In Vitro Evaluation of the Ophthalmic Toxicity Profile of Chlorhexidine and Propamidine Isethionate Eye Drops

Abstract

Purpose:

Acanthamoeba keratitis causes frequent epithelial lesions that fully expose the corneal stroma. The aim of this study was to determine the toxic profile of chlorhexidine and propamidine eye drops.

Methods:

We used primary human keratocytes in cell culture in combination with a novel technology that evaluates dynamic real-time cytotoxicity through impedance analysis. Additional studies such as a classic cell viability test (WST-1^®), a bovine corneal opacity and permeability assay, and an irritation eye study (Hen's Egg Test [HET]) have been made.

Results:

Both eye drop formulations showed a time- and concentration-dependent toxicity profile, in which long periods and high concentrations were more detrimental to cells. In prolonged times of exposure, propamidine is more harmful to cells than chlorhexidine. On the contrary, no irritation has been detected in using the HET-chorioallantoic membrane test and no alterations in the corneal transparency nor permeability was produced by the treatment with both eye drops.

Conclusions:

In culture assay, chlorhexidine eye drops have proven to be less cytotoxic than Brolene^® for a long contact period of time, but no signs of irritation or alterations in transparency or permeability have been observed in the cornea after both treatments.

Introduction

Free-living amoeba of the Acanthamoeba spp. genus is the causal agent of Acanthamoeba keratitis, an ocular infection that poses a risk to corneal integrity. The number of reported cases worldwide is increasing yearly, especially among contact lens users, and the rate of successful treatment is low despite advances in antimicrobial chemotherapy and supportive care.¹

Currently, no universal protocols exist for therapeutic treatment of this pathology. The most common protocol is an initial intensive treatment that combines a diamidine and a biguanide, which produces a rapid lysis of trophozoites to prevent them from reverting into more resistant cystic forms. After an initial intensive treatment, the number of instillations is reduced due to the frequent incidence of toxic phenomena. In certain cases, it is even necessary to continue with monotherapy using only one of the medications. Therapeutic treatment with topical ophthalmic pharmacologic agents is prolonged, with a minimum duration of 4 weeks, and some authors recommend therapy for at least 6 or 12 months.²

The low investment in research and development for low-incidence pathologies hinders the development and commercialization of effective medications by the pharmaceutical industry.³ Currently, no ophthalmic topical treatment for Acanthamoeba keratitis is commercially available in Spain. Thus, hospital pharmacies are responsible for the foreign acquisition and transaction involved in importing diamidines, such as propamidine isethionate 0.1% (Brolene^® or Golden Eye^®), and for developing biguanides through compounded formulation of chlorhexidine 0.02% eye drops.⁴

Typically, compounded ophthalmic formulations made at hospital pharmacies use commercially available medications intended for use via the parenteral route. The medications are dissolved or diluted in tamponades that are compatible with the ocular route. However, these medications are not designed or adapted for that route; therefore, they may have side effects at the ophthalmic level.⁵ Conversely, only ocular toxicity and dose–response studies are available for the pharmaceutical drugs used, which is why the concentrations used are adapted and used based on results and clinical experience gained over time. Furthermore, poor adaptation to the ocular route and the tendency to use very simple formulations generate systems that lack effectiveness because they have a very low retention time on the ocular surface due to high precorneal clearance.⁶

Dosage intervals based on frequent instillations of eye drops with high drug concentrations and for long periods of time are prescribed with the goal of maintaining therapeutic concentrations. This practice leads to increased patient discomfort, resulting in reduced treatment adherence and, thus, reduced therapeutic effectiveness.⁷ Furthermore, toxic ocular effects are frequent and are caused by commercial medications such as nonsteroidal anti-inflammatory drugs and antiglaucomatous prostaglandins. These medications, despite being tightly controlled by regulatory agencies, sometimes present unacceptable toxicities.⁸

One essential aspect in the development of new ophthalmic drugs is the evaluation of the local tolerance. The effects of the eye drops on the ocular structures are usually studied using in vitro cellular studies, ex vivo assays, or in vivo animal models such as mice, rabbits, or rats. The in vivo standard model proposed to study ocular irritation is the Draize test,⁹ however, the use of vertebrates to evaluate the safety has been widely criticized based on scientific and ethical reasons. The concept of the 3Rs (refinement, replacement, and reduction) stimulates the development of alternative methods such as in vitro methods, ex vivo methods, or the use of “lower” organisms as models (invertebrates, plants, and microorganisms).¹⁰ According to different articles, the Hen's Egg Test–chorioallantoic membrane (HET-CAM) method, bovine corneal opacity and permeability assay, or the in vitro cellular studies are a good alternative to the use of animal model in toxicity eye studies.^11–13

Cellular models of detection of toxicity, including cytotoxicity assays, based on the observation of changes on morphology, viability, metabolism, or cell adhesion can be a good alternative to determine the behavior that the cells adopt in vivo. The most commonly used cytotoxicity assays are the neutral red dye, MTT^®, WST-1^® or AlamarBlue^®. In addition, novel detection systems in real time without cell markings as xCELLigence real-time cell analyzer system impedance analysis (RTCA) are increasingly used.¹⁴ RTCA is a novel methodology with several advantages over the classical methods because it does not require the use of labels, is noninvasive, it does not interfere with dyes, and allows to make continuous measurement in time. Also, previous articles have demonstrated a good correlation between RTCA and classical cytotoxicity assays.^15,16

It is of interest to guarantee that the preparations are suitable for their intended use and pose no threat to the patient. To achieve this, the toxic potential of these formulations must be established to develop estimations of the benefits and risks that the formulations can have in complicated clinical situations where the integrity of the cornea is compromised. This study aims to evaluate the toxicity profile of anti-Acanthamoeba eye drops.

Methods

This is an in vitro experimental study developed to determine the toxicity of the compounded ophthalmic formulation made in hospital pharmacies, 0.22 mM chlorhexidine eye drops and 0.17 mM propamidine isethionate eye drops (Brolene). Table 1 shows the composition of the evaluated eye drops.

Table 1.

Composition of the Evaluated Anti-Acanthamoeba Eye Drops

0.22 mM Chlorhexidine eye drops	Propamidine isethionate eye drops (Brolene^®)
• Bohmclorh chlorhexidine 0.5%	• Ammonium chloride
• Water for injection	• Benzalkonium chloride
Compounded formulation made at pharmacy hospitals under sterile conditions established by the Guide for Good Practices in the Preparation of Medicinal Products in Healthcare Establishments.¹⁰	• Sodium chloride
	• Sodium hydroxide
	• Water for injection

Cell culture assays

Isolation of the stromal keratocyte primary culture

Keratocyte primary cultures were obtained from human cornea fragments remaining from those used for corneal transplant in a manner consistent with institutional ethical standards and the Declaration of Helsinki. Cells were incubated for 10 min in trypsin at 37°C based on the modified Ramke method, followed by mechanical elimination of the endothelium and epithelium. The corneal stroma was cut into 2-mm sections that were submerged in Dulbecco's modified Eagle's medium/Ham F-12 supplemented with 10% fetal bovine serum, 2 mM L-glutamine, and antibiotics (100 IU penicillin, 100 mg/mL streptomycin, and 50 mg/mL gentamycin).

Cytotoxicity assay

Cell assays were performed using a cellular bioimpedance biosensor system (xCELLigence Real-Time System Cell Analyzer; RTCA).¹⁷ This system uses electronic microchips that measure changes in the impedance between the electrodes and the solution. When the cell adheres to the well (16-well E-plate; ACEA Biosciences), the resistance increases, which increases the impedance. The impedance values are transformed into a parameter known as the cellular index (CI)¹⁴ using an algorithm. A low CI indicates a lower number of cells adhering to the microelectrode, and an increase in the CI indicates an increase in the number of cells¹⁸; therefore, all changes are detected in a continuous manner in real time. To evaluate cytotoxicity, CI values are expressed as the normalized CI (NCI), where CI_i(t) is the CI at a given time, and CI_{i(t of dose)} is the CI at the moment at which the medications are added to the culture medium. The results were graphed to show the dynamic changes for each of the assayed concentrations, and NCI is represented over time starting with the addition of the compounds. The concentrations of compounds that reduced the CI by 50% were calculated by interpolation from the graphs that showed the normalized response (%), and this parameter was called the CI50.¹⁹

Cytotoxicity assays

Cytotoxicity was evaluated using an RTCA. Previous studies have shown that the optimal number of cells is 3,000 per well.⁸ On reaching a CI of ∼1.0, eye drops were added to the culture at different concentrations. After adding the medications, the cell behavior, measured as the CI, was recorded continuously and automatically every 15 min for 20 h. All assays were performed using cells between passages 4 and 10.

In addition, the colorimetric reaction assays WST-1 (Cell Proliferation Reagent WST-1 from Roche Applied Science), which are based on cell viability in relation to mitochondrial enzymes,²⁰ were performed. Both methods determined keratocyte viability after 30 min, 75 min, 8 h, and 24 h of incubation with these eye drops.

The Hen's Egg Test–chorioallantoic membrane

The HET-CAM test was used to determine the potential irritancy of the anti-Acanthamoeba eye drops. The method used was similar to that described in previous studies.²¹ Briefly, freshly fertilized white leghorn eggs, weighing 50–60 g, were incubated at 37.5°C with a relative humidity of ∼65% for 9 days in an incubator with an automatic rotating device. After removing the eggshell covering their air cell with an electric drill and cutting through the inner egg membranes, the test substance (0.3 mL) was applied onto the vasculated CAM of at least three eggs. The CAM was observed over time (300 s) under a stereomicroscope (Olympus SZ-STN) with an integrated standalone digital camera with full live HD video output (Leica IC80 HD) and scored for the following effects: hemorrhage, vascular lysis, and coagulation. Sodium hydroxide (0.1 N) served as the positive control, and sodium chloride 0.9% served as the negative control. These anti-Acanthamoebas were placed in the CAM to determine the irritation score (IS) by the classic methodology described in Protocol No. 96 de INVITTOX.

Bovine corneal opacity and permeability assay

The procedure used was a variation of the method described by Parekh et al.²² The bovine whole eye balls were obtained from a local slaughter house and collected immediately after the cows were slaughtered. They were transported to the laboratory in balanced salt solution (BSS) in cold condition (4°C). Corneas were excised along with 1–2 mm of surrounding scleral tissues and were placed in warm BSS for 50 min. The corneas were placed between two cylindrical supporting black holders (fabricated with polylactic acid filaments using a 3D print, Witbox BQ) that have a special shape adapted to the cornea curvature and a hole (diameter = 11.5 mm), which allows the transmission of the light between the light source (Olympus Highlight 200 pipe light in Brightness position 3) and the light probe, which detects the transmitted light measured by an illuminance meter (Gossen 5032C USB). An initial baseline opacity measurement (%T_initial, difference of light intensity without and with cornea) is performed for all the corneas in contact with BSS. The initial opacity reading is representative of a normal, untreated cornea. After the initial opacity reading, each cornea was placed in Franz cells with the epithelium oriented to the receptor chamber that contains 500 μL of the tested products (chlorhexidine 0.02% and Brolene eye drops and BSS^® as negative control). After 10 min of contact, the corneas were removed and the transparency was measured again (%T_10min). For permeability assay, the treated corneas were fixed in the Franz cell, fluorescein is added to the anterior chamber and PBS in the receptor chamber and incubated for 90 min. After this time, the receptor medium is transferred to a quartz cuvette and the optical density at 490 nm is determined using a spectrophotometer (Agilent 8453). Each product was tested in duplicate and the corneal surface in contact with the products was 0.785 cm².

Statistical analysis

The statistical analysis used to compare the results of cellular assays was a two-way analysis of variance using the Tukey test as multiple comparison test. Analyses were developed using GraphPad Prism 6.0 software.

Results

Cell culture assays

Adding anti-Acanthamoeba eye drops to the standard culture medium of keratocytes at confluence (Fig. 1) led to a marked decrease in cell viability, which can be observed in Figs. 2 and 3.

FIG. 1.

Immunostaining in cell cultures of keratocytes (10X). The cells presented cytoplasmic fibrillar immunoreactivity for vimentin.

FIG. 2.

Toxicokinetic profile of Brolene^® eye drops 20 h after exposure to different concentrations of the drug. (A) Changes in normalized cell index over time. (B) Changes in CI50 over time. CI, cellular index.

FIG. 3.

Toxicokinetic profile of chlorhexidine eye drops 20 h after exposure to keratinocytes using different drug concentrations. (A) Changes in normalized cell index over time. (B) Changes in CI50 over time. CI, cellular index.

Figure 2A reveals that the highest propamidine isethionate concentrations caused a pronounced decrease in cell viability, leading to null NCIs when testing the highest eye drop concentrations. In addition, the highest concentrations of eye drops led to an earlier decrease in the NCI, with a sharp reduction in viability during the initial period of exposure. These effects are shown in Fig. 2B, which demonstrates an increase in compound toxicity as the cell exposure time increases. Table 2 shows that the original eye drop concentrations (OEDC) used for therapy were several fold greater than the CI50 toxicity parameter. On increasing the time of exposure, the CI50 decreases from the initial value of 110 microM (which is 1.5-fold lower than the OEDC typically used for therapy) to 4 microM (the OEDC is 42 times greater than the CI50 after 20 h).

Table 2.

CI50 Evolution and Fold Change of the Original Eye Drop Concentration and the CI50 Over Time

	Brolene^® eye drops (foreign commercial product)		Chlorhexidine^® eye drops (compounded formulation)
Exposure time	CI50 (mM)	Number of times OEDC surpasses CI50	CI50 (mM)	Number of times OEDC surpasses CI50
20 min	110	1.5	94	2.3
1 h	94	1.8	40	5.5
8 h	12	14.2	32	6.9
10 h	10	17	29	7.6
20 h	4	42.5	22	10

CI, cellular index; OEDC, original eye drop concentration.

Figure 3A, as in the previous case, shows that chlorhexidine exhibits a concentration-dependent toxic effect, with a correlation between the most pronounced reductions in viability and the highest concentrations. In the case of chlorhexidine, the reductions in viability are less pronounced than those of propamidine because full loss of the total keratocyte population does not occur with any of the assayed concentrations. Meanwhile, as seen in the previous case, the highest eye drop concentrations led to earlier decreases in the NCI. The effects of chlorhexidine eye drop toxicity can be observed quantitatively in Fig. 3B. The data show a progressive decrease in the CI50 over time, from an initial value of 94 microM to 20 microM at 20 h, and the OEDC surpasses these values by 2.3- and 10-fold, respectively (see Table 2).

We performed a complementary assay of cell viability based on the measurement of the mitochondrial activity of the keratocytes in the presence of the anti-Acanthamoeba eye drops (WST-1 assay). The results obtained with this method are showed in Fig. 4. Two-way analysis of variance shows that both the time of contact (α<0.01) and the different treatments and concentrations (α<0.01) have significant influence on percentage of cellular viability. Cell death increases as the exposition time increases and also with Chlorhexidine^® or Brolene concentration. No significant differences were observed between both drugs for similar concentrations, but there is a tendency to present higher cell viability of chlorhexidine. These results are in accordance with those obtained using RTCA.

FIG. 4.

Cell viability in the presence of anti-Acanthamoeba eye drops obtained using WST-1^® assay.

Hen's Egg Test–chorioallantoic membrane

In Fig. 5, no damage to the blood vessels on the CAM surface after a period of 5 min of contact with the anti-Acanthamoeba eye drops was observed. A null IS for both eye drops was observed; thus, these can be classified as nonirritants.

FIG. 5.

Images of chorioallantoic membrane after a contact time of 5 min with anti-Acanthamoeba eye drops.

Bovine corneal opacity and permeability assay

Results shown in Table 3 indicate that no alterations in transparency nor permeability were produced in the cornea by the treatment with chlorhexidine 0.02% and Brolene. The values of percentage of transparency and the OD were similar in the formulations compared with BSS^®.

Table 3.

Bovine Corneal Opacity and Permeability Results

	Transparency				Permeability
	% T_initial		% T_10min		OD_Form-OD_BBS
	Mean	SD	Mean	SD	Mean	SD
BSS (negative control)	88	0.01	88	0.01	0	0
Chlorhexidine 0.02% eye drops	86	0.01	89	0.01	0.0026	0.0080
Brolene^® (propamidine isethionate 0.1%)	92	0.07	90	0.01	−0.0001	0.0106

BSS, balanced salt solution; OD, optical density; SD, standard deviation; T, transparency.

Discussion

Previous evidence has shown corneal toxicity due to the use of propamidine isethionate^23,24 and chlorhexidine²⁵ eye drops; however, their mechanism of action has not been determined. In this study, a new in vitro cellular method based on the use of a cellular impedance biosensor system in primary cultures of human keratocytes was used. The method, unlike classical techniques such as MTT or alamarBlue, allows for direct and continuous evaluation of changes in cell viability over time as a consequence of contact between keratocytes and the assayed substances, which provides significant dynamic information about toxicity.²⁶ Keratitis exhibits an initial phase of dendritiform epithelial defects, anterior stromal infiltrates, and radial keratoneuritis, such that the stroma becomes partially exposed to the prescribed pharmaceutical drugs. Therefore, detection of the toxicity of these products on stromal keratocytes is of interest. The present study is the first published in vitro study of the toxicity of these products on this cell line using the novel technique of cellular bioimpedance.

We have observed that despite having initial CI50 values similar to propamidine, chlorhexidine is more toxic because the OEDC surpasses the CI50 by 2.3-fold, while propamidine surpasses its respective OEDC by 1.5-fold. However, at prolonged times of exposure to cells, the value is inverted, with propamidine (with an OEDC 42 times larger than the CI50) becoming more harmful to cells than chlorhexidine (with an OEDC 10 times greater than the CI50). Classical WST-1 assay confirms the toxicity of both eye drops and their high dependence of contact time and drug concentration. Despite the formulation showing significant cytotoxicity, neither acute irritation nor alteration on permeability or transparency was observed. This behavior can produce an absence of discomfort and observable apparent injuries after their administration, so patients using these eye drops have to be carefully surveilled to avoid ocular adverse reactions.

Extrapolating data from in vitro to in vivo settings are complicated and should be performed with caution. The results here show that the currently used concentrations greatly surpass the in vitro toxicological margin of safety.²⁷ The use of high concentrations occurs, in part, due to the significant precorneal clearance that occurs in ocular physiology, which makes it difficult to achieve efficient medication concentrations for prolonged periods of time on the ocular surface. This strategy is likely inadequate because it favors periods of contact between the cornea or conjunctiva and high concentrations of the medication, leading to adverse effects. Instead, strategies should be promoted to determine alternative formulations that aim to increase the residence time of the medication on the ocular surface through adequate design of formulations for the ophthalmic route. Hydrogels are an example of a substance with the capacity for in situ gelification.¹⁶ These substances are used because of their ability to change consistency and structure when exposed to specific stimuli. The ocular surface provides conditions that can facilitate the formation of gels with certain polymers that are sensitive to stimuli. In the absence of stimuli, these gels behave similar to fluid systems and are easy to administer. Once administered, the formulations transform into a bioadhesive hydrogel film that remains adhered over the ocular surface for prolonged periods of time while delivering the active compound.²⁸

Meanwhile, it must be mentioned that often, issues of ophthalmic toxicity are related to the excipients used to produce the eye drops and those that coexist with the active compound. In many cases, these excipients or their combination with the active compounds are the main causes of ocular toxicity. It should be noted that in this study, the full eye drop formulation was used, and Brolene includes preservatives such as benzalkonium chloride, which must be considered when evaluating toxicity. Some authors have described that excipients commonly used in ophthalmologic treatments, such as benzalkonium chloride,²⁹ etilendiaminotetraacetic acid,³⁰ or cyclodextrins,³¹ are responsible for the toxicity of some ophthalmic products. Thus, the increased cytotoxicity exhibited by propamidine isethionate eye drops relative to chlorhexidine could be due to the presence of a preservative, while chlorhexidine also has shown antibacterial therapeutic efficiency.³²

Therefore, future research should be performed to evaluate the safety of excipients used in the production of eye drops. This will provide the information needed to select and optimize the composition of the formulations used in ophthalmologic treatments and promote the use of multidose administration systems with no preservatives, such as Hylabak^®,³³ because preservatives often cause irritation or toxic effects. Currently, several research lines are focusing on determining new pharmacological targets to treat Acanthamoeba keratitis. Pharmaceutical drugs such as statins³⁴ and miltefosine³⁵ are being studied, and strategies based on crosslinking are being developed.³⁶ However, these strategies have not been adopted in clinical treatment guides because most studies only show in vitro results or are case studies. To avoid the development of ocular toxicity episodes, it is crucial to determine the safety of the medications that are currently used as the first line of topical ophthalmic treatments for Acanthamoeba keratitis. In this study, both eye drops chlorhexidine and Brolene have proven to have a favorable cytotoxicity profile from ex vivo corneal transparence, permeability, and acute irritation point of view, but the cell culture assays shows that chlorhexidine eye drops are the less toxic current alternative. The use and efficiency of chlorhexidine eye drops have been supported by recently published studies.^37,38

Footnotes

Acknowledgments

A.F-F. acknowledges the support of Instituto de Salud Carlos III (Rio Hortega research grant CM15/00188). The authors also acknowledge the support of Fundación Mutua Madrileña and Fundación Española Farmacia Hospitalaria (AISEFH).

Author Disclosure Statement

No competing financial interests exist.

References

Lorenzo-Morales

, Martín-Navarro

C.M.

, López-Arencibia

, Arnalich-Montiel

, Piñero

J.E.

, and Valladares

Acanthamoeba keratitis: an emerging disease gathering importance worldwide?. Trends Parasitol. 29:181–187, 2013.

Lorenzo-Morales

, Khan

N.A.

, and Walochnik

An update on Acanthamoeba keratitis: diagnosis, pathogenesis and treatment. Parasite. 22:1–20, 2015.

Pammolli

, Magazzini

, and Riccaboni

The productivity crisis in pharmaceutical R&D. Nat. Rev. Drug Discov., 10:428–438, 2011.

Lim

, Goh

, Bunce

, et al. Comparison of polyhexamethylene biguanide and chlorhexidine as monotherapy agents in the treatment of Acanthamoeba keratitis. Am. J. Ophthalmol., 145:130–135, 2008.

Herreros

J.M.A.

Preparación de medicamentos y formulación magistral para oftalmología. Madrid, España: Ediciones Díaz de Santos; 2003.

Gaudana

, Ananthula

H.K.

, Parenky

, and Mitra

A.K.

Ocular drug delivery. AAPS J. 12:348–360, 2010.

Tatham

A.J.

, Sarodia

, Gatrad

, and Awan

Eye drop instillation technique in patients with glaucoma. Eye. 27:1293–1298, 2013.

Fernández-Ferreiro

, Santiago-Varela

, Gil-Martínez

, et al. Ocular safety comparison of non-steroidal anti-inflammatory eye drops used in pseudophakic cystoid macular edema prevention. Int. J. Pharm., 495:680–691, 2015.

Draize

J.H.

, Woodard

, and Calvery

H.O.

Methods for the study of irritation and toxicity of substances applied topically to the skin and mucous membranes. J. Pharmacol. Exp. Ther., 82:377–390, 1944.

10.

York

, and Steiling

A critical review of the assessment of eye irritation potential using the Draize rabbit eye test. J. Appl. Toxicol., 18:233–240, 1998.

11.

Zuang

, Schäffer

, et al. EURL ECVAM progress report on the development, validation and regulatory acceptance of alternative methods (2010–2013). JRC Scientific and Policy reports.

12.

OECD 405 (October 2012) Acute Eye Irritation/Corrosion, OECD Guidlines for the Testing of Chemicals. Available at www.oecd.org/env/ehs/testing/TG%20List%20EN%20Aug%202012.pdf (accessed February 15, 2016).

13.

Luepke

N.P.

Hen's egg chorioallantoic membrane test for irritation potential. Food Chem. Toxicol., 23:287–291, 1985.

14.

Atienza

J.M.

, Zhu

, Wang

, Xu

, and Abassi

Dynamic monitoring of cell adhesion and spreading on microelectronic sensor arrays. J. Biomol. Screen., 10:795–805, 2005.

15.

Quereda

J.J.

, Martínez-Alarcón

, Mendoça

, et al. Validation of xCELLigence real-time cell analyzer to assess compatibility in xenotransplantation with pig-to-baboon model. Transplant. Proc., 42:3239–3243, 2010.

16.

Fernández-Ferreiro

, González Barcia

, Gil-Martínez

, et al. In vitro and in vivo ocular safety and eye surface permanence determination by direct and Magnetic Resonance Imaging of ion-sensitive hydrogels based on gellan gum and kappa-carrageenan. Eur. J. Pharm. Biopharm., 94:342–351, 2015.

17.

Boyd

J.M.

, Huang

, Xie

, Moe

, Gabos

, and Li

X.-F.

A cell-microelectronic sensing technique for profiling cytotoxicity of chemicals. Anal. Chim. Acta., 615:80–87, 2008.

18.

Halle

The registry of cytotoxicity: toxicity testing in cell cultures to predict acute toxicity (LD50) and to reduce testing in animals. Altern. Lab. Anim., 31:89–198, 2003.

19.

Otero-González

, Sierra-Alvarez

, Boitano

, and Field

J.A.

Application and validation of an impedance-based real time cell analyzer to measure the toxicity of nanoparticles impacting human bronchial epithelial cells. Environ. Sci. Technol., 46:10271–10278, 2012.

20.

Peskin

A.V.

, and Winterbourn

C.C.

A microtiter plate assay for superoxide dismutase using a water-soluble tetrazolium salt (WST-1). Clin. Chim. Acta., 293:157–166, 2000.

21.

Fernández-Ferreiro

, González Barcia

, Gil Martínez

, Blanco Mendez

, Lamas Díaz

M.J.

, and Otero Espinar

F.J.

Analysis of ocular toxicity of fluconazole and voriconazole eyedrops using HET-CAM. Farm. Hosp., 38:300–304, 2014.

22.

Parekh

, Ferrari

, Ruzza

, Pugliese

, Ponzin

, and Salvalaio

A portable device for measuring donor corneal transparency in eye banks. Cell. Tissue. Bank., 15:7–13, 2014.

23.

Johns

K.J.

, Head

W.S.

, and O'Day

D.M.

Corneal toxicity of propamidine. Arch. Ophthalmol., 106:68–69, 1988.

24.

Alizadeh

, Silvany

R.E.

, Meyer

D.R.

, Dougherty

J.M.

, and McCulley

J.P.

In vitro amoebicidal activity of propamidine and pentamidine isethionate against Acanthamoeba species and toxicity to corneal tissues. Cornea. 16:94–100, 1997.

25.

Fraunfelder

F.T.

, Fraunfelder

F.W.

Jr. , and Chambers

W.A.

Drug-Induced Ocular Side Effects: Clinical Ocular Toxicology. Elsevier Health Sciences; 2014.

26.

Meindl

, Absenger

, Roblegg

, and Frohlich

Suitability of cell-based label-free detection for cytotoxicity screening of carbon nanotubes. BioMed. Res. Int., 2013:1–13, 2013.

27.

Yoon

, Campbell

J.L.

, Andersen

M.E.

, and Clewell

H.J.

Quantitative in vitro to in vivo extrapolation of cell-based toxicity assay results. Crit. Rev. Toxicol., 42:633–652, 2012.

28.

Geethalakshmi

, Karki

, Jha

S.K.

, Venkatesh

D.P.

, and Nikunj

Sustained ocular delivery of brimonidine tartrate using ion activated in situ gelling system. Curr. Drug. Deliv., 9:197–204, 2012.

29.

Mencucci

, Paladini

, Pellegrini-Giampietro

D.E.

, Menchini

, and Scartabelli

In vitro comparison of the cytotoxic effects of clinically available ophthalmic solutions of fluoroquinolones on human keratocytes. Can. J. Ophthalmol., 46:513–520, 2011.

30.

Epstein

S.P.

, Ahdoot

, Marcus

, and Asbell

P.A.

Comparative toxicity of preservatives on immortalized corneal and conjunctival epithelial cells. J. Ocul. Pharmacol. Ther., 25:113–119, 2009.

31.

Fernández-Ferreiro

, Fernández

Bargiela

, N., Varela

M.S.

, et al. Cyclodextrin-polysaccharide-based, in situ-gelled system for ocular antifungal delivery. Beilstein. J. Org. Chem., 10:2903–2911, 2014.

32.

E.Y.

, Shoff

M.E.

, Gao

, and Joslin

C.E.

Effect of low concentrations of benzalkonium chloride on acanthamoebal survival and its potential impact on empirical therapy of infectious keratitis. J. Am. Med. Assoc. Ophthalmol., 131:595–600, 2013.

33.

Brjesky

V.V.

, Maychuk

Y.F.

, Petrayevsky

A.V.

, and Nagorsky

P.G.

Use of preservative-free hyaluronic acid (Hylabak^®) for a range of patients with dry eye syndrome: experience in Russia. Clin. Ophthalmol. Auckl. NZ., 8:1169–1177, 2014.

34.

Martín-Navarro

C.M.

, Lorenzo-Morales

, Machin

R.P.

, et al. Inhibition of 3-hydroxy-3-methylglutaryl-coenzyme A reductase and application of statins as a novel effective therapeutic approach against Acanthamoeba infections. Antimicrob. Agents Chemother., 57:375–381, 2013.

35.

Barisani-Asenbauer

, Walochnik

, Mejdoubi

, and Binder

Successful management of recurrent Acanthamoeba keratitis using topical and systemic miltefosine. Acta. Ophthalmol. (Copenh.). 90, 2012.

36.

Khan

Y.A.

, Kashiwabuchi

R.T.

, Martins

S.A.

, et al. Riboflavin and ultraviolet light a therapy as an adjuvant treatment for medically refractive Acanthamoeba keratitis: report of 3 cases. Ophthalmology. 118:324–331, 2011.

37.

Mafra

C.S.P.

, Carrijo-Carvalho

L.C.

, Chudzinski-Tavassi

A.M.

, et al. Antimicrobial action of biguanides on the viability of Acanthamoeba cysts and assessment of cell toxicity. Invest. Ophthalmol. Vis. Sci., 54:6363–6372, 2013.

38.

Rahimi

, Hashemian

S.M.N.

, Tafti

M.F.

, et al. Chlorhexidine monotherapy with adjunctive topical corticosteroids for Acanthamoeba keratitis. J. Ophthalmic. Vis. Res., 10:106–111, 2015.