelsknownascanaliculi,whichjointoform thelacrimalsacbeforedrainingintothenasolacrimal ductandeventuallythenasalcavity[22].Becausethis drainageprocessiscontinuous,asignificantportionof anadministeredeyedropcanberapidlyclearedfrom theeye.Inmanycases,upto80%oftheinstilleddrug maybelostwithin15–30secondsafteradministration. Larger drop volumes can further increase drug loss bytriggeringreflextearingandoverflow.Inaddition, drugsthatenterthenasolacrimalpathwaymaybeabsorbedthroughthenasalmucosa,potentiallyleading tosystemicexposureandunwantedsideeffects.The conjunctivarepresentsanotherdynamicfactoraffectingoculardrugdistribution. Thisthinmucousmembranecoversthefrontoftheeyeandlinestheinner surfacesoftheeyelids. Itcontainsadensenetwork of blood vessels that supply nutrients to the ocular surface.Whiletheconjunctivacanallowcertaindrugs topenetrateitsepitheliallayer,theunderlyingblood vesselscanrapidlytransportthesemoleculesintothe systemiccirculation. Asaresult,theamountofdrug reachingdeeperoculartissuesmaybereduced,and theriskofsystemicabsorptionmayincrease.Takentogether,thesedynamicbarrierssignificantlyinfluence thefateofdrugsadministeredtotheeye.Theircombinedeffectsoftenleadtorapiddrugeliminationand limitedocularbioavailability.Forthisreason,thedevelopmentofeffectiveophthalmictherapiesfrequently requiresinnovativeformulationapproachesorspecializeddeliverysystemsdesignedtoprolongdrugresidencetimeandimprovedrugpenetrationintotarget oculartissues.[23]
Ocularimplantshaveemergedasanimportantinnovation in ophthalmic drug delivery, providing an alternativetotraditionalmethodssuchaseyedropsor repeatedinjections. Theseimplantablesystemsare specificallydesignedtoreleasetherapeuticagentsat atargetedocularsiteoveracontrolledperiodoftime [24].Bymaintainingsustaineddruglevelsdirectlyat thesiteofaction,ocularimplantscanreducethefrequencyofdrugadministrationandminimizetheneed for repeated invasive procedures, which may otherwiseincreasetheriskofinfectionorinflammation.One ofthekeyadvantagesofocularimplantsistheirability to deliver drugs locally while maintaining stable therapeuticconcentrationsforextendedperiods.The placementoftheimplantdependsonthediseaselocationwithintheeye. Fordisordersaffectingtheanteriorsegment,implantsareoftenpositionedinthe subconjunctivalregion,whereasdiseasesoftheposteriorsegmentaretypicallytreatedusingintravitrealor suprachoroidalimplants.Anotheroptionisintrascleralimplantation,whichcanbeusedforbothanterior andposteriorsegmentconditions[25].Becausethese devicesarepositionedclosetothetargettissue,they canenhancedrugdeliverywhilereducingsystemicexposure.[26]
The implantation procedure varies according to the locationofdelivery.Subconjunctivalimplantsareusuallyinsertedontothesclerathroughasmallincisionin theconjunctiva.Intrascleralimplantsareplacedwithin asmallpocketcreatedinthescleraltissue,typicallyat abouthalfthescleralthickness,whichallowsdrugdiffusiontowardposterioroculartissueswithlimitedsystemicabsorption.Intravitrealimplantsareintroduced directlyintothevitreouscavityusingspecializedapplicators.Thisapproachprovidesdirectaccesstoretinal tissuesandeliminatestheneedforsutures,asthedevicecanbedeliveredthroughaneedleorsmallsclerotomy.Theparsplanaregioniscommonlyselectedas theinsertionsitebecauseitminimizestheriskofdamagetosensitiveocularstructures.Basedonthematerialsusedandthemechanismofdrugrelease,ocularimplantsaregenerallyclassifiedintobiodegradableand non-biodegradablesystems[27].Non-biodegradable implantsarecapableofreleasingdrugsatacontrolled rateforextendeddurations,sometimeslastingseveral years.However,thesedevicesmusteventuallyberemovedorreplacedoncethedrugsupplyisexhausted. Incontrast,biodegradableimplantsgraduallydegrade withinthebodyafterdeliveringthedrug,eliminating theneedforsurgicalremoval[28]. Theycanalsobe manufactured in various shapes and may allow implantation procedures to be performed more easily. Severalocularimplantproductshavealreadyreceived approvalfromtheU.S.FoodandDrugAdministration (FDA).Amongthese,mostnon-biodegradableimplants relyonpolymerssuchaspolyvinylalcohol(PVA)toregulatedrugrelease. Biodegradableimplantsaretypicallycomposedofbioabsorbablepolymers,including poly(lacticacid)(PLA)andpoly(lactic-co-glycolicacid) (PLGA).OnerecentlyapprovedimplantisiDoseTR,a smalltravoprost-containingdevicedesignedtocontinuouslyreleasemedicationforglaucomamanagement. AnothernotablesystemisSuvismo,arefillableintravitrealimplantthatdeliversranibizumabthroughaport deliverysystem,allowingrefillingapproximatelyevery sixmonths.[29]
Insitugellingsystemsrepresentanadvancedapproach foroculardrugdelivery.Theseformulationsaretypicallyadministeredasliquideyedrops,butoncethey reachtheocularsurface,theyundergoatransformationintoasemi-solidgel.Thisphasetransitionoccurs withintheconjunctivalsac(ocularcul-de-sac)andcan betriggeredbyenvironmentalfactorssuchaschanges in temperature, pH, ionic concentration, or light exposure.Aftergelformation,theformulationbecomes moreviscousandremainsincontactwithoculartissuesforalongertimecomparedwithconventionaleye drops.Thepolymersusedinthesesystemsplayakey roleinthegelationprocess.Theycreateaprotective andadhesivematrixovertheocularsurface,allowing thedrugtoremainatthesiteofactionforextended periods.Becauseofthisprolongedresidencetime,in situgelsenhancedrugabsorptionandreducetheneed forfrequentadministration. Theviscoelasticpropertiesofthegelenableittoadhereeffectivelytoocular tissuessuchasthecorneaandconjunctiva, improvingdrug–tissueinteractionandfacilitatingdiffusion throughepithelialbarriers. Inaddition,manyinsitu gelsaredesignedtoreleasethedruggradually,which helpsmaintaintherapeuticdruglevelsoverasustained period.Asaresult,thesesystemshavethepotentialto overcomesomemajorlimitationsoftraditionalophthalmicdosageforms,includingrapiddrugeliminationandpoorbioavailability.[30]Recentresearchhas increasinglyfocusedonthermosensitiveinsitugels, whichrespondtotemperaturechangesafteradministration.Thesesystemsremaininliquidformatroom orstoragetemperaturebutformagelonceexposedto thephysiologicaltemperatureoftheeye.Avarietyof polymershavebeenexploredforthispurpose,includingchitosan,polyethyleneglycol,polycaprolactone, poloxamers,polylactide,andpolyglycolide.Thesematerialscanself-assembleintotemperature-dependent micellarstructuresthattransformintogelswhenthe temperatureincreases,therebyimprovingdrugretention and potentially enhancing delivery to both the anteriorandposteriorsegmentsoftheeye.SeveralexandGlaucoma-RelatedDisorders
Ganciclovir Vitrasert 4.5mg cytomegalovirus
Fluocinolone Managementof Intravitreal Around30 Polyvinylalcohol Retisert 0.59mg acetonide chronicuveitis
Dexamethasone Ozurdex 0.7mg
Fluocinolone Intravitreal Nearly36 Iluvien 0.19mg diabetic acetonide
Fluocinolone Intravitreal Upto36 combinedwith Yutiq 0.18mg non-infectious acetonide
Bimatoprost Durysta 10 µg pressure intracameral including 2020
Ranibizumab Susvimo 2mg age-related
Travoprost iDoseTR 75 µg intraocular
perimentalstudieshavedemonstratedthepotentialof thistechnology.Forexample,formulationscontaining gatifloxacinpreparedwithalginateandhydroxypropyl methylcellulose(HPMC)showedfavorablerheological propertiesandeffectivegelformationinsimulatedtear fluid.Theoptimizedcompositionprovidedcontrolled drugreleasewithoutirritatingoculartissues. Inanotherstudy,researchersdevelopedamicroemulsionbasedinsitugelcontainingbetamethasoneandmoxifloxacinforuseafterocularsurgery.Thisformulation remainedliquidduringstoragebutgelledatocularsurfacetemperature,enablingaccuratedosingandsustaineddrugrelease.Exvivoexperimentsusingporcine cornealtissueindicatedimproveddrugpermeability comparedwithconventionalformulations. Otherinvestigationshaveincorporateddrug-loadednanoparticleswithininsitugelstofurtherenhanceperformance. For instance, moxifloxacin-loaded nanoparticle gels demonstrated prolonged drug release and effective antibacterialactivityagainstkeratitis. Experimental studiesalsoconfirmedgoodoculartoleranceandre-
Polyvinylalcohol Treatmentof Intraocular Approximately (PVA)and 1996 implant 5–8months ethylene-vinylacetate (CMV)retinitis (EVA)polymers 2005 implant months (PVA)matrix Diabeticmacular Upto6 edema(DME), Poly months retinalvein Intravitreal (lactic-co-glycolicacid) dependingon 2009 occlusion implant (PLGA) disease (RVO),and biodegradablesystem condition uveitis Treatmentof Polyimidetubewith PVAandsilicone 2014 implant months macularedema adhesivecomponents Polyimidetubing Therapyfor 2018 implant months PVAand uveitis siliconeadhesive Reductionof Biodegradable intraocular Ophthalmic/ polymers Several months (IOP)in implant PLGA,PDLA,PDLLA, glaucoma andPEG3350 Managementof Intravitreal Upto6 neovascular(wet) Implantable port months port-based 2021 delivery between macular drugdeliverydevice system refills degeneration Titaniumreservoir Lowering Intracameral About3 with 2023 implant months asemipermeable pressure membranecoating
ducedbacterialloadininfectedcornealtissue. Similarly,temperature-responsivegelsystemshavebeen exploredfordeliveringcysteamine,adrugusedtotreat cystinosis.Thesesystemswereabletoreleasethedrug graduallyoverseveralhourswhilepromotingefficient cornealabsorptionwithminimalsystemicexposure. [31]
Nanomicellesarenanosizedcarrierstypicallyranging fromabout10to100nm.Theyconsistofahydrophobiccoreformedbyfattyacylchainsandahydrophilic outershellcomposedofpolarheadgroups.Becauseof thisamphiphilicstructure,nanomicellescansolubilize bothhydrophilicandpoorlywater-solubledrugs,makingthemeffectivedeliveryvehiclesfortargetingboth theanteriorandposteriorsegmentsoftheeye.Previousresearchhashighlightedimportantconsiderations forusingnanomicellesinoculardrugdelivery,particularlyregardingtheirabilitytocrossocularbarriers. [32]Thesesystemsofferseveraladvantages,including simplepreparationmethods,smallparticlesize,and ahighcapacityfordrugencapsulation.Suchfeatures contributetoimproveddrugstability,enhancedocularbioavailability, andbettertherapeuticoutcomes. In addition, nanomicelles are increasingly being exploredforoculargenedeliveryapplications.Particle sizeplaysacrucialroleintheirperformance,assmaller nanomicellesgenerallyexhibitbetterpermeabilityand moreefficienttransportacrossoculartissues.Recent experimentalstudieshavedemonstratedthepotential ofnanomicellesinimprovingdrugdeliveryinocular diseasemodels.Theiramphiphilicnature,nanoscale size,andefficientdrug-loadingabilityenablethemto significantlyincreasethesolubilityandbioavailability ofpoorlysolubledrugs.Forexample,PingLuandcolleaguesdevelopedTPGS-basednanomicellesloaded withbutenafine(BTF)forthetreatmentoffungalkeratitis[33].TheformulationdramaticallyincreasedBTF solubility and improved its permeability across the cornea. Inanimalmodels,thisnanomicellarsystem producedbettertherapeuticoutcomesthanconventionalformulations,mainlyduetoprolongedretention ontheocularsurfaceandimproveddrugpenetration intocornealtissues. Overall,nanomicelle-basedsystemsrepresentapromisingstrategyforovercoming oculardrugdeliverybarriersandenhancingtreatment effectivenessforvariouseyedisorders.[34]
Microparticlesrepresentanimportantstrategyinocular drug delivery because they can provide prolongeddrugreleaseandimprovedocularbioavailability.Thesetinypolymer-basedparticles,typicallyrangingfrom1–1000 µm,aredispersedinaliquidmedium andcanencapsulatedrugseitherbyphysicalentrapment within the polymer matrix or by chemical attachment to the polymer structure. Based on their structure,microparticlesaremainlyclassifiedasmicrospheresandmicrocapsules. [35]Aftertopicaladministration,microparticlestendtosettleintheocularcul-de-sac,whichhelpsextendtheprecornealresidence time of the drug. The medication is then releasedgraduallythroughmechanismssuchasdiffusion,polymerdegradation,orchemicalreactions.This sustainedreleasereducestheneedforfrequentdosingandenhancestheoveralltherapeuticavailability ofthedrugintheeye. However,therelativelylarger particle size may sometimes cause ocular irritation. Therefore, polymers used for ophthalmic microparticles should ideally be biodegradable, biocompatible, andcapableofadheringtooculartissues. Typically,onlysmallvolumes(50–100 µL)ofmicroparticle formulations can be safely injected into the eye. Severalstudieshighlightthetherapeuticpotentialof microparticlesystems. Forinstance,PLGA-basedmicroparticlescontainingEPO-R76E,amodifiedformof erythropoietin,demonstratedneuroprotectiveeffects onretinalneuronsandopticnerveaxonswithoutsignificantly increasing hematocrit levels in glaucoma models. Another investigation utilized PLGA microspheresloadedwithdexamethasone,melatonin,and coenzymeQ10,whichshowednotableneuroprotective effectsinamousemodelofchronicocularhypertensionafterasingleintravitrealinjection[36]. Inaddition,dexamethasone-loadedPLGAmicrosphereswere reportedtoeffectivelyreduceinflammationinrabbit eyesfollowinglipopolysaccharide-inducedocularinflammationwithin30daysoftreatment.[34]
Liposomes are microscopic vesicular carriers composedofoneormorephospholipidbilayerssurrounding aqueous compartments. Based on the number of lipid layers and particle size, liposomes are commonlyclassifiedintothreetypes:multilamellarvesicles(MLVs),whichcontainmultiplelipidbilayersand usuallyexceed300nminsize;smallunilamellarvesicles (SUVs), typically ranging from 10–100 nm; and large unilamellar vesicles (LUVs), generally measuringbetween100–300nm. Inoculardrugdelivery,liposomesareparticularlyadvantageousbecausetheir lipidcompositioncloselyresemblesbiologicalmembranes,enablingthemtointeracteffectivelywiththe cornealandconjunctivalsurfacesandenhancedrugabsorption[35].Thissystemisespeciallyusefulfordrugs withpoorsolubility,limitedpermeability,lowpartition coefficients,orrelativelylargemolecularweights.The effectivenessofliposomesinophthalmicapplications islargelyattributedtotheirexcellentbiocompatibilityandtheirabilitytoencapsulatebothhydrophilic andhydrophobictherapeuticagents.Severalstudies have reported their potential in delivering drugs to bothanteriorandposteriorsegmentsoftheeye.Surfacechargealsoplaysanimportantroleindeterminingtheirperformance;positivelychargedliposomes tendtoexhibitimprovedadherenceandabsorptionon thenegativelychargedcornealsurfacecomparedwith neutralornegativelychargedvesicles.Inaddition,liposomesaregenerallybiodegradableandcanprovide controlledorsustaineddrugrelease.Despitetheseadvantages,certainlimitationsexist.Forexample,liposomalsuspensionsmaysometimescausetemporaryvitreousopacityorblurredvision.Moreover,ininflamed oculartissues,liposomesmayexhibitshorterhalf-lives, possiblyduetochangesintheblood–retinalbarrier and increased degradation by infiltrating inflammatorycells. [36]Recentresearchhasexploredvarious liposomalformulationsfortopicaloculartherapy.For instance, a study conducted in 2022 developed a liposomalformulationoffluconazoleandcomparedit withaconventionalcommercialproduct.Theliposomalsystemdemonstratedsustaineddrugreleasewith zero-orderkineticsandshowedhigherexvivocorneal permeation.Pharmacokineticanalysisindicatedsignificant increases in both the area under the curve andthemeanresidencetimecomparedwiththemarketedformulation,whilepreclinicalstudiesreported noocularirritationorhemolyticeffects[37].Inanother investigation,liposomescarryingthepoorlysoluble antiglaucomadrugacetazolamidewereincorporated intoahydroxypropylmethylcellulose(HPMC)-based formulation. Thishybridsystemmarkedlyimproved ocularbioavailability,mainlyduetoincreasedviscosityfromHPMC,whichprolongedprecornealresidence time,andadispersionmediumdesignedtomimicthe naturaltearfilmenvironment. Studieshavealsoexaminedhowliposomecharacteristicsinfluenceretinal delivery.Factorssuchasparticlesize,surfacecoating, andchargesignificantlyaffectretinalpenetration.Experimentalfindingssuggestthatsmallerliposomes, around50nmindiameter, canpenetrateretinaltissuesmoreefficientlythanlargervesicles,whilePEGylationandnegativelychargedsurfaceshelpachieve moreuniformretinaldistribution.[38]Althoughliposomaldrugdeliveryhasshownconsiderablepromise, thedevelopmentandregulatoryapprovalofsuchformulationsremainchallenging.Onlyalimitednumber ofliposomalproductshavereceivedapprovalfromthe U.S.FoodandDrugAdministration(FDA),largelydue tothecomplexityoftheirformulationandmanufacturing processes. The FDA issued specific guidance forliposomaldrugproductscoveringchemistry,manufacturing,controls,pharmacokinetics,andlabeling requirements. Currently,onlyafewliposome-based productsareavailableforocularuse.Forexample,Visudyne®containsverteporfinandisadministeredintravenouslyforphotodynamictherapyinconditionssuch as age-related macular degeneration and choroidal neovascularization.Anotherproduct,TearsAgain®,isa liposomalspraydesignedtostabilizethetearfilmand providelubricationinpatientswithdryeyedisease. Clinicalstudieshavedemonstratedthatthisliposomal sprayoffersimprovedsymptomaticreliefcompared withsomeconventionalocularlubricants.[39]
Polymeric nanoparticles are colloidal carriers typically smaller than 1000 nm and are prepared using biodegradablepolymers.Inoculardrugdelivery,the particlesizeplaysacriticalroleindeterminingdrug loading,stability,anddistribution.Generally,nanoparticlesrangingfrom200–400nmareconsideredoptimal for topical ophthalmic administration because they canbeefficientlyabsorbedthroughthecorneaand conjunctiva[40].Theirsmallsizehelpsreduceocular irritation,enhancepenetrationacrossocularbarriers, andimproveinteractionwiththemucosalsurfaceof theeye. Surfacechargealsoinfluencestheirperformance;positivelychargednanoparticlesoftenremain longerontheocularsurfaceduetoelectrostaticattractionwithnegativelychargedoculartissues.Avariety ofnaturalandsyntheticpolymersareusedtodevelop thesesystems,includingPLA,PLGA,chitosan,sodium alginate,albumin,andpolycaprolactone.Structurally, nanoparticles are generally classified into nanocapsulesandnanospheres.Nanocapsulesconsistofacore surroundedbyapolymericshellwherethedrugisenclosed,whereasnanosphereshavethedruguniformly dispersedthroughoutthepolymermatrix.Thesesystemsprovideadvantagessuchasimproveddrugprotection,controlledrelease,andenhanceddrugdistributionwithinoculartissues. [41]Duetotheirability toprovidesustaineddrugdeliveryandminimizeirritation,nanoparticlesarewidelyexploredfortreating bothanteriorandposterioreyediseases. Theirprolonged drug release profile can reduce the need for frequentdosing,whichimprovespatientadherence. Tofurtherenhanceocularretention,researchershave alsodevelopedmucoadhesivenanoparticlesthatcan remainontheocularsurfaceforextendedperiods.Recentstudieshighlightthepotentialofthesesystems [42].Forexample,Khaliletal.designedahybridocular insertcontainingazithromycin-loadedPLGAnanoparticlesembeddedinelectrospunpoly(N-vinylpyrrolidone)nanofibers[107]. Thisbiodegradableandmucoadhesivesystemenableddrugreleaseformorethan tendays, reducingdosingfrequencyandimproving therapeuticeffectiveness.Inanotherstudy,tacrolimusloaded nanoparticles prepared with gellan gum usingionotropicgelationdemonstratedprolongedprecornealretentionandsustaineddrugreleaseforthe treatmentofdryeyedisease.Safetyevaluationsconfirmedtheirocularcompatibility.Additionally,chitinbasednanoparticleshavebeeninvestigatedfordeliveringfluconazole,showingacontrolledtwo-stagereleaseprofilewithminimaldrugreleaseduringtheinitial48hours.Overall,nanoparticle-basedsystemsrepresentapromisingstrategyforachievinglong-lasting oculardrugdelivery.[42]
Oculariontophoresisisanemergingnon-invasivetechnique for delivering drugs to both the anterior and posterior segments of the eye. This method uses a mildelectricalcurrenttodrivechargeddrugmolecules acrossoculartissues. Drugtransportmainlyoccurs throughtwomechanisms:electro-osmosisandelectromigration.Duringtheprocedure,anelectrodewith the same charge as the drug is placed near the eye tofacilitatedelivery, whileacounterelectrodewith the opposite charge is positioned elsewhere on the bodytocompletetheelectricalcircuit[43].Inthisprocess,thedrugsolutionalsofunctionsastheconductive medium.Althoughclinicalstudiesinhumansremain limited,severalanimalstudieshavedemonstratedthat bothtranscornealandtransscleraliontophoresiscan effectivelydelivervariousophthalmicagents,includingoligonucleotides.Comparedwithintravitrealinjectionsorsurgicalimplants,iontophoresisislessinvasive andavoidsmanyassociatedcomplications.However, itdoesnotsignificantlyprolongtheintrinsichalf-life of the drug. Experimental studies have shown that transscleraliontophoresiscanfacilitatethedeliveryof anti-VEGFagentstotheretinaandchoroid.Forexample,applyingacurrentdensityofabout6.25mA/cm2 hasbeenreportedtoincreasecornealdrugaccumulationtoapproximately120ng/mg,creatingadepotthat slowlyreleasesthedrugintotheaqueoushumor.[44] Threemainapproachestooculariontophoresishave beendeveloped: transcorneal,transscleral,andcorneoscleraliontophoresis.Thescleraisparticularlysuitablefordrugtransporttotheposteriorsegmentdue toitsrelativelyhighpermeability,largesurfacearea, andhighhydrationlevel.Comparedwithconventional eyedrops,iontophoresiscanimprovedrugpenetration andreducerapiddrugclearance.Itmayalsoenhance patientcompliancecomparedwithrepeatedintraocularinjections[45].However,themethodhascertain limitations,suchastheneedforrepeatedtreatments, possiblemilddiscomfortduringapplication,andlimiteddurationofdrugaction. Recentresearchhasexplored combining iontophoresis with nanoparticlebaseddrugdeliverysystems.Forinstance,latanoprostloadedPLGAnanoparticleshavebeendeliveredusing iontophoresisforglaucomatherapy. Invivostudies indicated that nanoparticles around 300 nm in size producedthemostprolongedtherapeuticresponse. Thisapproachextendedthedrug’seffectformorethan sevendaysanddemonstratedsignificantlygreaterefficacycomparedwiththecommerciallatanoprostformulationXalatan®. Severaldeviceshavebeendeveloped to support ocular iontophoretic delivery. The OcuPhor™system,designedfortransscleraldrugdelivery,includesanapplicator,dispersiveelectrode,anda dosecontrollertoregulatetheelectricalcurrent[46]. Anothersystem,Visulex™,enablestargetedmovement of ionized molecules across the sclera. Iontophoresishasalsobeeninvestigatedforthedeliveryofseveralantibiotics,suchasgentamicin,tobramycin,and ciprofloxacin, as well as therapeutic agents includingdexamethasoneandantisenseoligodeoxyribonucleotides. However, drugswithveryhighmolecular weight,suchasvancomycin,showlimitedsuitability forthistechnique. Overall,oculariontophoresisrepresentsapromisingalternativeapproachfortargeted ophthalmicdrugdelivery.[47]
Contactlensesarepolymer-baseddevicesplacedon thecorneaprimarilytocorrectrefractiveerrors.They maybefabricatedfromeitherhydrophilicorhydrophobicpolymersandaregenerallyclassifiedashardor softlenses.Beyondvisioncorrection,therapeuticcontact lenses have emerged as a convenient platform for ocular drug delivery. Their ability to adhere to the moist ocular surface through surface tension allowsdrugstoremainincontactwitheyetissuesfor longerperiods. Asaresult,drugdiffusionacrossthe corneaincreaseswhiledruglossthroughthelacrimal drainagesystemdecreases. Comparedwithconventionaleyedropsthatrequirefrequentadministration, drug-loadedcontactlensescansignificantlyimprove drugbioavailabilityandreducedosingfrequency[48]. Onecommonmethodforpreparingdrug-loadedlenses involvessoakingthelensesinadrugsolutionsothat thedrugbecomesincorporatedintothepolymermatrix. Drug-eluting contact lenses function as solid dosageformsthatprovidesustaineddrugrelease,ensuringprolongeddrugresidenceonthecornealsurface. Thisapproachminimizestheoveralldrugdose requiredanddecreasessystemicabsorption.In2022, thefirstcommercialdrug-releasingcontactlens,ACU- VUE®Theravision®withKetotifen,wasintroducedfor thetreatmentofocularallergysymptoms. [49]The mechanismofdrugdeliveryinvolvesdiffusionofthe drugintothetearfilmlocatedbehindthecontactlens, followedbydispersionwithinthetearfluidandabsorptionthroughthecornea. Continuousdruguptakeby thecorneamaintainsalowerdrugconcentrationin thetearfilmcomparedwiththelens,creatingaconcentration gradient known as the sink effect, which supportssustaineddrugreleaseandstabletherapeuticlevels.Severalstudieshighlightthepotentialofthis technology.Wuetal.developedpirfenidone-loaded contactlensesusingaPVAinsertbetweensiliconeelastomerlayers,whichincreaseddrugresidencetimein tearsandaqueoushumorwhileusingamuchsmaller drugdosethaneyedrops[108]. Dingetal. designed a contact lens containing microtubes for glaucoma therapy that allowed drug release for up to 45 days andrespondedtochangesinintraocularpressureto regulatedrugdelivery[109]. Similarly, Xuetal. createdamicelle-basedcontactlenscarryinglatanoprost andtimolol, whichsuccessfullyreducedintraocular pressureformorethanaweek[110].Inanotherstudy, Jiaoandcolleaguespreparedanantimicrobialhydrogelcontactlensincorporatingquaternaryammonium chitosanandtannicacid,showingnearlycompleteantibacterialactivityagainstEscherichiacoliandStaphylococcusaureuswhilealsoreducingoxidativestress. [50]
Microneedles(MNs)representanovelandminimally invasivestrategyforoculardrugdelivery. Thesetiny needlestructurescanpenetrateonlyafewhundred micrometersintothesclera,enablinglocalizeddrug administrationwhileminimizingdamagetodeeperoculartissues.Comparedwithconventionalapproaches suchasintravitrealinjections,microneedlesofferadvantagesincludingcontrolleddrugrelease,targeted delivery, ease of administration, and relatively low manufacturingcost. Theyhaveshownparticularpotentialinthetreatmentofposterioreyedisorderssuch asdiabeticretinopathy,age-relatedmaculardegeneration,andposterioruveitis[51].Bydepositingdrugs directlynearthesclera,therapeuticagentscandiffuse into deeper tissues such as the choroid and retina, thereby improving ocular bioavailability and reducingdrugwastage.Variousmicroneedledesignshave beenexplored,includingsolid,hollow,dissolving,and polymer-coated systems. [52] Recent studies have demonstratedthetherapeuticpromiseofmicroneedlebased ocular systems. For instance, dissolving microneedlepatchescomposedofhyaluronicacidand polylactic acid have been investigated for treating fungal keratitis, a major cause of corneal blindness [53].Inexperimentalmodels,thesepatchestemporarilypenetratedthecornealepitheliumandpromoted rapidhealing,showinggreatertherapeuticeffectivenessthanconventionaleyedrops.Similarly,pen-type microneedledeviceshavebeendevelopedtoimprove insertionprecisionandenablelocalizeddrugdelivery withinthecornealstroma. Dissolvingpolymericmicroneedlearraysmadefrompolyvinylalcohol(PVA) andpolyvinylpyrrolidone(PVP)havealsobeenstudiedforproteindelivery. Thesesystemswereableto encapsulateproteinswithoutstructuraldegradation andreleasemorethan80%oftheloadedcargowithin 48hours,demonstratingefficientdrugdeliveryacross scleraltissues.Despitetheseadvantages,certainlimitationsstillexist.Microneedleapplicationmaycause localirritation,inflammation,orinfectionattheinsertionsite,andprolongedplacementontheeyecanlead topatientdiscomfort.Inaddition,challengessuchas limited drug-loading capacity, variability in drug release,andpotentialsafetyconcernsmustbecarefully addressed.Asaresult,althoughmicroneedletechnologyiswidelystudied,microneedle-basedophthalmic formulationshavenotyetreachedroutineclinicaluse. Furtherresearchandoptimizationarenecessarytoimprovetheirsafety,efficacy,andpatientacceptability forlong-termoculartherapy.[54]
Hydrogels are widely explored in ocular drug delivery because of their hydrophilic nature and threedimensionalpolymernetwork,whichallowsthemto retainlargeamountsofwater.Thispropertyprovides severaladvantagescomparedwithconventionaleye drop formulations. Hydrogels can increase the residencetimeofdrugsontheocularsurface,enablesustaineddrugrelease,andallowthedeliveryofmultiple therapeuticagentssimultaneously.Unliketraditional eyedropsthatarerapidlyremovedfromtheeyedue toblinkingandteardrainage,hydrogelsremainlonger atthesiteofapplication,improvingdrugavailability. [55]Hydrogelscanbeadministeredthroughdifferent routes,includingtopical,intravitreal,andintracameral delivery. Theirporousstructuresupportshighdrug loading and controlled release behavior. Recent researchhasfocusedondevelopingstimuli-responsive hydrogels that release drugs when exposed to specifictriggerssuchastemperature,pH,orions.Forinstance,ahydrogelsystemcomposedofcyclodextrins, hyaluronicacid,carrageenan,andgellangumwasdesignedtoimprovetheoculardeliveryoftheantifungaldrugeconazole,demonstratingslowerandmore controlleddrugreleasecomparedwithconventional solutions.Similarly,athermosensitiveglycol-chitosan hydrogelcontaininglevofloxacinshowedimprovedocularbioavailabilityandsignificantlyprolongeddrug retentiononthecornealsurface.[56]
Bispecificantibodies(bsAbs)areengineeredproteins capableofrecognizingtwodifferentepitopeseither onthesameantigenorontwoseparatetargets.This dual-bindingabilityprovidesbroadertherapeuticpotentialcomparedwithtraditionalmonoclonalantibodies. BsAbsaregenerallycategorizedintotwogroups basedonwhethertheycontainafragmentcrystallizable(Fc)region. ThefirstcategoryincludesIgG-like bsAbs,whichpossessanFcregionandstructuralcharacteristicssimilartoconventionalIgGantibodies.Examplesincludetechnologiessuchasknobs-into-holes (KiH),CrossMab,andDuoBody.Thesemoleculestypically have molecular weights close to 150 kDa and exhibitlongerhalf-lives,improvedserumstability,and effectorfunctions.However,theirlargersizemaylimit tissuepenetrationandtheirproductionprocessescan be complex. [57] The second category consists of fragment-basedornon-IgG-likebsAbs,whichlackan Fcregion.Examplesincludesingle-chainvariablefragments(scFv),bispecificT-cellengagers(BiTEs),anddiabodies.Becauseoftheirsmallersize(approximately 25–110kDa),thesemoleculesgenerallydemonstrate bettertissuepenetrationandreducedimmuneactivation. However, they often have shorter half-lives andlowerstabilityincirculation.BsAbscanfunction throughseveralmechanisms,suchaslinkingimmune effectorcellstodiseasedcells, modulatingmultiple signalingpathways,mimickingenzymaticcofactors,or facilitatingtargeteddeliverythroughpiggybackmechanisms[58]. Currently, severalbsAbshavebeenapprovedforclinicaluse,mainlyinoncology.Anexample relevant to ophthalmology is faricimab (Vabysmo®), whichtargetsbothangiopoietin-2(Ang-2)andvascularendothelialgrowthfactor-A(VEGF-A).Byinhibiting thesepathways,faricimabreducesinflammation,vascularleakage,andabnormalbloodvesselformation.It isadministeredasanintravitrealinjectionandisused to treat conditions such as neovascular age-related maculardegeneration,diabeticmacularedema,and retinalveinocclusion.[59]
Overthepastthreedecades,genetherapyhasemerged as a promising strategy for treating several ocular diseases, includingglaucoma. Researchershaveexplored gene-based approaches for many inherited retinal and optic nerve disorders such as X-linked retinoschisis,Ushersyndrome,retinitispigmentosa, choroideremia, Leber congenital amaurosis, achromatopsia, and Leber’s hereditary optic neuropathy. Inaddition,genetherapyisbeinginvestigatedforacquiredretinalconditionslikewetage-relatedmacular degeneration(AMD),diabeticmacularedema,andretinalveinocclusion. Differenttechniquesareusedto introducegeneticmaterialintoocularcells[60].Viral vectorssuchasadenovirusesandadeno-associated viruses(AAV)arecommonlyusedbecausetheyprovide efficientgenetransfer.Adenoviralvectorscancarryrelativelylargegeneticsequences;however,theirusein retinaltherapyislimitedduetorapidclearancecaused bypre-existingimmuneresponses.Non-viraldelivery strategies,includingelectroporation,magnetofection, aptamer-basedsystems,geneguns,andotherphysical approaches,havealsobeenexplored.Thesemethods ofteninvolvedeliveringnakednucleicacidssuchas plasmidDNA,siRNA,mRNA,ormiRNA,althoughtheir stabilityintissuescanbelimitedduetorapiddegradation.[61]Amajormilestoneinoculargenetherapywas theapprovalofLuxturna™(voretigeneneparvovec)by theUSFDAin2017forthetreatmentofLebercongenitalamaurosistype2.ThistherapyusesanAAVvectorto deliverafunctionalRPE65geneencodingtheisomerohydrolaseenzymeinpatientscarryingbiallelicmutations.Genetherapiescanbeadministeredthroughintravitreal,subretinal,orsuprachoroidalroutes.While subretinalandsuprachoroidalinjectionsenableprecisetargetingofretinaltissues,theyoftenrequirespecializedsurgicalproceduressuchasvitrectomy.Suprachoroidaldeliveryisgainingattentionbecauseofits relativelysimpleradministrationanditscurrentclinicaluseintreatmentslikeXipere®(triamcinoloneacetonide)foruveitis.RNA-basedandgene-editingtechnologies are also being investigated for ocular therapy.Forexample,SYL040012,asmallinterferingRNA (siRNA)drugcandidate,targetsthe β 2-adrenergicreceptorintheciliarybodytoreduceintraocularpressure [62]. Studies in rabbits demonstrated about a 20%reductioninintraocularpressurelastingseveral days.Advancedgenome-editingapproachessuchas CRISPR/Cas9havealsoshownpromise.Researchers haveusedthistechnologytotargetviralgenomesin infectionslikeherpessimplexkeratitisandtocorrect mutations associated with glaucoma, including the MYOC gene mutation linked to elevated intraocular pressure. Gene-silencingapproachesarealsobeing exploredfordiseasesinvolvingabnormalbloodvesselgrowth.Forinstance,thesiRNAcompoundSirna- 027(AGN211745)targetsVEGFR-1tosuppressvascular endothelialgrowthfactorsignalinginwetAMD[63]. Experimentalstudieshaveshownthatintravitrealor periocularadministrationofthismoleculecanpersist inretinaltissuesforseveraldays,leadingtoreduced VEGFR-1expressionandinhibitionofpathologicalneovascularization.[64]
Oneoftheearliestlong-actingophthalmicimplants approvedbytheU.S.FoodandDrugAdministration (FDA)wasVitrasert,anon-biodegradableintravitreal implantcontainingganciclovir. Itwasdevelopedfor themanagementofcytomegalovirus(CMV)retinitisin patientswithacquiredimmunodeficiencysyndrome. Thedeviceusedpolymerssuchaspolyvinylalcohol (PVA)andethylene-vinylacetate(EVA)toregulatedrug release and was surgically placed through the pars
planaintothevitreouscavity[65]. Onceimplanted, it delivered ganciclovir at an approximate rate of 1 µgperhourforaboutfivetoeightmonths.Although theimplantprovidedsustainedantiviraltherapy,its usewasassociatedwithseveralcomplications,including endophthalmitis, retinal detachment, and vitreous hemorrhage. In addition, the device required surgicalreplacementafterthedrugreservoirwasdepleted. [66]Followingthisdevelopment,additional non-biodegradableimplantscontainingfluocinolone acetonide were introduced between 2005 and 2018 forthetreatmentofconditionssuchasnoninfectious uveitisanddiabeticmacularedema(DME).Thesesystems, also based on PVA-type polymers, were capableofmaintainingdrugreleaseformuchlongerdurations,insomecasesuptothreeyears.However,incorporatinglargebiomoleculesintonon-biodegradable implantshasprovendifficult,whichhasencouraged theexplorationofalternativedeliverystrategies.One promisingapproachisencapsulatedcelltechnology (ECT),acell-baseddeliverysystemdesignedforsustainedintraoculartherapy.Inthismethod,genetically modifiedcellsthatproducetherapeuticagentsareenclosedwithinasemipermeablepolymercapsule.The membraneallowsnutrients,oxygen,andtherapeutic proteinstodiffuseoutwardwhilepreventingimmune cellsfromenteringthecapsule. Thedeviceissealed atbothendsandanchoredneartheparsplana,where itisattachedtothesclerausingasmalltitaniumloop tomaintainitspositionwithintheeye.[67]Inrecent years,biodegradableocularimplantshavegainedconsiderableattentionbecausetheycandeliverdrugsin acontrolledmannerwhilegraduallydegradingwithin thebody.Thesesystemsareoftenusedforconditions thatrequireaninitialhighdosefollowedbyagradual declineindruglevelsoverperiodsrangingfromafew daystoseveralmonths. Biodegradableimplantsare commonlyfabricatedusingpolymerssuchaspolylacticacid(PLA)andpoly(lactic-co-glycolicacid)(PLGA). Drug release from these materials generally follows first-orderkinetics,typicallyshowinganinitialburst release followed by a gradual decrease in drug concentration. ByadjustingtheratioofPLAtoPLGA,researcherscanmodifythedegradationrateofthepolymerandtherebytailorthedurationofdrugrelease, whichmayrangefromseveralweekstomanymonths. PLGA has been widely utilized in ocular implant designbecauseitisbiodegradable,biocompatible,and capableofsustainingdrugdeliveryoverextendedperiods[68].Duetotheirrelativelylargersizeandlower surface-to-volumeratio,theseimplantscanoftenaccommodatehigherdrugloadscomparedwithsmaller deliverysystems.Awell-knownexampleofthistechnologyisOzurdex,abiodegradableimplantthatdelivers dexamethasone for the treatment of macular edema.ApprovedbytheFDAin2009,thisrod-shaped devicecontains0.7mgofdexamethasoneembedded withinaPLGAmatrix. UsingAllergan’sNOVADURdeliverytechnology,theimplantgraduallybreaksdown intolacticandglycolicacidwhilecontinuouslyreleasingdexamethasoneinthevitreouscavityforuptosix months.Theimplantisadministeredthroughanintravitreal injection, representing an important advancementinsustainedoculardrugdelivery.[69]
Conventionaloculardrugdeliverysystemssuchaseye dropsandoralmedicationshavelongbeenemployed forthemanagementofeyediseases.Althoughthese approachesareclinicallyeffective,theyareoftenassociatedwithseveraldrawbacks,includingpoorpatientadherence,rapidprecornealdrugloss,lowocularbioavailability,andtheinconvenienceofrepeated dailyadministration.Inaddition,thevolumethatcan be retained in the conjunctival cul-de-sac is limited to nearly 30 µL, restricting the amount of drug that canbedeliveredthroughtopicalformulations.These challengeshaveencouragedthedevelopmentofalternativeophthalmicdrugdeliverysystemssuchasmedicatedcontactlenses, ocularinserts, lacrimalplugs, intraocularimplants,andbiodegradableconjunctival devicesdesignedtoprovideprolongedtherapeuticaction.[70]Inrecentyears,three-dimensional(3D)printingtechnologyhasemergedasapromisingplatform forthefabricationofpersonalizedophthalmicdevices and drug delivery systems. However, despite growinginterestinthisfield,researchinvolvingtheapplicationof3Dprintingforoculartherapeuticsremainsat arelativelyearlystage.[71]Three-dimensionalprinting,commonlyknownasadditivemanufacturing(AM), referstothefabricationofobjectsthroughthesequentialdepositionofmaterialsinalayer-by-layermanner usingdigitalcomputer-aideddesign(CAD)models.Owingtoitsautomatedoperation,designflexibility,and abilitytoproducecustomizedstructures,thistechnologyhasgainedconsiderableattentioninpharmaceuticalandbiomedicalapplications.Personalizeddosage formsandmedicaldevicescanbetailoredaccording to the individual requirements of patients, thereby improvingtherapeuticoutcomesandtreatmentefficiency. Furthermore, the compact and portable natureof3Dprintingsystemsallowson-demandmanufacturingathospitals,pharmacies,remotehealthcare settings, andevenenvironmentssuchasspacemissions. Anotherimportantadvantageisitssuitability forsmall-batchproductionduringpreclinicalandclinicalinvestigations,wherefrequentdoseadjustments basedonpharmacokineticandpharmacodynamicresponsesmayberequired.Consequently,additivemanufacturinghasbecomeanattractivestrategyfordevelopingadvancedophthalmicformulationsandimplantableoculardevices.[72]
FusedFilament-Based3DPrinting: Amongtheavailable3Dprintingmethods,materialextrusion-based technologieshavebeenwidelyexploredforthepreparationofoculardrugdeliverysystemsandimplantable devices. Depending on the feed material employed during the extrusion process, these techniques are categorizedintofuseddepositionmodeling(FDM),directpowderextrusion(DPE),andsemi-solidextrusion (SSE),alsoreferredtoasbioprinting. Fuseddepositionmodelingisoneofthemostextensivelyinvestigatedapproachesinpharmaceuticalmanufacturing. Inthistechnique,thermoplasticpolymericfilaments containing active pharmaceutical ingredients (APIs) areheatedabovetheirmeltingtemperatureandextrudedthroughanozzleontoabuildplatformaccordingtoadigitallyprogrammeddesign.Layer-by-layer depositionenablesthefabricationofcomplexgeometrieswithhighprecision.AlthoughFDMcangenerate intricatestructureswithrelativelyfineresolution,additionalpost-processingproceduressuchassupport removal,polishing,andsurfacefinishingareoftennecessarytoimprovethefinalappearanceandqualityof printedproducts.[73]FDMoffersmultipleadvantages, includinglowequipmentcost,rapidproduction,mechanicalrobustnessofprintedobjects,andcompatibilitywithnumerousFDA-recognizedGenerallyRegarded asSafe(GRAS)polymers. Drugincorporationintofilamentscanbeachievedeitherthroughhot-meltextrusion (HME) or by soaking preformed filaments in drugsolutionstopermitpassivediffusionoftheAPI. Amongtheseapproaches,HME-assistedFDMismore frequentlypreferredbecauseitensuresbetterdrugdistributionthroughoutthefilament.Nevertheless,the elevatedtemperaturesinvolvedduringprocessingmay limittheapplicationofthismethodforthermosensitivecompounds. Incontrast,passivediffusionmethodsgenerallyresultinlowerdrug-loadingefficiency, restrictingtheiruseinformulationsrequiringhigher doses. Topical treatment of glaucoma is frequently compromisedbyrapidtearturnover,blinking,nasolacrimaldrainage, andlimitedcornealpermeability, allofwhichreduceoculardrugretention. Injectable therapies,althougheffective,maycausediscomfort andrequireadministrationbytrainedhealthcareprofessionals.Toovercometheselimitations,researchers have explored 3D-printed ocular devices capable of sustaineddrugrelease. Mohamdeenandcoworkers fabricatedpersonalizedcontactlensescontainingtimolol maleate using a combination of HME and FDM techniques. Initially, the lenses exhibited a yellowishappearanceduetothermalexposureduringprinting;however,transparentlensesweresuccessfullyobtainedbyoperatingbelowthemeltingtemperatureof thedrug.Thedevelopedlensessustaineddrugrelease forapproximatelysevendays, althoughissuessuch asinitialburstreleaseandlimitedcornealpermeabilitywerestillobserved. [74]Inanotherinvestigation, Wuandcolleaguesexaminedtheinfluenceofprinting temperatureconditionsonthefabricationofpolylacticacid(PLA)-basedmicroneedlearrayscoatedwith rhodamineBasamodelcompound.Theresearchers demonstratedthatincreasingthenozzleandstagetemperaturesimprovedinterlayerbondingandreduced voidformationwithintheprintedstructures,ultimately enhancingthemechanicalpropertiesofthemicroneedles.[75]
FDM, which utilizes solid filaments as feed material,semi-solidextrusion(SSE)printingemploysgels, pastes,orhydrogel-basedformulations.Thistechnologyhasattractedconsiderableinterestindrugdelivery, tissue engineering, and bioprinting applications becauseitoperatesatrelativelylowtemperaturesandenablessingle-stepmanufacturing.Inaddition,theuseof disposablesyringesorcartridgesmaysimplifyregulatorycomplianceandreducecontaminationrisks.Since filamentpreparationisnotrequired,SSEcanimprove manufacturingefficiencyandformulationconsistency. Theincreasingavailabilityofbiocompatiblehydrogels andbio-inkshasencouragedtheinvestigationofSSE printingforophthalmicapplications.Tagamiandcolleaguesdevelopedlevofloxacin-loadedocularpatches intended for antibacterial corneal therapy using hydrogelinkscomposedofhydroxypropylmethylcellulose(HPMC),mannitol,andxylitol.Afterprintingand freeze-drying,thepatchesexhibitedmucoadhesivebehaviorandformedhydrogelsuponhydration,thereby prolongingocularresidencetimeandreducingdrug drainage.Similarly,Ioannouandcoworkersfabricated biodegradableimplantscontaining5-fluorouracil(5- FU)forpreventingconjunctivalfibrosisfollowingglaucomasurgery[76].Theimplantswerepreparedusing polycaprolactoneandchitosanmatricesloadedwith 1%5-FU.Thedevelopedsystemdemonstratedfavorablebiocompatibility,biodegradability,andsustained drugreleaseoveraneight-weekperiod,potentiallyreducingtheneedforrepeatedpostoperativeinjections. AnotherstudybyPaleelandassociatesfocusedonintracameralimplantscontaininghighconcentrationsof timololforglaucomatherapy.TheSSE-fabricatedimplantsachievedsustaineddrugreleasefornearlyeight weekswithoutproducingsignificanttoxicitytoward humantrabecularmeshworkcells.Collectively,these studieshighlighttheabilityofSSEtechnologytocreate personalizedandbiocompatibleocularimplantscapableofimprovingtreatmentadherenceandlong-term therapeuticoutcomes.[77]
Although FDM and SSE are commonly employed in pharmaceutical manufacturing, their application in ophthalmicformulationsmaybelimitedwhenproducinghighlyintricatestructuresduetorelativelylarge nozzlediametersandtherequirementforextensive post-processing. To address these challenges, vat photopolymerization(VPP)-basedtechniquessuchas stereolithography (SLA) and digital light processing (DLP) have gained attention for ophthalmic device fabrication.Thesemethodsoffersuperiorresolution, smoothersurfacecharacteristics,andthinnerlayerformationcomparedwithextrusion-basedtechnologies. InVPPsystems,liquidphotopolymerizableresinisselectively cured by light exposure in a layer-by-layer fashiontoproducesolidstructures.SLAutilizesultravioletlaserlighttopolymerizespecificregionsofthe resin,whereasDLPprojectsanentireimagelayersimultaneously,enablingfasterprintingspeeds.Since printingoccursunder relatively mildthermal conditions,thesemethodsareparticularlysuitableforthermolabiledrugsandsensitivebiomolecules[78]. SLA technologyhasbeeninvestigatedforthefabrication ofmicroneedlemoldsandoculardeliverysystems.Fitaihiandcoworkersdesignedcone-shapedmicroneedleswithoptimizedaspectratiosforeffectiveocular tissuepenetration.Theirfindingsemphasizedthesignificanceofadjustingparameterssuchasorientation angleandlayerthicknesstoachieveaccurateandmechanically stable structures. In another study, SLAgeneratedmoldswereusedtopreparedissolvingmicroneedlescomposedofHPMCandpolyvinylpyrrolidone (PVP K90). Among the materials employed in SLAandDLPsystems,polyethyleneglycoldiacrylate (PEGDA) is one of the most widely used photopolymerizableagentsbecauseofitsfavorablebiocompatibility.XuandcolleaguesfabricateddexamethasoneloadedpunctalplugsusingDLPprintingwithPEGDA andPEG400.FormulationscontainingPEG400exhibiteddrugreleaseforapproximatelyoneweek,whereas plugscomposedsolelyofPEGDAsustainedreleasefor morethanthreeweeks. Thesefindingsindicatethe potentialofDLPprintingforproducingpersonalized punctalplugsfordryeyemanagement.[79]Gotoand collaborators further explored DLP printing to fabricateazithromycin-loadedcontactlensesusingPEGDAbased formulations. The lenses demonstrated sustaineddrugreleasealongwithantimicrobialactivity comparabletocommerciallyavailableeyedrops.VariationsintheratioofPEGDAtoPEG400significantly influencedswellingbehaviorandreleasekinetics.In another notable study, researchers developed DLPprintedcontactlensesintegratedwithnanoscalesurfacepatternsgeneratedbylaserablation.Thesemodificationsenabledthelensestofunctionaswearable biosensors for monitoring ocular parameters. Such advancementsdemonstratethegrowingpotentialof DLPprintingnotonlyinsustainedoculardrugdeliverybutalsointhedevelopmentofsmartophthalmic diagnosticsystems.[80]
technicalandregulatorybarrierscontinuetolimitits broaderimplementation.Challengesassociatedwith thistechnologyincludethelimitedavailabilityofsuitablepharmaceuticalexcipients,constraintsrelatedto softwareandprintingequipment,andconcernsregardingthemechanicalqualityofthefinalprintedproducts, suchasinadequatestrength,roughsurfacecharacteristics,andvariabilityinperformance.[81]Amongthe different3D-printingapproachesusedforocularformulationsanddrug-deliverydevices,semi-solidextrusion(SSE)andvatphotopolymerization(VPP)arecommonlyemployed. However,thepresenceofresidual solventsinmanyprintedformulationsremainsasignificantconcern.Inaddition,thedependenceonphotopolymerizingandcrosslinkingagents,manyofwhich havelimitedbiocompatibility,furtherrestrictstheclinical applicability of these systems. Therefore, there isacriticalneedtodevelopnovelexcipientsthatare safe,biodegradable,biocompatible,environmentally sustainable,andsuitableforocularuse. Anotherimportantlimitationisthatmanycurrentlyavailable3D printersarenotfullycompliantwithGoodManufacturingPractice(GMP)standards,raisingconcernsregardingthesafety,reproducibility,andqualityofproducts intendedforhumanuse.Highinstallationandoperationalcostsassociatedwithprintersandrawmaterials alsocreateeconomicchallenges,particularlyinhealthcaresettingswherelarge-scaleproductionmaystill offergreatercostefficiency.Toimproveproductreliabilityandmanufacturingconsistency,non-destructive analyticalapproachessuchasprocessanalyticaltechnologies(PATs)maybeincorporatedeitherduringor aftertheprintingprocessinsteadofrelyingsolelyon destructive testing methods. At the same time, robustquality-controlsystemsmustbeimplementedto ensurecompliancewithregulatoryexpectationsand productsafetystandards.[82]
Hot-Melt Extrusion: Hot-melt extrusion (HME) is a pharmaceutical manufacturing technique originally adapted from the plastics industry. The process involvesfeedingdrug(s)andexcipient(s)intoanextruder, whererotatingscrewsapplyheatandmechanicalshear tomelt,mix,andtransportthematerials.Themolten massisthenpushedthroughadie, cooled, andcollectedasanextrudate. Processingisusuallycarried outattemperaturesabovethepolymerglasstransition temperature(Tg)ormeltingtemperature(Tm). HME providesefficientmolecular-levelmixingbetweenAPIs and excipients, resulting in formulations with excellentcontentuniformityandimprovedproductquality. Thetechnologyiswidelyusedfordevelopingoral,ocular, topical, andmodified-releasedosageforms. It is particularly valuable for enhancing the solubility andbioavailabilityofpoorlywater-solubledrugsby producingamorphousdispersionsandco-crystalsystems.ThemajoradvantagesofHMEincludecontinuousmanufacturing,solvent-freeprocessing,scalability, automation,andhighreproducibility.Screwconfigurationscanalsobecustomizedtoachievetherequired mixingintensityfordifferentformulations.Inaddition, HME can be integrated with advanced technologies suchas3DprintingandmonitoredusingprocessanalyticaltoolslikeUV–VISandNIRspectroscopy.However,heat-sensitivebiomoleculessuchasproteinsand peptidesmaynotbesuitableforthisprocess,making preformulationstudiesessentialbeforeproductdevelopment.[83]
Injection molding is one of the
mostcommonlyemployedfabricationmethodsinthe medicaldevicesectorbecauseitenablestheproduction of highly precise and structurally complex systems with good manufacturing consistency. In oculardrugdelivery,thistechniquehasgainedconsiderableimportancefordevelopingimplantabledevices capableofprovidingprolongedandcontrolledrelease oftherapeuticagents. Bothbiodegradableandnonbiodegradable polymeric implants can be prepared usingthisapproachdependingontheintendedclinical application. During formulation development, several factors must be carefully optimized, includingbiocompatibilityofthematerials,preservationof drugintegrity,andthedesireddrugreleasebehavior. Biodegradablepolymerssuchaspoly(lactic-co-glycolic acid)(PLGA)arefrequentlyselectedduetotheirfavorablesafetyprofileandcontrolleddegradationcharacteristics[84]. Inaddition, theselectionofpolymers, excipients,anddrugincorporationstrategiesshould betailoredtotheocularenvironment,whichischaracterizedbyspecificphysiologicalconditionssuchas pHvariations,enzymaticactivity,andrapidclearance mechanisms. Injectionmoldingprovidesimportant advantagessuchaseaseoflarge-scaleproduction,reproducibility,andtheabilitytomanufactureimplants withuniformdimensions. Nevertheless,certainlimitationsremainassociatedwiththisprocess.Maintainingthestabilityofheat-sensitivedrugsduringmanufacturingisparticularlychallengingbecauseelevated processingtemperaturesmayleadtodegradationof theactivecompound. Therefore,parametersincludingmoldingtemperature,injectionpressure,andcoolingconditionsrequirecarefuloptimizationtomaintain bothdrugstabilityandpolymerintegrity,asthesefactorsdirectlyinfluencethefinalreleaseprofileofthe implant.Furthermore,theminiaturesizeandsophisticatedarchitectureofocularimplantsmakeitdifficult toachievehomogeneousdrugdistributionandprecise dosingthroughoutthedevice.[86] OcularDevices
Drug-loaded Timololmaleate contactlenses (1%w/w)
RhodamineB Microneedle (model Polylacticacid(PLA) systems compound)
Moldsfor PLAfilaments Galantamine dissolvable alongwith hydrobromide microneedles PVA/PVPblends
Ocular Levofloxacin methylcellulose patches (0.5%w/w) (HPMC),
Intracameral Timololmaleate Polycaprolactone implants (5–10%w/w) (PCL)
Ocular 5-Fluorouracil Polycaprolactone implants (1%w/w) andchitosan
Betamethasone Hydrogel sodium Polyethyleneiminescaffolds phosphate -basedhydrogel (2.5%w/w)
Dissolving Placebo (PVP)and microneedles formulation polyvinylalcohol
Moldsfor Placebo dissolvable Polylacticacid(PLA) formulation microneedles
Dexamethasone Punctalplugs PEGDAandPEG400 (10–20%w/w)
Thefabricatedlenses demonstratedaninitial Hot-meltextrusion rapidreleasephase Ethylenevinylacetate fuseddeposition followedbyacontrolled (EVA)and modeling releasepatternextending polylacticacid(PLA) (HME-FDM) uptothreedays,indicating suitabilityfor glaucomatherapy.
Theproducedmicroneedles werefoundsuitablefor surfacecoatingwith Fuseddeposition therapeuticagentsand modeling(FDM) mayservemultiple biomedicaland oculardeliverypurposes.
Customizedmoldsfabricated through Fuseddeposition FDMtechnologyenabled modeling(FDM) thepreparationof drug-incorporateddissolving microneedles.
Thepatchesexhibited antibacterial Hydroxypropyl potentialagainstocular Semi-solid infectionswiththe extrusion(SSE) majorityofdrug mannitol,andxylitol releaseoccurringwithin 1–2hours.
Long-actingimplantswere engineeredtomaintain Semi-solid sustaineddrugdelivery extrusion(SSE) fornearlyeightweeks, offeringpotentialbenefits inglaucomamanagement.
Theimplantdesignaimedto minimizeconjunctivalscarring Semi-solid andfibrosisfollowing extrusion(SSE) glaucoma-relatedsurgical procedures.
Bioprintedhydrogelscaffolds Semi-solid demonstratedpromisefor extrusion(SSE) localizedtreatmentofocular inflammatoryconditions.
Prototypedissolvable Polyvinylpyrrolidone microneedles Stereolithography weresuccessfullyproducedas (SLA) exploratorysystemsfor (PVA) ophthalmicdrugdelivery applications.
SLA-fabricatedmoldswere effectivelyutilizedfor Stereolithography manufacturingHPMC-and (SLA) PVPK90-baseddissolving microneedles.
Formulationspreparedentirely withPEGDAachieved Digitallight prolongedcorticosteroid processing(DLP) releaselasting beyondthreeweeksfordry eyemanagement.
Antimicrobial Azithromycin contactlenses (1%w/w)
Hollow Placebo microneedles formulation
Active Device/ Pharmaceutical Materials/ Delivery Ingredient ExcipientsUsed System (DrugLoading)
Ophthalmicdrugdevelopmentpresentsuniquechallengesbecauseoftheeye’sspecializedanatomy,complexphysiologicalbarriers,anddistinctroutesofdrug administration.Atthesametime,advancedimaging techniquesanddigitalbiomarkerscreateopportunitiesformoreprecisetherapeuticdevelopment.Artificialintelligence(AI)anddigitalhealthcaretechnologies areincreasinglytransformingthisfieldbyaccelerating drug discovery, improving prediction accuracy, and supportingprecisionmedicineapproachesforocular diseases.[87]
AI-BasedTargetDiscovery: Severaloculardisorders, including age-related macular degeneration (AMD), glaucoma, cataracts, and corneal dystrophies, are linkedtoproteinmisfoldingandabnormalcellularsignaling. Understanding the three-dimensional structureoftheseproteinsisessentialforidentifyingtherapeutictargets. ConventionaltechniquessuchasXraycrystallography,nuclearmagneticresonancespectroscopy, and cryo-electron microscopy provide detailedstructuralinformationbutareexpensive,laborintensive, and time-consuming. AI-driven computationaltoolshavesignificantlyimprovedthisprocess. DeepMind’sAlphaFoldplatformrevolutionizedprotein structurepredictionbyaccuratelyestimatingprotein conformations from amino acid sequences. The introductionofAlphaFold2andAlphaFold3furtherenhanced prediction accuracy and expanded capabilitiestoincludeinteractionsinvolvingDNA,RNA,and smallmolecules. Thesedevelopmentshaveacceleratedcomputer-aideddrugdesignbyrapidlyidentifying potentialbindingsitesandreducingthecostanddurationofearly-stagedrugdiscovery.AI-poweredmultiomicsplatformssuchasPandaOmicsalsosupporttargetidentificationbyintegratingtranscriptomics,pronotableantimicrobialefficacy Digitallight withinhibitionzones PEGDAandPEG400 processing(DLP) comparableto marketedophthalmic formulations.
Angled-printingstrategies produced Commercial Digitallight hollowmicroneedleswith biocompatible processing(DLP) improvedstructuralgeometry resins comparedwithflat-printed counterparts.
KeyOutcomesand PrintingTechnique PotentialApplications
teomics,andbiomedicalliteraturedatabases.These systemsprioritizenovelanddruggabletargetsmore efficientlythanconventionalapproaches. SimilarAI methodologieshavebeenappliedinpeptidediscoveryfrommicrobiomeandextinct-organismdatasets, openingnewpossibilitiesforantimicrobialtherapies andpeptide-basedoculartreatments.[88]
CADDandMolecularOptimization: Onceatherapeutictargetisidentified,computationaldrugdesigntools areusedtopredicthowcandidatemoleculesinteract withbiologicalsystems.Structure-baseddrugdesign reliesonthethree-dimensionalstructureoftargetproteins,whileligand-baseddrugdesignfocusesonthe propertiesofpreviouslyknownactivecompounds.AIenhancedmoleculardockingenablesrapidscreening ofmillionsofcompoundsbypredictingbindingaffinityandinteractionpatternswithtargetproteins.This approachhasshownpotentialinidentifyingtherapies targetingvascularendothelialgrowthfactor(VEGF),an importantmediatorinretinaldiseasessuchasAMD. Moleculardynamicssimulationsfurtherimprovepredictionaccuracybymodelingthebehaviorofdrugsand proteinsindynamicbiologicalenvironments.Machine learning-basedquantitativestructure-activityrelationship (QSAR) models are also widely used to predict therapeuticactivityandoptimizeleadcompounds.Recentdeeplearningapproaches,includinggenerativeAI systems,candesignentirelynewmoleculeswithdesirablepharmacologicalproperties.Thesetechnologies havesignificantlyshortenedthetimelinefordiscoveringnovelantimicrobialpeptidesandmaysimilarly accelerateoculardrugdevelopment.[89]
ExplainableAIinDrugDiscovery: Onelimitationof conventional AI systems is the lack of transparency in prediction mechanisms. Explainable AI (XAI) addressesthisissuebyimprovingtheinterpretabilityof AI-generatedoutputs. TechniquessuchasSHAPand LIME help researchers identify which molecular featurescontributemoststronglytodrug-targetinteractions.PlatformssuchasSmartCADDintegrateAIwith quantummechanics-basedapproachestoimprovevirtualscreeningandpredictdrugabsorption,distribution,metabolism,excretion,andtoxicity(ADMET)profiles more accurately. Explainable models are especiallyvaluableinophthalmologybecauseoculardiseasesinvolvehighlycomplexmolecularpathwaysrequiringreliableandinterpretablepredictions.[90]
ditional animal models often fail to fully replicate humanocularphysiologyandmayraiseethicaland translationalconcerns.Asalternatives,organoidsand organ-on-a-chip(OoC)systemshaveemergedasadvanced preclinical platforms. Retinal organoids derived from stem cells can mimic the structure and functionofhumanretinaltissue,whileOoCsystems recreatephysiologicalmicroenvironmentsusingmicrofluidictechnology.[91]RecentretinalOoCmodels have successfully reproduced photoreceptor-retinal pigmentepitheliuminteractionsandchoroidalneovascularization,enablingmorerealisticdrugscreeningandtoxicityevaluation. AIintegrationfurtherenhances these systems by automating image analysis,optimizingexperimentalconditions,andprocessing large datasets generated from high-throughput screeningplatforms. Deeplearningalgorithmshave also been applied for angiogenesis analysis, vessel segmentation, and phenotypic screening in retinal organoids. These technologies improve experimentalreproducibilityandmaysignificantlyacceleratethe discoveryoftherapiesforretinaldisordersandocular infections.[92]
AIinADMETPrediction: Physiologicallybasedpharmacokinetic(PBPK)modelingisincreasinglyusedto predicttheabsorption,distribution,metabolism,and eliminationofophthalmicdrugs.However,accurateocularPBPKmodelingischallengingbecauseoftheeye’s complexanatomy,multipledeliveryroutes,andphysiologicalbarrierssuchastearturnover,blinking,and cornealpermeability.AIoffersapromisingsolutionby extractingmeaningfulphysiologicalparametersfrom largedatasetsanddevelopingadaptivenonparametric modelscapableofpredictingcomplexdrugbehaviors. Machinelearningsystemscanestimatecornealpermeability,intraoculardistribution,andocularmetabolism moreaccuratelythantraditionalmethods.Advanced neuralnetworkmodels,includingNeural-ODEandlong short-termmemory(LSTM)architectures,arealsobeingexploredforoptimizingtreatmentregimensand improvingpharmacokineticpredictions.[93]
Recentprogressinhigh-throughputsequencingand single-cell technologies has transformed the understanding of how drugs interact with biological systems, creatingnewopportunitiesforprecisiontherapeutics. These advances enable researchers to studymolecularalterationsingreaterdepthandsupportadaptivepatientselectionstrategiesduringclinical trials. For example, whole-genome sequencing can classify patients according to genetic variations, helping clinicians tailor therapies more effectively andimprovetreatment outcomes. Unlike traditionalpharmacogenomics,whichmainlyexamines individual genetic mutations, artificial intelligence (AI) enables a broader systems-level evaluation of gene–geneandgene–environmentinteractionsthat influencetherapeuticresponses. [94]Growinginterestindrug–omicsassociationshasfurtherimproved theunderstandingofcomplexinteractionsbetween drugsandbiomolecules. Ophthalmologyisparticularlysuitedfortheseapproachesbecausetheeyecontains highly specific biomarkers, including proteins, nucleicacids,andmetabolitesfoundintears,aqueous humor,andvitreoushumor. Sincethesebiomarkers arelocallyconcentrated,theyprovideaclearerrepresentationoftheocularmicroenvironmentcompared withsystemicmarkers. Liquidbiopsytechniquesallow these biomarkers to be collected through minimallyinvasiveprocedures,offeringreal-timeinsights intodiseaseprogressionandtherapeuticresponse.For instance,complement-relatedmoleculesdetectedin aqueoushumorhavebeenstronglylinkedwithagerelated macular degeneration (AMD) and glaucoma, makingthempromisingtherapeutictargets.[95]Advances in computational biology have also encouragedtheuseofmulti-omicsapproachesintegrating epigenomic,transcriptomic,proteomic,lipidomic,and metabolomic data to better understand ocular diseases.Researchershavealreadydevelopedtranscriptomicandproteomicmapsforhealthyanddiseased retinal tissues, including diabetic retinopathy models. AI-drivenanalyticalsystemscanidentifybiologicalpathwaysassociatedwithretinalagingandtreatmentresponse,openingnewpossibilitiesfortargeted therapies.Proteomicbiomarkersarealsoincreasingly usedtomonitortherapeuticoutcomesinconditions suchasgeographicatrophy.[96]Onemajorchallenge inomics-basedresearchisthepresenceofconfoundingvariablesthatobscuremeaningfulbiologicalsignals.AI-basedframeworkssuchascontext-awaredeconfoundingautoencoders(CODE-AE)helpovercome thisissuebyseparatinggenuinebiomarkersfromtechnicalnoise,batcheffects,andpatientvariability.Such approachesimprovethepredictionofpatient-specific drugresponsesandstrengthenthereliabilityofprecisionmedicinemodels.Applyingthesetechnologiesin ophthalmologycouldsignificantlyenhancebiomarkerguideddiagnosisandtherapeutictargeting.[97]
Imagingtechnologiesplayacentralroleinophthalmic diagnosis and monitoring. Modalities such as color fundusphotography,fluoresceinangiography,optical coherencetomography(OCT),andOCTangiography generatedetailedstructuralandfunctionalinformation about the eye. Emerging imaging systems, includingvisible-lightOCT,multispectralimaging,and fluorescence-basedtechniques,provideevendeeper insightsintoretinalmetabolismandphysiology.AIalgorithmsgreatlyimprovetheinterpretationofthese imagingdatasetsbyrapidlyidentifyingdisease-related featuressuchasretinalthickness,vascularalterations, andstructuralabnormalities.Theseautomatedanalysesreducedependenceonmanualgrading,improve consistency,andaccelerateclinicaldecision-making. Recent advances in unsupervised learning and vision–languagemodelshavealsoenabledAIsystemsto identifysubtlepathologicalpatternswithoutrequiring predefined labels. Such systems can recognize disease-associatedmorphologicalchangesinconditionslikeAMD,sarcoidosis,andrareretinaldisorders thatmaynotbeeasilyvisibletoclinicians.AIalsosupportstreatmentmonitoringbydetectingearlystructural changes and predicting therapeutic response. Biomarkersextractedfromimagingdatacanhelpdeterminetreatmentfrequencyandidentifypatientsmost likelytobenefitfromspecificinterventions.Generative AI(GenAI)furthercontributesbysimulatinghypotheticalpost-treatmentimages,allowingcliniciansandpatientstovisualizepotentialoutcomesmoreeffectively. TheseAI-derivedmetricsmayeventuallyserveassecondaryendpointsinclinicaltrials,supportingfaster decision-makingduringdrugdevelopment. Inglaucoma research, for example, AI can evaluate retinal nervefiberlayerthicknessandotherneuralstructural changesratherthanrelyingsolelyonintraocularpressuremeasurements.Suchinsightsareparticularlyvaluableforthedevelopmentofneuroprotectivetherapies aimedatpreservingretinalandopticnervefunction. [98]
Patientrecruitmentandmonitoringremainmajorlimitationsinconventionalclinicaltrials,oftencontributingtohighfailurerates.Traditionalrecruitmentmethodsbasedonelectronichealthrecords(EHRs)andmanualscreeningcanoverlookeligibleparticipantsorincludeunsuitablecandidates.AI-drivenpredictiveanalyticsimprovesthisprocessbyanalyzinglargedatasets thatincludeclinicalhistory,laboratoryfindings,imagingresults,andlifestyle-relatedfactors.Thisstrategy, knownasclinicaltrialenrichment,increasesthelikelihoodofselectingappropriateparticipants.Inophthalmology, AIanalysisofOCTscanshasdemonstrated improvedaccuracyinidentifyingeligibleAMDtrialparticipantscomparedwithconventionalEHR-basedapproachesalone.AIalsofacilitatesdecentralizedclinicaltrials,allowingparticipantstocontributeremotely ratherthantravelingtocentralizedresearchsites.This isespeciallybeneficialforelderlypatientsaffectedby retinaldiseasessuchasAMDanddiabeticretinopathy. Anotherimportantadvancementisfederatedlearning, whichenablesmultipleclinicalcenterstocollaborate whilemaintainingpatientprivacy.Insteadoftransferringsensitivedata,federatedlearningallowsAImodels tolearnfromdistributeddatasets,resultinginstronger andmorediverseclinicalevidence.[99]AI-supported adaptive clinical trials further improve efficiency by dynamicallymodifyingstudyprotocolsaccordingto interimfindings.Unliketraditionalfixedtrialdesigns, adaptivesystemscanredirectresourcestowardmore promisingtherapeuticstrategiesanddiscontinueineffectiveinterventionsearlier.Automationofroutine taskssuchasadverse-eventreporting,patientfollowup,anddatamanagementalsoreducesworkloadand operationalcosts. Areal-worldexampleisthe“Vibe Up”study,anAI-drivendecentralizedtrialevaluating digitalinterventionsforpsychologicaldistress.Using adaptivealgorithms,participantallocationchangesin realtimeaccordingtoemergingevidence,improvingefficiencywhilereducingsamplesizerequirements.The introductionofAI-poweredDigitalTwinsrepresents anothertransformativeinnovation.DigitalTwinsare virtualpatientmodelscreatedusingreal-worldclinical, imaging,genomic,andbiomarkerdata.Thesemodels cangeneratesyntheticcontrolgroups,reducingthe needforlargeplaceboarmsandminimizingethical concernsassociatedwithwithholdingtreatment.RegulatoryagenciessuchastheEuropeanMedicinesAgency (EMA)havealreadyacknowledgedAI-supportedstatisticalframeworkslikeprognosticcovariateadjustment (PROCOVA),whichimprovetrialefficiencywhilemaintainingscientificrigor. DigitalTwinscouldintegrate multimodalimagingandmolecularbiomarkerstopredicttreatmentresponseanddiseaseprogression.Applicationsareemerginginproliferativevitreoretinopathyandglaucoma,wherevirtualpatientmodelsmay helpoptimizeindividualizedtreatmentstrategiesand reducedependenceonextensiveinvivotesting.[100] Wearabledevicesandsmartphone-basedAItoolsare alsoreshapingclinicalmonitoringbyenablingcontinuouspatientengagementandreal-timedatacollection.AI-assistedmobileapplicationscansupportoculardiseasemonitoring,improvetreatmentadherence, andstrengthencommunicationbetweenpatientsand healthcareproviders.Studieshavealreadyshownthat AI-supportedcaremodelsimproveself-management andfollow-upcomplianceindiabeticretinopathypatients.
Developinganewdrugiscostly,time-intensive,and associated with high failure rates. Drug repurposingoffersanattractivealternativebyidentifyingnew therapeuticapplicationsforalreadyapprovedmedications.Sincethesafetyprofilesofthesedrugsarealreadyestablished,repurposingreducesdevelopment risks,lowerscosts,andacceleratesclinicalavailability. Theriseofbigdataandglobaldata-sharingplatformshasmadeAIincreasinglyvaluableinidentifying repurposingopportunities.Naturallanguageprocessing(NLP)anddeeplearningalgorithmscananalyze enormousvolumesofscientificliterature,clinicaltrial data, and EHRs to uncover hidden therapeutic relationships.DuringtheCOVID-19pandemic,AIsystems identifiedremdesivir,originallydevelopedforEbola, as a potential antiviral therapy against SARS-CoV-2, demonstratingthepracticalvalueofAI-drivenrepurposing.RecentdevelopmentsinGenAIandlargelanguagemodels(LLMs)haveexpandedthesecapabilities further. Thesemodelscanprocesslargeamountsof unstructuredbiomedicalinformationanddetectnonobviousdrug–diseaseassociations.Researchershave alreadyexploredAI-guidedrepurposingstrategiesfor neurologicaldiseasessuchasAlzheimer’sdisease,and similarapplicationsareexpectedtoemergerapidlyin ophthalmology.[101]
Topical and locally administered ophthalmic drug products are categorized by the United States Food and Drug Administration (FDA) as complex formulationsbecauseofthesignificantchallengesinvolved intheirchemistry,manufacturing,andcontrol(CMC) processescomparedwithconventionaldosageforms. Over the years, the FDA has actively supported researchinitiativesthroughgrantsandcollaborativeprograms involving academic institutions and contract researchorganizationsworldwidetoaddressregulatory and scientific hurdles associated with complex genericophthalmicproducts. Theseinitiativeshave focusedonavarietyofophthalmicdosageforms,includingemulsions,ointments,inserts,andimplants. Outcomesfromthesecollaborativeeffortshavecontributedtoscientificpublications,strengthenedthe regulatoryjustificationforabbreviatednewdrugapplications(ANDAs)andnewdrugapplications(NDAs), supported the revision and development of regulatory guidelines, and accelerated the approval pathway for complex generic ophthalmic products. Recently, the FDA introduced the draft guidance titled “QualityConsiderationsforTopicalOphthalmicDrug Products”in2023.Inparallel,severalproduct-specific guidance documents have also been updated or releasedtoprovidemoredetailedrecommendationsfor theformulationdevelopment,characterization,and evaluationofophthalmicproducts.Therefore,reviewingbothgeneralandproduct-specificguidancedocumentsisessentialfordeveloperstoremainaligned with current regulatory expectations. [102] Despite considerable progress in ophthalmic product development, establishing reliable in vitro–in vivo correlations(IVIVCs)remainsamajorscientificchallenge. TheFDA’soriginalIVIVCguidance,publishedin1997 forextended-releaseoralformulations,laidthefoundation for adapting IVIVC concepts to other dosage forms. Although regulatory agencies do not universallyrequireIVIVCstudiesforsubmissions,establishingmeaningfulcorrelationscangreatlybenefitformulationoptimizationandproductevaluation. Among theavailableapproaches,LevelAIVIVC,whichrepresentsapoint-to-pointrelationshipbetweeninvitroreleaseandinvivoperformance,isconsideredthemost informativeandisgenerallypreferredbyregulatory authorities. However,generatingLevelAIVIVCmodelsforophthalmicsystemsisconsiderablymoredifficultthanforconventionaloraldosageformsdueto the anatomical and physiological complexity of the eye.SuccessfulIVIVCdevelopmentdependsonseveral criticalfactors,includingthepredictivecapabilityof theinvitroreleasemethod,therelevanceofpharmacokineticprofilesindifferentoculartissuesfollowing administration,andtheselectionofappropriatemodelingapproachessuchasnumericaldeconvolutionor physiologicallybasedpharmacokinetic(PBPK)models. Anotherlimitationisthepracticaldifficultyofmeasuringdrugconcentrationsdirectlywithinvariouscompartmentsofthehumaneye. Consequently,animal models,particularlyrabbits,arefrequentlyexplored fordevelopingpreclinicalIVIVCrelationshipsdespite knowninterspeciestranslationallimitations.[103]Artificialintelligence(AI)isincreasinglytransformingdrug discoveryanddevelopment;however,itsintegration intopharmaceuticalresearchalsointroducesseveral scientific,ethical,andregulatoryconcerns.AI-driven systemsrequireaccesstoextensivedatasetsformodel training,raisingimportantquestionsregardingdata privacy,security,andpatientconfidentiality.Sharing healthcare-related data across institutions or countriesmayincreasetheriskofunauthorizedexposure ofsensitiveinformation. Toaddresstheseconcerns, compliancewithregulationssuchastheHealthInsurance Portability and Accountability Act (HIPAA) and the General Data Protection Regulation (GDPR) has becomeincreasinglyimportant. Federatedlearning hasemergedasapromisingstrategytoimprovedata privacywhileenablingcollaborativeAImodeldevelopment. In this approach, data remain within their originalinstitutions,whileonlymodelparametersare sharedwithacentralizedserverforaggregationand optimization. Thisdecentralizedtrainingframework minimizesdirectdataexposurewhilestillsupporting thedevelopmentofhigh-performingmachinelearningmodels.Federatedlearninghasshownpotentialin advancingAI-assisteddrugdiscoveryplatforms,includingtechnologiessuchasAlphaFoldandorgan-on-chip systems.Regulatoryagenciesworldwideareactively attemptingtoadaptexistingframeworkstothegrowinginfluenceofAIinhealthcareandpharmaceutical innovation.TheFDAhasreviewednumeroussubmissionsinvolvingAI-basedmethodologiesand,inMarch 2024,introducedarisk-basedframeworkemphasizing fourmajorpriorities:encouragingcollaborationtoprotectpublichealth,promotingharmonizedstandards andbestpractices,supportinginnovation-friendlyregulatoryapproaches,andadvancingresearchforevaluatingandmonitoringAIperformance.Similarly,organizationssuchastheEuropeanMedicinesAgency(EMA) and the Medicines and Healthcare products RegulatoryAgency(MHRA)areworkingtowardstandardized protocolsforAIandmachinelearningapplicationsin drugdevelopment.Nevertheless,therapidevolution ofgenerativeAIandlargelanguagemodels(LLMs)continuestochallengeexistingregulatorysystems,which currentlylackcomprehensiveprovisionsaddressing unpredictableAI-generatedoutputs.RegulatoryoversightmayalsoneedtoexpandtowardAI-supported biomarkeranalysisinocularfluidssuchasaqueous humorandvitreoushumortoensureethicalandscientificallyreliableimplementation. Atpresent,thereis nouniversallyacceptedinternationalregulatoryframeworkgoverningAIindrugdiscovery,highlightingthe needforcoordinatedglobalstandardsandharmonized oversightstrategies.Furthermore,althoughlife-cycle managementapproachesforcontinuouspost-market monitoringofAIsystemshavebeenproposed,practicalimplementationremainscomplex. [104]TheincreasinguseofAIinpharmaceuticalinnovationhas alsogenerateddebateregardingintellectualproperty rightsandpatenteligibility.IntheUnitedStates,patent lawsrecognizeinventionsthatrepresenta“newand usefulprocess”;however,algorithmsthemselvesare oftenregardedasabstractconceptsandmaynotqualifyforpatentprotection.Additionaluncertaintyarises fromlegislationsuchasthePatientEligibilityRestorationActof2023,whichexcludesmathematicalformulasandmentalprocessesfrompatentablesubjectmatter[105].Questionshavealsoemergedconcerningthe originalityandinventorshipofAI-assisteddiscoveries. CurrentlegalinterpretationsundertheFederalCircuit requireinventorstobenaturalpersons, meaningAI systemscannotbelistedasinventorsonpatentapplications.PatentapplicationsinvolvingAItechnologies mayalsoencounterchallengesrelatedtoenablement andwrittendescriptionrequirementsunderSection 112, particularlywhenvaguereferencestomachine learningorneuralnetworksareprovidedwithoutadequatemethodologicaldetail.Toreducesuchissues, patentspecificationsshouldprovidecomprehensive, clear,andtechnicallydetaileddescriptionsthatfully explaintheinventionanditsoperationalmethodology. [106]
The development of effective ophthalmic therapies continuestofacesignificantscientific,technological, andregulatorychallengesduetothecomplexanatomy andprotectivebarriersoftheeye.Conventionalocular dosageformsoftenfailtoachieveadequatetherapeuticconcentrations,particularlyinposteriorsegment diseases,highlightingtheneedformoreadvancedand targeteddeliveryapproaches.Recentprogressinocularimplants,nanoparticles,liposomes,hydrogels,microneedles,nanomicelles,iontophoresis,andinsitu gellingsystemsdemonstratesthegrowingpotential ofnext-generationophthalmicdrugdeliverytechnologiestoimprovebioavailability,prolongdrugresidence time,andreducedosingfrequency.Inparallel,additive manufacturingtechniquessuchasthree-dimensional printingandhot-meltextrusionareopeningnewpossibilitiesforpersonalizedoculartherapiesandpatientspecificmedicaldevices.Artificialintelligencehasfurther accelerated innovation in ophthalmic drug discoverybyenablingrapidtargetidentification,molecularoptimization,predictivemodeling,andbiomarker analysis. Emerging AI-driven tools, combined with digitalhealthcaretechnologies,maysubstantiallyreducedevelopmenttimelinesandimprovetherapeutic precisioninocularmedicine.However,issuesrelated tomanufacturingscalability,regulatorycompliance, long-termsafety,dataprivacy, andethicalconcerns remainimportantobstaclestoclinicaltranslation.Futureprogressinophthalmictherapeuticswilldepend onmultidisciplinarycollaborationamongpharmaceuticalscientists, clinicians, engineers, computational researchers,andregulatoryauthorities.Continuedadvancements in both drug delivery technologies and AI-drivenpharmaceuticaldevelopmentareexpected totransformthemanagementofoculardiseasesand improveglobalvisualhealthoutcomes.
AuthorshavedeclaredthattheyhavenoknowncompetingfinancialinterestsORnon-financialinterestsOR personal relationships that could have appeared to influencetheworkreportedinthispaper.