Design and Development of an In Vitro Tear Replenishment System

SAMAN MOHAMMADI,1 CAMERON POSTNIKOFF,1 ANN M. WRIGHT,2 and MAUD GORBET1,3

1Systems Design Engineering, University of Waterloo, 200 University Ave., Waterloo, ON, Canada; 2Alcon/CIBAVision,Duluth, GA, USA; and 3School of Optometry and Vision Science, University of Waterloo, Waterloo, ON, Canada

(Received 20 December 2013; accepted 26 May 2014)

Associate Editor Estefanıa Pena oversaw the review of this article.

Abstract—Understanding the cellular and molecular mech-anisms of the corneal tissue and translating them intoeffective therapies requires organotypic culture systems thatcan better model the physiological conditions of the front ofthe eye. Human corneal in vitro models currently exist,however, the lack of tear replenishment limits corneal in vitromodels’ ability to accurately simulate the physiologicalenvironment of the human cornea. The tear replenishmentsystem (TRS), a micro-fluidic device, was developed to mimicthe in vivo tear replenishment in the human eye in an in vitrocorneal model. The TRS is capable of generating adjustableintermittent flow from 0.1 lL in every cycle. The TRS is asterilizable device that is designed to fit standard 6-well cellculture plates. Experiments with the corneal models demon-strated that exposure to the TRS did not damage theintegrity of the stratified cell culture. Contact lenses ‘‘worn’’by the in vitro corneal model also remained moist at all timesand the cytotoxicity of BAK could also be verified using thismodel. These in vitro results confirmed that the TRS presentsnovel avenues to assess lens-solution biocompatibility anddrug delivery systems in a physiologically relevant milieu.

Keywords—Tear replenishment, Artificial tear solution,

In vitro cornea model, Micro-fluidic system, Ophthalmic

biomaterial.

INTRODUCTION

The cornea is the eye’s ‘‘window’’ to the outsideworld and consists of three main cellular layers: theepithelium, the fibroblastic stroma, and the endothe-lium. The corneal epithelium is a stratified structurethat contributes to both the visual quality and theprotective barrier functions of the eye. Contact lensesinteract directly with the corneal epithelium and rep-resent the most widely used medical device (withan estimated 125 million wearers worldwide27). It is

recognized that, as with any biomaterial, the presenceof an ophthalmic biomaterial will alter the physiolog-ical balance of the ocular environment, which may alsoaffect cells in the anterior eye.16 In the past decade,clinical and epidemiological studies have indicated thatlens wear as well the interaction between the contactlens material and the solution used to clean the lensescan lead to biocompatibility problems.6,7,17 With thecurrently available technologies (slit lamp, confocalmicroscopy and optical coherence tomography), it canbe difficult to evaluate the cellular effects that contactlens materials and solutions have on the human corneain vivo as the detailed biochemical changes that occurwithin these cells cannot be assessed. To evaluatebiochemical changes in the cornea, animal experimentswith rabbits and rats wearing contact lenses have beenperformed to assess biocompatibility or cytotoxicitywithin the physiological ocular environment. However,such in vivo studies are often costly, involve the sacri-fice of animals, and may be difficult to interpret givenboth interspecies variation and animal-to-animalinconsistency.3,8,9,14,18,23,25,32–35,40–42 With the increas-ing interest in using contact lenses as drug delivery andsensing devices8,23,34,41 and the requirement to betterunderstand the complex interactions between lens caresolutions and contact lenses, there is an increased needfor in vitro corneal models that can better mimic thedynamic ocular environment to more effectively assessbiocompatibility in vitro.

Human corneal in vitro models are recognized ascost effective substitutes compared to in vivo studies33

and have been used previously to study in vitro oculartoxicity.14,32 Corneal epithelium models have beenshown to correlate well with in vivo results.40 Fur-thermore, it is recognized that the corneal epitheliumacts as the primary barrier and is responsible for 99%of corneal electrical resistance,18,25 it is also the tis-sue that interacts most directly with a contact lens

Address correspondence to Maud Gorbet, Systems Design

Engineering, University of Waterloo, 200 University Ave., Waterloo,

ON, Canada. Electronic mail: mgorbet@uwaterloo.ca

Annals of Biomedical Engineering (� 2014)

DOI: 10.1007/s10439-014-1045-1

� 2014 Biomedical Engineering Society

material. Thus, the corneal epithelium alone can beconsidered as a good proxy for an in vitro cornealmodel as it can further reduce cell culture time andallow for higher throughput testing of cell-biomaterialinteractions and drug cytotoxicity studies.14 In vitroocular toxicity of lens care products and ophthalmiceye drops have been recently tested on static cell cul-ture models of the corneal epithelium either using amonolayer of primary or immortalized human cornealepithelial cells (HCEC)2,5 or using a stratified cultureof corneal epithelial cells.4,21,29 In an effort to bettermimic the in vivo situation, recent studies have alsoincluded solution-soaked contact lenses in the in vitrocell model.11,20,28 While the latter studies consider theinteraction of the corneal epithelium with the oph-thalmic materials, similarly to the solution studies,they do not take into account the fluid exchange thatoccurs with ‘‘tear replenishment’’.

A human eye holds approximately 7–30 lL of tearswith a turnover rate of 0.5–2.2 lL/min.1,24,25 Sponta-neous blinking happens at the rate of 6–15 times/minand helps to mechanically spread the tear film evenlyacross the ocular surface and to remove any foreignobjects in contact with the eye.25 To keep the cornealmodel hydrated, corneal epithelial models are con-ventionally immersed in cell culture medium or artifi-cial tear solution. However, the immersion cell culturepoorly models the tear replenishment occurring in thehuman eye via the lacrimal system. To simulateblinking, a rocker mechanism has previously been usedto generate intermittent flow on top of the cell culturemodel,13 however this model did not allow forreplenishment of fluid. Geerling et al.10 also used dif-ferent incubation times and concentrations to accountfor the dilution that can occur in the eye. These dilu-tions, though, are merely an approximation of thein vivo conditions and poorly simulate the releasemechanism from drug delivery or solution-soakedcontact lens materials. To more closely mimic the dy-namic fluid environment that the corneal epithelium isexposed to, we present a prototype for a tear replen-ishment system (TRS): a dynamic in vitro environment,whereby artificial tear solution regularly wets the sur-face of a contact lens resting on the curved multilayerepithelial model while fluid accumulation is preventedby automated timed aspiration. Following design andimplementation of the TRS, a series of experimentswere performed to confirm the ability of the in vitrotear replenishment device to hydrate the topical surfaceof a cornea epithelial model without damaging it. Tofurther test the proof of concept of the potential effectthat replenishment may have on mechanisms ofrelease, lenses soaked in a known cytotoxic compoundwere also tested and results compared to the conven-tional immersion method.

A MICRO-FLUIDIC APPROACH TO TEAR

REPLENISHMENT

Mimicking tear replenishment in the eye requiresthe delivery of a tear analog to the surface of thein vitro corneal model within the range of a physi-ological tear flow rate and in a recurring fashion.Our objectives were to cover the surface of thecorneal model with a small amount of artificial tearliquid that is sufficient to hydrate the surface of themodel while maintaining the air–liquid interface andto remove liquid to prevent full immersion of themodel.

To bring the artificial tear liquid to the surface ofthe in vitro corneal model, the injection of liquidthrough a needle or a nozzle was first considered. Toachieve a physiological tear flow rate, the size of theneedle needs to be smaller than 10 lm. Such a needlesize would be prone to clogging thus this was not asuitable solution. On the other hand, spraying a spe-cific amount of liquid at a frequency similar to humanblinking can be used to recreate the effect of tearreplenishment and, if designed properly, will also allowthe entire surface of the in vitro corneal model to bemoistened. Spraying occurs when the cohesive anddisruptive forces applied to the surface of a cylindricaljet of liquid generates small perturbations which, underfavorable conditions, can amplify to disintegrate the jetinto small droplets.19

At small flow rates (similar to the physiological tearflow rate), the liquid jet will breakup after traveling thedistance L, referred to as ‘‘Rayleigh varicose breakuplength’’, according to the empirical equation shownbelow:

L ¼ 19:5d0ð1þ 3OhÞ0:85ffiffiffiffiffiffiffi

Wep

in which Oh, Ohnesorge number is defined as below:

Oh ¼ lffiffiffiffiffiffiffiffiffiffi

qrd0p ¼

ffiffiffiffiffiffiffi

Wep

Re

where We and Re are the Weber and Reynolds num-bers and d0 is the diameter of the nozzle.

Using the fluid properties of tears presented in Ta-ble 1, the Rayleigh varicose break-up length will be asshort as a few micrometers for a nozzle diameter of50 lm, which implies that the liquid jet will aerosolizeas soon as exiting the nozzle. Although properties of

TABLE 1. Tear fluid properties.

Symbol Parameter Value

q Tear density 1 g/mL

l Tear viscosity 4.4–8.3 mPa s37

r Surface tension against the air 0.043 N/m38

MOHAMMADI et al.

any artificial tear solution will likely be slightly dif-ferent from human tears,22 such differences will notaffect the ability of the solution to aerosolize as soon asit exits the nozzle.

Based on these calculations, the micro-fluidic systemdescribed in Fig. 1 is proposed for use as a TRS. Thesystem uses a pressurized supply line from which anartificial tear solution is sprayed through a nozzle ontothe surface of the in vitro corneal model. The amountof sprayed solution is controlled through a series ofsolenoid isolation valves by adjusting the fraction oftime that each valve is open. The air–liquid interface ismaintained by collecting the sprayed solution anddraining it from each well.

To assess the feasibility of the microfluidic system’sconcept, mathematical modeling of the proposed sys-tem was performed and indicated that the desired flowrate could be reached if each isolation valve wasopened for the adjustable period of 25–1000 ms. Basedon the solved numerical model, the accumulativetransferred volume in each well over a period of 1 minis 2.2 lL. This volume corresponds well with thereported volume of 0.5–2.2 lL/min that is spread overthe ocular surface (Fig. 2).1,24,25

MECHANICAL DESIGN

Based on the modeling analysis, the designed systemneeds to provide an adjustable amount of tear fluid of1 lL/min or above, with the frequency of 10 times aminute. While Fig. 1 provides a schematic of the

micro-fluidic system, Figs. 3 and 4 present in moredetail the implemented solution. A solenoid operatedmicro-pump from Bio-Chem FluidicsTM is used topump the artificial tear liquid through TeflonTM tubing(BIO-CHEM Fluidics, NJ, USA). A custom-builtmanifold with a channel width of 1/16¢¢ was machinedfrom polycarbonate plates. The artificial tear solutionis pumped through the manifold and returns to theartificial tear solution reservoirs after passing througha pressure relief valve that maintains the pressure of 20psi inside the channels. The pressurized sterile artificialtear solution passes through the channels and thesolenoid operated isolation valves (BIO-CHEM Flu-idics, NJ, USA) and ceramic nozzles (CoorsTek�,

FIGURE 1. Schematic of the micro-fluidic system: a solenoid pump provides enough back pressure to a series of isolation valvesto control the flow rate as well as the period of time that the artificial tear liquid is being sprayed over the curved multilayer model.

FIGURE 2. Numerical solution of the accumulative trans-ferred volume in a well over one minute period using a 250 msopening time of the isolation valve.

In Vitro Tear Replenishment Model

USA) before being sprayed on the surface of the cor-neal model and then drained using needles connectedthrough another solenoid-operated micro-pump. Allthe components in the device are sterilizable and bio-logically inert. The device is also designed to fit in a cellculture incubator (37 �C, 5% CO2).

EXPERIMENTAL

To ensure the validity of this novel in vitro system,the ability of the in vitro tear replenishment device to

hydrate the topical surface of a cornea epithelial modelwith and without contact lenses was tested and com-pared to the conventional immersion method.

In Vitro Cell Culture

Immortalized HCECs were cultured in keratino-cyte serum free medium supplemented with bovinepituitary extract, recombinant epidermal growthfactor, and Penicillin/Streptomycin (Pen/Strep)(KSFM) (ScienCell, Carlsbad, California) at 37 �Cand 5% carbon dioxide (CO2). Fresh medium was

FIGURE 3. (a) Exploded view of the TRS and its various components. (b) Schematic cross section of the TRS with the nozzleassembly and the curved multilayer model.

MOHAMMADI et al.

added every other day and cells were grown to 90%confluency in tissue culture treated flasks. Adherentcells were removed using TryplExpress (Life Tech-nologies, Burlington, Ontario, Canada) dissociationsolution. Cells were routinely observed for anymorphological changes and were used before theireleventh passage.

The stratified curved corneal epithelia were pre-pared based on a previously developed protocol.30 Inbrief, 30 mm cellulose inserts were permanently de-formed into a curve by a die with the dimension andcurvature of an average human cornea. HCEC werecultured on the curved inserts in a keratinocyte serum-free medium supplemented with growth factors. Onday seven, cells were differentiated into a stratifiedmultilayer by exposing the monolayer to an air–liquidinterface. At this time, cells were fed only on thebasal side with 2% FBS in 1:1 DMEM/F12 andmedium was exchanged daily. The cells grew underthese conditions for 7 days and were then ready forexperiments.

Preparation of Contact Lenses

Two commercially available contact lens materials,balafilcon A and etafilcon A were used in this studyand their properties are listed in Table 2. Lenses hadback vertex power of 23.00. Phosphate buffered sal-ine, previously shown to be biocompatible,12 was usedas the negative control lens solution. A sterile oph-thalmic solution of benzalkonium chloride (BAK) wasused as the positive control: BAK is a preservativeused in several commercially available eye drop for-mulations and is a well-known irritant.29 The com-mercially available Moisture Eyes (ME; Bausch &Lomb, Rochester, NY, USA) has a BAK concentra-tion of 0.01% w/v. Contact lenses were removed fromtheir packaging and soaked for 24 h either in Phos-phate Buffered Saline (PBS, Lonza, Allendale, NewJersey) or in the undiluted BAK-containing solution.Each lens was totally immersed in the solutions in asterile 12-well polystyrene plate and all procedureswere performed under sterile conditions.

FIGURE 4. The TRS. (a) Polycarbonate manifold, nozzles and culture plate can be identified. (b) A cell culture plate with in vitrocorneal models, the corneal models appear dark purple as they have been stained with a metabolic dye following experiments inthe TRS; (c) Closed TRS: the system is built to ensure that testing with the in vitro corneal model is performed under asepticconditions. (d) The TRS is designed to fit in a cell culture incubator so that cells are exposed to the proper growing environment.

In Vitro Tear Replenishment Model

Tear Replenishment Experiments

The microfluidics system was sterilized by running70% ethanol through the device followed by a rinsewith PBS. The in vitro corneal models grown in 6-wellplates were transferred into the device in a sterileenvironment. Cells were fed with fresh KSFM on thebasal side of the cell culture inserts. Apically, KSFMwas then added to wet the surface prior to lens place-ment. BAK-soaked and PBS-soaked contact lenseswere then placed concave-side down on top of themultilayer. The contact lenses, onlayed on the curvedcorneal models, covered the entire surface of the mul-tilayer cornea model, in a fashion similar to how a lenswould contact the human corneal epithelium. Controlin vitro corneal models only received KSFM to wet thesurface. The entire device was then placed in a cellculture incubator (37 �C, 5% CO2). In each experimentreported, six corneal models were tested for 2 h, 3 weresubjected to tear replenishment conditions (spray withkeratinocyte serum-free medium as an artificial tearsolution) and 3 were simply immersed in medium (no-spray/tear replenishment conditions; it was necessaryto immerse the cornea model + lens in medium toprevent the contact lens from drying). In the ‘‘spray’’well, where tear replenishment took place, as soon asthe device was started, aspiration occurred and thusthe medium that had been added to allow lens place-ment was immediately removed, allowing the cell-contact lens model to only be exposed to the inter-mittent flow/spray.

After 2 and 6 h of incubation, lenses and mediumwere removed. To assess the effect of tear replenish-ment on cells, viability of the multilayer corneal modelwas measured by a modified MTT viability assay.30

Briefly, dimethyl thiazoyl blue tetrazolium bromide(0.5 mg/mL, MTT, Sigma Aldrich, Oakville, ON,Canada) was added to the apical and basal sides of thecell culture insert and was incubated for 3 h at 37 �Cand 5% CO2. The MTT solution was then removedand isopropanol was added to both the apical andbasal sides of the insert and plates were agitated for2 h. The solutions in the apical and basal sides weremixed together and samples were read in a UV–Vis

spectrophotometer at optical density of 595 nm with areference at 650 nm. All results are expressed as therelative viability compared to control cells: cells incu-bated in the absence of a contact lens and with nospray (immersion model).

All experiments were performed on different days.Results are reported as the mean of the experi-ments ± SD. To evaluate the significance of the dif-ferences in cell viability, an analysis of variance(ANOVA) was performed, followed by multiple pair-wise comparisons using the Holm-Sidak test and apairwise comparison using T tests according to Sidakcorrection of Bonferroni inequality in SigmaPlotTM.All lens combinations were tested at least three timesand controls (no spray/no lens) were run each time.

Experimental Results

Exposure to replenishment conditions for up to 6 h,in the absence of a lens, did not affect the cornealmultilayer model as shown by a viability of 96 ± 11%compared to control (no lens, no spray) (p> 0.5).Furthermore, no difference in viability was observedwith the multilayer exposed to PBS-soaked lenses withor without replenishment (Table 3).

As a means to identify a positive control and to gaina better understanding of the potential effect ofreplenishment, BAK-soaked lenses were tested andused with the TRS to determine if the release of acytotoxic compound from a contact lens on the cornealmultilayers would be affected by the replenishmentconditions. While no significant differences wereobserved at 2 h (Table 3), following a 6 h exposure, asignificant reduction in viability was observed withBAK-soaked Etafilcon A under both replenishmentand static immersion conditions (Fig. 5). Furthermore,tear replenishment had a significant effect on viabilitywhereby exposure to BAK-soaked Etafilcon A in adynamic fluid-exchange system resulted in a signifi-cantly lower viability when compared to static/noreplenishment conditions (p = 0.03). With BAK-soaked balafilcon A, a reduction in cell viability, albeitnot statistically significant when compared to the

TABLE 2. Properties of the contact lens materials used in the TRS study.

Commercial name Acuvue 2 Purevision

(US adopted name) Etafilcon A Balafilcon A

Manufacturer Johnson & Johnson Bausch & Lomb

Water content 58 36

Center thickness at (23.0 D) in mm 0.084 0.09

Principal monomer HEMA + MA NVP + TPVC +

NVA + PBVC

FDA Group IV

High water content, ionic

III

Low water content, ionic

MOHAMMADI et al.

no-lens and PBS-soaked Balafilcon A control (p> .1),was observed and no effect of tear replenishment wasobserved (p> 0.75).

DISCUSSION

Results from both the no-lens control samples andthe PBS-soaked lenses demonstrate that exposure to thespray from the TRS does not damage the multilayerand thus the TRS can be safely used to assess lens-cellinteractions in a dynamic system. While BAK toxicityhas been reported before, in vitro models have mostlybeen used to investigate the cytotoxicity of BAK solu-tion alone.2,39 To our knowledge, this is the first timethat HCEC are exposed to BAK-releasing materials.The lack of toxicity observed with BAK-soaked bala-filcon A lenses compared to BAK-soaked etafilcon Ahighlights the difference in material properties and theensuing difference in the mechanisms of BAK uptakeand release. Similar results have been observed previ-ously with differential uptake and release profiles of

lens care solutions and ophthalmic drugs with differentcontact lens materials.11,15,26,31

In a dynamic system, if one were to compare with astatic incubation in vitro model, two different mecha-nisms may occur: (1) the in vitro cytotoxic effect maynot be as pronounced compared to the static incuba-tion model; this would be due to the dilution of thetoxic substance, not accumulating, and being washedaway and removed by the TRS, as it would be in tears,or (2) cytotoxicity may increase due to an increase inconcentration of the toxic substance being releasedinto a much smaller residual artificial tear liquid,especially between the contact lens and the cornealcells, as it might be the case in tear film between theepithelium and the contact lens. The latter mechanismwas confirmed by our results from a 6 h exposure toBAK-soaked Etafilcon A whereby a statistically sig-nificant difference in cell viability between replenish-ment vs. no replenishment was observed. Interestingly,the BAK-soaked balafilcon A lens did not act as astrong cytotoxic stimulus to cells, which in turn made itdifficult to identify differences between dynamic vs.

FIGURE 5. Effect of tear replenishment vs. immersion conditions on corneal epithelial cell viability after 6 h incubation in TRSdevice. Viability was measured by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay and is expressed as apercentage relative to in vitro corneal model without contact lens and spray. n 5 3–4 6 SD. *Significantly different from no lensand PBS-soaked lens control, p < 0.01; #Significantly different from no spray/no tear replenishment sample, p 5 0.03.

TABLE 3. Effect of tear replenishment on viability of cells exposed to PBS- and BAK-soaked lenses for 2 h.

Without tear

replenishmenta viability (%)

With tear replenishment

viability (%)

p value, test for with vs.

without replenishment

Balafilcon A-PBS 106 ± 5 100 ± 10 1.00

Balafilcon A-BAK 86 ± 10 86 ± 5 0.99

Etafilcon A-PBS 101 ± 14 113 ± 15 1.00

Etafilcon A-BAK 80 ± 15 76 ± 18 0.54

n = 3 to 4 ± SD.aCells-contact lens system was fully immersed in solution for 2 h.

In Vitro Tear Replenishment Model

static conditions. The surface treatment of balafilcon Aassociated with the amphiphilicity of BAK is likelyresponsible for the limited uptake (and then release ofBAK),31,36 which makes this BAK-lens combination oflimited use when investigating the effects of tearreplenishment on cytotoxicity.

Our experiments demonstrated that the imple-mented design to mimic tear replenishment can deliverartificial tear fluid to the multilayered corneal epithelialmodels at a flow rate and frequency that are similar tothe human eye. Due to the static nature of the currentin vitro models, accumulation of the toxic agents in theculture medium may occur; thus, the release profilemay also be affected by the lack of replenishment. Inthe dynamic system presented with the tear replenish-ment, we demonstrated that replenishment can beachieved without damaging the integrity of the strati-fied cell culture, that the spray conditions allowed thecontact lenses to remain moist at all times withoutrequiring immersion in solution and that tear replen-ishment may have significant effect on measured out-comes (here cytotoxicity).

Through modeling the microfluidics of the TRS, thisnovel method will contribute to a better understandingof corneal cell-lens interactions in vitro. In the future,we hope to use this method to better understand howthe eye interacts with ophthalmic materials and solu-tions such as contact lenses, lens cleaning solutions,and drug delivery platforms. The dynamic environ-ment provided by tear replenishment impacts ocularinflammation, lens cleaning solution biocompatibilityand fluorescein staining, as well as drug delivery to theeye; this model provides one of the first approaches tostudy these effects in vitro in a more physiologicallyrelevant system.

ACKNOWLEDGMENTS

The funding for this project was provided by theNatural Sciences and Engineering Research Council ofCanada and CibaVision (now Alcon). Authors alsowould like to thank Jason Benninger for his technicalassistance in the manufacturing of the microfluidicsparts and Christopher Amos from CIBA Vision forfruitful discussion related to the development of thisin vitro model.

CONFLICT OF INTEREST

Ann M. Wright is an employee of Alcon (formerlyCIBA Vision). In the past 4 years, the correspondingauthor (MG) has received funding from CIBA Vision/Alcon and Johnson & Johnson.

REFERENCES

1Baca, J. T., D. N. Finegold, and S. A. Asher. Tear glucoseanalysis for the noninvasive detection and monitoring ofdiabetes mellitus. Ocul. Surf. 5:280–293, 2007.2Baudouin, C., A. Labbe, H. Liang, A. Pauly, and F.Brignole-Baudouin. Preservatives in eyedrops: the good,the bad and the ugly. Prog. Retin. Eye Res. 29:312–334,2010.3Castro-Munozledo, F. Corneal epithelial cell cultures as atool for research, drug screening and testing. Exp. Eye Res.86:459–469, 2008.4Cavet, M. E., K. L. Harrington, K. R. Vandermeid, K. W.Ward, and J. Z. Zhang. In vitro biocompatibility assess-ment of multipurpose contact lens solutions: effects onhuman corneal epithelial viability and barrier function.Contact Lens Anterior Eye 35:163–170, 2012.5Cavet, M. E., K. R. VanDerMeid, K. L. Harrington, R.Tchao, K. W. Ward, and J. Z. Zhang. Effect of a novelmultipurpose contact lens solution on human corneal epi-thelial barrier function. Contact Lens Anterior Eye33(Suppl 1):S18–S23, 2010.6Chalmers, R. L., L. Keay, J. McNally, and J. Kern. Mul-ticenter case-control study of the role of lens materials andcare products on the development of corneal infiltrates.Optom. Vis. Sci. 89:316–325, 2012.7Chalmers, R. L., H. Wagner, G. L. Mitchell, D. Y. Lam, B.T. Kinoshita, M. E. Jansen, et al. Age and other risk fac-tors for corneal infiltrative and inflammatory events inyoung soft contact lens wearers from the Contact LensAssessment in Youth (CLAY) study. Investig. Ophthalmol.Vis. Sci. 52:6690–6696, 2011.8Chu, M., T. Shirai, D. Takahashi, T. Arakawa, H. Kudo,K. Sano, et al. Biomedical soft contact-lens sensor forin situ ocular biomonitoring of tear contents. Biomed.Microdev. 13:603–611, 2011.9Debbasch, C., C. Ebenhahn, N. Dami, M. Pericoi, C. Vanden Berghe, M. Cottin, et al. Eye irritation of low-irritantcosmetic formulations: correlation of in vitro results withclinical data and product composition. Food Chem. Toxi-col. 43:155–165, 2005.

10Geerling, G., J. T. Daniels, J. K. Dart, I. A. Cree, and P. T.Khaw. Toxicity of natural tear substitutes in a fully definedculture model of human corneal epithelial cells. Investig.Ophthalmol. Vis. Sci. 42:948–956, 2001.

11Gorbet, M. B., N. C. Tanti, B. Crockett, L. Mansour, andL. Jones. Effect of contact lens material on cytotoxicitypotential of multipurpose solutions using human cornealepithelial cells. Mol. Vis. 17:3458–3467, 2011.

12Gorbet, M. B., N. C. Tanti, L. Jones, and H. Sheardown.Corneal epithelial cell biocompatibility to silicone hydrogeland conventional hydrogel contact lens packaging solu-tions. Mol. Vis. 16:272–282, 2010.

13Griffith, M. (inventor). Artificial cornea. USA patentnumber 6645715, 2003.

14Hornof, M., E. Toropainen, and A. Urtti. Cell culturemodels of the ocular barriers. Eur. J. Pharm. Biopharm.60:207–225, 2005.

15Hui, A., A. Boone, and L. Jones. Uptake and release ofciprofloxacin–HCl from conventional and silicone hydrogelcontact lens materials. Eye Contact Lens 34:266–271, 2008.

16Jacob, J. Biomaterials: contact lenses. In: Biomaterialsscience: an introduction to materials in medicine3rd, editedby B. D. Ratner, A. S. Hoffman, F. J. Schoen, and J. E.Lemons. New York: Elsevier, 2013, pp. 909–917.

MOHAMMADI et al.

17Keay L, Stapleton F, Schein O. Epidemiology of contactlens-related inflammation and microbial keratitis: a 20-yearperspective. Eye Contact Lens 33:346–353, discussion 62–63, 2007.

18Klyce, S. D. Electrical profiles in the corneal epithelium. J.Physiol. 226:407–429, 1972.

19Lefebvre, A. H. Atomization and sprays. New York:Hemisphere Press, 1989.

20Lehmann, D. M., and M. E. Richardson. Impact of assayselection and study design on the outcome of cytotoxicitytesting of medical devices: the case of multi-purpose visioncare solutions. Toxicol. In Vitro 24:1306–1313, 2010.

21Lim, M. J., R. K. Hurst, B. J. Konynenbelt, and J. L.Ubels. Cytotoxicity testing of multipurpose contact lenssolutions using monolayer and stratified cultures of humancorneal epithelial cells. Eye Contact Lens 35:287–296, 2009.

22Lorentz, H., M. Heynen, D. Trieu, S. J. Hagedorn, and L.Jones. The impact of tear film components on in vitro lipiduptake. Optom. Vis. Sci. 89:856–867, 2012.

23Mansouri, K., F. A. Medeiros, A. Tafreshi, and R. N.Weinreb. Continuous 24-hour monitoring of intraocularpressure patterns with a contact lens sensor: safety, toler-ability, and reproducibility in patients with glaucoma.Arch. Ophthalmol. 130:1534–1539, 2012.

24Mishima, S., A. Gasset, S. D. Klyce, Jr., and J. L. Baum.Determination of tear volume and tear flow. Investig.Ophthalmol. Vis. Sci. 5:264–276, 1966.

25Mitra, A. Ophthalmic drug delivery systems (2nd ed.). NewYork: Informa Healthcare, 2003.

26Mohammadi, S., L. Jones, and M. Gorbet. Investigation ofglaucoma drug-release from contact lens materials usingin vitro cell models. Investig. Ophthalmol. Vis. Sci. 53:E-Abstract 5471, 2013.

27Morgan, P. B., N. Efron, M. Helland, M. Itoi, D. Jones, J.J. Nichols, et al. Twenty first century trends in siliconehydrogel contact lens fitting: an international perspective.Contact Lens Anterior Eye 33:196–198, 2010.

28Mowrey-McKee, M., A. Sills, and A. Wright. Comparativecytotoxicity potential of soft contact lens care regimens.CLAO J. 28:160–164, 2002.

29Pauly, A., M. Meloni, F. Brignole-Baudouin, J. M. War-net, and C. Baudouin. Multiple endpoint analysis of the3D-reconstituted corneal epithelium after treatment withbenzalkonium chloride: early detection of toxic damage.Investig. Ophthalmol. Vis. Sci. 50:1644–1652, 2009.

30Postnikoff, C., R. Pintwala, S. Williams, A. Wright, D.Hileeto, and M. Gorbet. Development of a curved, strati-fied, in vitro model for assessment of ocular biocompati-bility. PLoS ONE 9(5):e96448, 2014.

31Powell, C. H., J. M. Lally, L. D. Hoong, and S. W.Huth. Lipophilic vs. hydrodynamic modes of uptake andrelease by contact lenses of active entities used in mul-tipurpose solutions. Contact Lens Anterior Eye 33:9–18,2010.

32Reichl, S., J. Bednarz, and C. C. Muller-Goymann. Humancorneal equivalent as cell culture model for in vitro drugpermeation studies. Br. J. Ophthalmol. 88:560–565, 2004.

33Reichl, S., C. Kolln, M. Hahne, and J. Verstraelen. In vitrocell culture models to study the corneal drug absorption.Expert Opin. Drug Metab. Toxicol. 7:559–578, 2011.

34Sanchez, I., V. Laukhin, A. Moya, R. Martin, F. Ussa, E.Laukhina, et al. Prototype of a nanostructured sensingcontact lens for noninvasive intraocular pressure monitor-ing. Investig. Ophthalmol. Vis. Sci. 52:8310–8315, 2011.

35Sjoquist, B., and J. Stjernschantz. Ocular and systemicpharmacokinetics of latanoprost in humans. Surv. Oph-thalmol. 47(Suppl 1):S6–S12, 2002.

36Soluri, A., A. Hui, and L. Jones. Delivery of ketotifenfumarate by commercial contact lens materials. Optom. Vis.Sci. 89:1140–1149, 2012.

37Tiffany, J. M. The viscosity of human tears. Int. Ophthal-mol. 15:371–376, 1991.

38Tiffany, J. M., N. Winter, and G. Bliss. Tear film stabilityand tear surface tension. Curr. Eye Res. 8:507–515, 1989.

39Whitson, J. T., and W. M. Petroll. Corneal epithelial cellviability following exposure to ophthalmic solutions con-taining preservatives and/or antihypertensive agents. Adv.Ther. 29:874–888, 2012.

40Xiang, C. D., M. Batugo, D. C. Gale, T. Zhang, J. Ye, C.Li, et al. Characterization of human corneal epithelial cellmodel as a surrogate for corneal permeability assessment:metabolism and transport. Drug Metab. Dispos. 37:992–998, 2009.

41Yao, H., A. J. Shum, M. Cowan, I. Lahdesmaki, and B. A.Parviz. A contact lens with embedded sensor for monitor-ing tear glucose level. Biosens. Bioelectron. 26:3290–3296,2011.

42York, M., and W. Steiling. A critical review of the assess-ment of eye irritation potential using the Draize rabbit eyetest. J. Appl. Toxicol. 18:233–240, 1998.

In Vitro Tear Replenishment Model