Sign up
Title:
IN VITRO METHOD FOR TESTING TOXICITY AND EFFICACY OF A TEST MATERIAL TO AN OCULAR SURFACE
Kind Code:
A1
Abstract:
This invention relates to an in vitro method and kit for testing the toxicity through the characterization of the healing efficacy of a test material. The test method includes the following steps: i) performing a monolayered or multilayered culture of epithelial corneal or conjunctival cells including quiescent and dividing cells, said culture being obtained by cultivating cells to confluence followed by the scraping of the cell layer(s) resulting into a cleft, ii) applying the test material onto said culture, iii) measuring cell viability and closure of the cleft with the same culture treated with a reference solution, such as for example a saline solution.


Inventors:
GARRIGUE, Jean-Sébastien (8 voie de l'Aulne, Verrières Le Buisson, F-91370, FR)
DAULL, Philippe (34 rue du Cimetière, Soisy-sur-Seine, F-91450, FR)
BAUDOUIN, Christophe (17 avenue Marcel Proust, Paris, F-75016, FR)
BRIGNOLE-BAUDOUIN, Françoise (17 avenue Marcel Proust, Paris, F-75016, FR)
LIANG, Hong (10 square Albert Einstein, Evry, F-91000, FR)
Application Number:
EP2010/068859
Publication Date:
06/09/2011
Filing Date:
12/03/2010
Assignee:
NOVAGALI PHARMA SA (Bâtiment Genavenir IV, 1 rue Pierre Fontaine, Evry, F-91000, FR)
GARRIGUE, Jean-Sébastien (8 voie de l'Aulne, Verrières Le Buisson, F-91370, FR)
DAULL, Philippe (34 rue du Cimetière, Soisy-sur-Seine, F-91450, FR)
BAUDOUIN, Christophe (17 avenue Marcel Proust, Paris, F-75016, FR)
BRIGNOLE-BAUDOUIN, Françoise (17 avenue Marcel Proust, Paris, F-75016, FR)
LIANG, Hong (10 square Albert Einstein, Evry, F-91000, FR)
International Classes:
G01N33/50
View Patent Images:
Attorney, Agent or Firm:
ICOSA (142 rue de Rennes, Paris, F-75006, FR)
Claims:
CLAIMS

1. An in vitro method for testing toxicity and/or healing properties of a test material to an ocular surface including the steps of

- performing a monolayered or multilayered culture of epithelial corneal or conjunctival cells including quiescent and dividing cells, said culture being obtained by cultivating cells to confluence followed by a scraping of the cell layer (s) resulting into a cleft,

- applying the test material onto said culture,

- measuring cell viability and closure of the cleft within said culture, and within the same culture treated with a reference solution, such as for example a saline solution as a control.

2. The in vitro method of claim 1, wherein the scrap results in a cleft having a width of 100 to 500 pm more preferably 200 to 400 pm. 3. The in vitro method of claim 1 or claim 2, wherein the test material is a new chemical entity, an active principle, and an excipient, formulated either as liquid formulations such as for example aqueous solutions, eye drops, emulsions, oily solutions, micelles; or fluid formulations such as for example ointments, creams, lotions.

4. The method of anyone of claim 1 to claim 3, wherein the test material is a pharmaceutical or a dermatological composition, such as for example a pharmaceutical composition including a prostaglandin or a prostaglandin derivative, such as for example latanoprost or derivative thereof.

5. The method of anyone of claims 1 to 4, wherein the test material is a detergent, an excipient, a cosmetic product, such as for example a make-up product.

6. The in vitro method of anyone of claims 1 to 5, wherein the test material is selected from the group consisting of pharmaceutical compositions useful for enhancing healing of the ocular surface affected by surgery, ulcers, dry eye, injury, scars.

7. Kit for implementation of the method according to anyone of claims 1 to 6 comprising cells, a solution to be used as a negative control and a dye, preferably selected from the group comprising trypan blue and bromodeoxyuridine .

Description:
IN VITRO METHOD FOR TESTING TOXICITY AND EFFICACY OF A TEST MATERIAL TO AN OCULAR SURFACE

Field

The present invention relates to a sensitive method for toxicity and/or efficacy evaluation of material preparation without the use of animals.

Background of the invention

Numerous adverse effects of various preservatives, excipients and active ingredients on the cornea and conjunctiva have been described like corneal erosions, allergic reactions, destabilization of the tear film, corneal deposits or delayed wound healing rate These corneal and conjunctival adverse effects have been investigated by a variety of techniques- in vitro, ex vivo or in vivo - and using different animal models - rabbits or rodents. The tests currently used to evaluate the safety of xenobiotics in human eyes with vertebrate animals have been heavily criticized because of ethical and scientific considerations. Indeed, the Draize rabbit eye test (Draize et al . , 1944), a test adopted worldwide by regulatory bodies for the assessment of ocular irritancy, and based mainly on subjective scoring of observed macroscopic changes in the rabbit cornea, conjunctiva and iris after exposure to a test compound is well established. However, these observations have been discussed and criticized because of their subjectivity (Buehler, 1974; Heywood and James, 1978), the high dose of test material used (Freeberg et al . , 1986; Chambers et al . , 1993; Lambert et al . , 1993), their intra- and inter-laboratory variability (Marzulli and Ruggles, 1973; McDonald and Shaddhuck, 1977), their over-predictiveness of human responses (Marzulli and Simon, 1971; Buehler, 1974), and because they harm animals (Rowan, 1984; Zbinden, 1985) .

Considerable work has been devoted to developing alternatives for ocular safety testing. The principal alternative to animal testing is in vitro testing. Among these tests, one can cite for example: the bovine corneal opacity and permeability (BCOP) , Cytosensor® Microphysiometer (CM) , hen's egg test-chorioallantoic membrane (HET-CAM) , isolated chicken eye (ICE), and isolated rabbit eye (IRE) test. These methods are in vitro test methods used to predict the extent of ocular damage that might occur in vivo without requiring the use of live animals. The numerous in vitro assays developed generally use tissue or organs, or cell cultures such as corneal epithelial cell cultures, and measure a wide range of end-points such as cell cytotoxicity (i.e. survival), proliferation, membrane permeability or metabolism (Bruner, 1992; Chu and Toft, 1993) . Their advantages over living animal testing include high throughput, cost, effectiveness, simplicity, reproducibility, and reduced suffering in animals (Rohde, 1997) . However, numerous in vitro test methods are not predictive enough of the in vivo toxicity. In its 2010 evaluation report of alternative in vitro methods the Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM) concludes that the accuracy and reliability of the BCOP, CM, HET-CAM, ICE test methods do not support their use as a screening test to distinguish substances not labelled as irritants (ICCVAM In Vitro Ocular Evaluation Report, 2010) . Therefore, these methods cannot be recommended as screening tests to identify substances not labelled as irritants (i.e. the vast majority of excipient used in ophthalmology for instance) . Hence, in vitro tests with a strong correlation with in vivo data regarding the toxicity of excipients are urgently needed. Although there are numerous screening tests available, some of which have been well-validated (Balls et al . , 1995), none has yet gained full acceptance (Jester et al . , 1996; Schneider et al . , 1997). The reason for this lack of replacement is in part due to the difficulty in correlating in vitro data with subjective in vivo test results (Jester et al . , 1996).

Evaluation of cell damage on the culture was automated by several microscopic methods such as fluorescence reader (R.W. Yee, H.M. Sorour, S.B. Yee, A.Z. Chuang and X. Zhao Comparison of Relative Toxicity of Four Ophthalmic Antibiotics Using the Human Cornea Epithelial Cell Culture System Invest Ophthalmol Vis Sci 2004; 45: E-Abstract 4939.) or confocal microscope (Engelke M, Patzke J, Tykhonova S, Zorn-Kruppa M. Assessment of ocular irritation by image processed quantification of cell injury in human corneal cell cultures and in corneal constructs. Altern Lab Anim. 2004 Oct ; 32 ( 4 ) : 345-53 ) . The advantage of this automation is to avoid subjective interpretation, operator defects and to enhance reproducibility of tests. However, these in vitro tests evaluate the live/death state of the cells in culture. In these common in vitro tests, cells are in a quiescent state, which renders them more resistant to xenobiotics, as their metabolism is slowed down (i.e. a higher dose is necessary to induce the sought after toxicity endpoint) . Hence, this evaluation is not fully satisfactory as it did not reflect in vivo situations. In vivo, corneal cells are in constant regeneration and the toxicity of xenobiotics is more likely to affect primarily the cells under active multiplication and regeneration .

The present invention relates to a single method for evaluating both the toxicity and its underlying mechanism by assessing the detrimental or beneficial effects on the healing properties of a test material.

The method of the invention is of high interest with regards to above-mentioned technical issues, in that it is implemented in vitro, thus avoiding any use or suffering of animals, whereas reflecting accurately the in vivo situation and leading to data which are fully correlated with the expected results in vivo. The method of the invention is also easy to perform and reproducible.

The method of the invention involves a scraping step. Even though scraping tests are known for a long time by cell biologists to characterize the cell migration and intracellular skeleton reorganization, no one, in the extent of the Applicant's knowledge, has ever applied the scraping technique, i.e. the characterization of the healing properties as a point of interest for the evaluation of the toxicity of test materials onto the surface of the eye. For that purpose, the evaluation of healing properties and mechanism of healing to screen test materials for their toxicity is new.

The method of the invention involves the use of cultured cells. There are some publications or patents that describe the use of either a monolayer of quiescent cells or synthetic cornea epithelia (skinEthic model). However, the Applicant noticed that the use of quiescent cells could not lead to satisfactory results, as mentioned above. The Applicant, taking into account the specificity of the corneal and conjunctival cells, realized that, when carrying out a scrap, the cells present in the culture were turning into two states: first, some quiescent cells remain, but also, due to the scraping a number of other cells are non quiescent anymore; i.e. leave the GO state, re-enter in mitosis, and restart cell cycling (cells are in a "division state") . The Applicant noticed that this in vitro model leads to a situation very close to an in vivo situation, and that the results of the in vitro test method of the invention could be correlated with in vivo results, thus making in vitro results predictive of in vivo future behaviours.

This invention is thus of great advantage, as it is a very simple test method for the evaluation of the toxicity of test materials susceptible to enter into contact with the eye surface. It is important to bear in mind that ocular surface is constantly regenerated, with dividing cells being in a state that is mimicked by the scrap and the creation of the wound in the in vitro test method. This in vitro test method is also of great use for the evaluation of the healing properties of specific test materials designed to treat ocular lesions such as surgery, ulcers, dry eye, injury, scars, wounds, trauma, and lesions.

According to the bivalent method of the invention, toxicity and healing properties may be evaluated simultaneously or sequentially, or separately, on the same cultured cells batch. In a preferred embodiment, when the test method is implemented, both toxicity and healing properties are searched. The invention makes it possible to observe both degradations due to toxicity, and condition improvements due to healing properties, on the same in vitro model and/or batch. According to the Applicant's knowledge, this in vitro model is far more superior to any in vivo model in terms of cost and simplicity of use. Summary of the invention

This method is innovative, as it takes the normal behaviour of the corneal epithelial and conjunctival cells into account to characterize the toxic effect of material which may be applied on, or enter in contact with the eye surface. For that purpose cell toxicity is not only assessed by the sole survival state of the cells, as in classic and less sensitive in vitro tests, but mainly by the proper response of the cells to an extracellular challenge that precedes the application of the said material to be tested.

This invention thus relates to an in vitro method for testing toxicity and/or healing properties of a test material to an ocular surface including the steps of:

- performing a monolayered or multilayered culture of epithelial corneal or conjunctival cells including quiescent and dividing cells, said culture being obtained by cultivating cells to confluence or almost to 100% confluence followed by a scraping of the cell layer (s) resulting into a cleft,

- applying the test material onto said culture,

- measuring cell viability and the closure rate of the cleft within said culture and within the same culture treated with a reference solution, such as, for example a saline solution as a control.

According to an embodiment, the above-mentioned scraping results in a cleft having a width of 100 to 500 pm, more preferably 200 to 400 pm.

According to an embodiment, the test material is in a form selected from the group consisting of liquid formulations, such as for example aqueous solutions, eye drops, emulsions, oily solutions and micelles or fluid formulations, such as for example ointments, creams, lotions. Advantageously, the liquid formulation is selected from the group consisting of emulsions, aqueous or oily solutions, and micelles. According to another embodiment, the test material is an active principle.

According to a further embodiment, the test material is a pharmaceutical or a dermatological composition, such as for example a pharmaceutical composition including any active principle, such as for example prostaglandins or prostaglandin derivatives, such as for example latanoprost or derivative thereof, or an excipient .

According to another embodiment, the test material is an excipient, a detergent, a cosmetic product, such as for example a make-up product.

According to still another embodiment, the test material is selected from the group consisting of pharmaceutical compositions useful for enhancing healing of the ocular surface affected by surgery, ulcers, dry eye, injury, scars, wounds, traumas and lesions.

According to another embodiment, the test material is selected among the group of any new chemical entities (NCE) or excipient that may be applied onto the ocular surface or enter in, or in contact with, the eye through their normal or accidental use.

The main advantage of the method of the invention is that the toxicity of the test materials is evaluated on a cell monolayer including cells that are both in a quiescent state (cells far away from the scrap) and activated state (i.e. dividing cells close to the edge of the scrap) . Thus this test method is very useful and sensitive, as it can be used to assess the toxicity of the test material through the characterization of the healing properties (rate of cicatrisation) . Moreover, the test method can also be used to assess the efficacy of the test material at improving wound healing. It is therefore both a toxicity and an efficacy test.

Another object of this invention is a kit for implementation of the method of the invention.

Detailed description

This invention thus relates to an in vitro method for testing toxicity and/or healing properties of a test material to an ocular surface including a first step of performing a monolayered or multilayered culture of epithelial corneal or conjunctival cells including quiescent and dividing cells; for example 100 000 cells per well) may be plated in 6-well plates or any other appropriate support (12-, 24-, 48-, ... well plates) . The cells may be cultured for 24 h or until they reach the desired confluence; preferably, confluence is 80% or more.

At the desired confluence the cell monolayer or multilayer is scraped one time with a disposable tip or an appropriate standardized tool, in order to remove cells from a controlled surface area, thus resulting in a cleft. On the scraped surface 100% of the cells are removed. The width of the scraped surface ranges from 100 to 500 pm more preferably from 200 to 400 pm.

In a second step, the cell culture medium is replaced by fresh cell medium containing the desired amount of the test material. The incubation with the test material may last from a few seconds to 1 hour or more. A 30-minute incubation is generally sufficient to induce the desired response.

Once the incubation time is over, the cell medium containing the test material may be removed and replaced with fresh cell medium. The incubation in fresh cell medium may last up to 5 days, during which the scratch closure is assessed at predefined times ranging from 30 minutes to 5 days. Incubation times ranging from 30 minutes to 24 hours, preferably 12-20 hours, are generally sufficient prior to performing the assessment of toxicity and/or healing efficacy.

In a third step, a measure of cell viability and of the closure of the cleft within said culture is performed, and the same measure is carried out on the same culture treated with a reference solution, such as for example a saline solution as a control .

Measure of cell viability and of the closure of the cleft may be performed by any suitable means known by the skilled artisan. According to an embodiment, measure of cell viability and of the closure of the cleft may be assessed by any means that allows for an easy assessment of the scraping's closure (such as for example cell migration and multiplication) . Means for assessment of cell viability and/or healing properties, including assessment of the cleft, may include observation through a white light or polarized light inverted microscope (such as for example Olympus 1X70 fluorescence microscope) . The intensity of the contrast may be improved by the use of the appropriated dyes (such as trypan blue, bromodeoxyuridine ) , contrast agents or other probes for apoptosis or proliferation, to name only a few.

In order to search for sensitive markers of the healing mechanism, cells may also be treated with antibodies targeting proteins involved in healing mechanism (cytoskeleton proteins, transcription factors, scafolding proteins...) and quantified by immunofluorescence.

The test may be automation compatible. The observation of cells can be automated and computerized to obtain reproducible, objective and un-biased results.

Another object of the present invention is a kit for implementation of the test method as described hereabove. Advantageously, the kit of the invention comprises cells, a solution to be used as a negative control and a dye, preferably selected from the group comprising trypan blue and bromodeoxyuridine .

According to an embodiment, the kit of the invention comprises a vial of cells, preferably of cells selected from the group comprising epithelial corneal and conjunctival cells. According to an embodiment, the vial of cells is a frozen vial of cells. Advantageously, the number of cells in the vial of cells is at least 10000 cells, preferably at least 100000 cells, more preferably at least 1000000 cells. According to an embodiment, the kit of the invention comprises a concentrated solution of a negative control. In an embodiment, said solution is a saline solution. In another embodiment, said solution is a solution of BAK. In an embodiment, the solution is at a concentration of at least 0.1% w/v, preferably of at least 1% w/v, more preferably of at least 5% w/v.

According to an embodiment, the kit of the invention further comprises yellow tips and/or plates for cultivating cells, preferably 6-well plates. Definitions

Test material (s): any product, active principle, new chemical entity or excipient in the form of a liquid solution or in a form that can be applied on the cultured cells, which is tested according to the test method of the invention. This material may comprise any new or usual ingredients such as for example preservatives, osmotic agents, surfactants, buffer etc. It may comprise any emulsion, solution or suspension including-or not- an active ingredient or any formulation which can be applied on the eye surface such as eye drops, ointments, creams etc...

Epithelial and conjunctival toxicity: can be defined as the potential of a material to damage the cells, i.e. to be irritant on the eye surface and/or to create or prevent healing of lesions, including superficial lesions, ulcers, injury, scars, and wounds, on the epithelial corneal and conjunctival cells.

Brief description of the drawings

Figure 1. Pictures of cell cultures and graph, showing an evaluation of the sensitivity of the test method through the characterization of the dose-response curve of a known ocular excipient. (A) Representative pictures taken immediately after the scraping, at 2 hours, day 1 and day 3 after the scraping for control and various concentrations of benzalkonium chloride (BAK) . Note that the bar on the picture illustrates the size of the scratch at the beginning of the experiment. (B) Bar graph compilation of the scraping data. Note that for the high concentrations (0.01%BAK or higher) at the long time points the measurement of the wound closure was not possible since the toxicity was too important. *, p<0.03 when compared to CTL; #, p<0.02 compared to 0.001% BAK; $, p<0.003 when compared to 0.001% BAK, p<0.02 when compared to 0.005% BAK and p<0.02 when compared to 0.01% BAK.

Figure 2. Pictures of "wounded area" 2 hours after the scrapping, showing a toxicity evaluation of different treatments on cell survival and migration 2 hours post scraping and eye drop challenge.

Figure 3. Pictures of the 'wounded area' 24 hours after the scraping. Toxicity evaluation of different treatments on cell survival and migration 24 hours post scraping and eye drop challenge.

Figure 4. Bar graph compilation of the data at 0, 2, and 24 hours post scraping and eye drop challenge. *, p<0.01 when compared to PBS or Latanoprost Emulsion; **, p<0.002 when compared to PBS or Latanoprost Emulsion. Note that the greater the distance, the more toxic the compound.

Figure 5. Microscope pictures showing Ki67 immuno- histological identification of the treated cells. Note that dividing cells appear green on the pictures. Figure 6. Pictures at day 4 (at the end of the treatment period) of the cornea. In vivo correlation of in vitro toxicity data: clinical evaluation of the cornea. The white scar present for rat treated with 0.02% BAK or Xalatan are a clear indication of their toxicity. Figure 7. Graph showing the measured area of corneal opacity at the end of the treatment period. *, p<0.05 when compared to Xalatan or 0.02% BAK. Note that the greater the surface the more toxic the compound.

Figure 8. IVCM pictures of the corneal epithelium. Pictures were taken at day 0 (DO) and day 4 (D4) . Examples

Example 1: in vitro toxicity dose-response curve of a known ocular excipient

Example 1 is to be read with reference to Fig.l. Brief description of the protocol: human corneal epithelial (HCE) cells (100 000 cells per well) were plated in 6-well plates and cultured until the desired confluence. At confluence the cell monolayer was scraped with a yellow tip, the medium removed and incubated for 30 min with 1/10 dilutions of benzalkonium chloride dilutions:

The starting dilutions tested are: 0.001, 0.005, 0.01, 0.02, and 0.1%.

Following the treatment, the test material was removed and replaced by fresh culture medium. The scraping closure was assessed at times ranging from 2 hours to 3 days post treatment .

Results : the data demonstrated that the test method is very sensitive, since very low concentrations of BAK were able to affect the normal wound healing process (see Fig.lA) . The lowest dose (0.001%) of BAK delayed the wound closure by almost 2 days, indicating that even at this concentration BAK presented some toxicity. Interestingly, the first signs of toxicity are seen as soon as 2 hours after the scraping; as evidenced by a delayed wound closure (for example compare the control and the 0.001% BAK at time point 2 hours) . It is important to keep in mind that BAK is found in eye drops at concentrations as high as 0.02% BAK, i.e. a dose that is described in the literature as being responsible for the induction of deleterious side effects in patients (see for example Baudouin C, Labbe A, Liang H, Pauly A, Brignole- Baudouin F. Preservatives in eye drops: the good, the bad and the ugly. Prog Retin Eye Res. 2010;29:312-34). In this scraping assay, 0.02% BAK was clearly identified as a toxic agent for HCE cells. The bar graph compilation clearly illustrates that increasing the dose of BAK alter HCE cells integrity (see Fig. IB) . At doses of 0.01% or higher the wound healing process is dramatically reduced, and the toxicity evident. Thus, the scraping method is very sensitive as it was able to identify a known ocular excipient as a toxic compound, and this even at very low doses.

Example 2 : In vitro toxicity of PG analogue-containing eye drops

Example 2 is to be read with reference to Fig.2-5.

Brief description of the protocol: human corneal epithelial (HCE) cells (100 000 cells per well) were plated in 6-well plates and cultured until the desired confluence. At confluence the cell monolayer was scraped with a yellow tip, the medium removed and incubated for 30 min with 1/10 dilutions of the different test materials:

PBS (phosphate buffer solution) ,

Xalatan® (aqueous solution of latanoprost at 0.005%), 0.02% BAK (benzalkonium chloride aqueous solution), and Latanoprost-containing emulsion.

Following the treatment, test materials were removed and replaced by fresh culture medium. The scraping closure was assessed 24h post treatment.

Results : 2 and 24h after the scraping, pictures of the cell monolayer were taken to assess the efficacy of the different treatments on the scraping closure. Clearly, latanoprost-containing emulsion displays a positive effect on cell survival and migration (see Figures 2 & 3) . There were less dead cells, and the closure of the cleft was merely complete 24h post scraping with emulsion with latanoprost, while Xalatan, 0.02% BAK solution and even PBS displayed no or only little closure, therefore demonstrating the toxicity of Xalatan and 0.02% BAK solution. Hence, the method is at the same time able to discriminate compounds or formulations negatively affecting the wound healing process (i.e. toxic compounds), to others with positive effects on the wound healing process. Neutral compounds would have behaved as PBS, a buffer with neither toxicity nor beneficial effects on the wound healing process.

The compilation of these data for the different conditions in the test demonstrated that latanoprost-containing emulsion was able to increase the pace of wound closure, while both Xalatan and 0.02% BAK solution altered the process of wound healing. No closure was observed with the latter two test items (see Figures 4), as demonstrated by the % of wound closure 24h post scraping that remains very low, 6% and 8% for Xalatan® and the 0.02% BAK solution, respectively. By opposition, latanoprost-containing emulsion has % of wound closure of 85%. This is confirmed by immunohistology data for Ki67, a proliferation marker (see Figure 5) . More positive cells for Ki67 (i.e. dividing cells) were present close to the edge of the scraping wound in plates treated with latanoprost- containing emulsion. Thus, confirming the utility of the test method at discriminating in one single operation toxic compounds to others having beneficial effect on the wound healing process.

Example 3: In vivo and in vitro correlation experiment

Example 2 is to be read with reference to Fig.6-8. Brief description of the protocol: 16 male Sprague-Dawley rats weighting 100 to 125g were randomly assigned to 4 groups (4 rats per group) : Gpl, PBS; Gp2, 0.02% benzalkonium chloride (BAK) ; Gp3, Xalatan; Gp4, Latanoprost Emulsion. Following general and local anesthesia with ketamine/xylazine and topical Oxybucaine, respectively, the upper part of the right eye cornea of each rat was scraped with a surgical scalpel following an application of 50 μΐ of 50% ethanol solution. The upper corneal and limbal epithelia were removed (at day DO) . One drop of Tobrex was instilled immediately after the scraping. 2 and 5 hours post scraping one drop of the different test articles were applied onto the cornea of the right eye. These latter instillations were repeated every day for 4 days, and corneal evaluation with an in vivo confocal microscope (IVCM-HRT) was performed on day 4 after the last instillation .

Results : Day 4 corneal opacity demonstrated that scarification of the cornea was dramatic following repeated instillations with either Xalatan or a 0.02% BAK solution. By opposition, Latanoprost Emulsion of the invention favours corneal healing without these sight-threatening scars. Figure 6 presents pictures of treated cornea at day 4 following scraping and repeated instillations of the different test materials. Clearly, the latanoprost-containing emulsion identified in the in vitro test method to have beneficial effects on the wound healing process favors a safe corneal wound healing, as very little white scar tissue is present onto the cornea. These in vivo data also confirmed the in vitro toxicity observed for Xalatan and the 0.02% BAK solution, therefore confirming the powerfulness of the in vitro test method of the invention. The surface of this white area was measured at day 4 for each of the treated eyes (see Figure 7) . Clearly, the latanoprost-containing emulsion presents a very nice healing profile when compared to Xalatan. The simplicity of the in vitro test method of the invention and its very good correlation with in vivo data makes it a very useful tool to assess the toxicity of any compound that may be applied onto the ocular surface.

IVCM (in vivo confocal microscopy) data confirmed the macroscopic observation made by the measurement of the scar area (Figures 6 &7) . The beneficial effects of the latanoprost-containing emulsion seen in vitro are demonstrated by the lack of scar tissue on the cornea, which was indeed very well healed (no formation of scars tissue, the white areas seen on the pictures of Figure 8) . Note that the scar tissue (hypereflectivity ) seen in the pictures obtained at day 4 for Xalatan and the 0.02% BAK solution is the result of corneal edema and tissue disorganisation, thus clearly indicating the toxicity of both test materials. By opposition, the absence of this hypereflectivity in eyes treated with the latanoprost-containing emulsion demonstrates the beneficial effect of this test material.

Altogether, these data confirmed that the test materials identified in the in vitro test method are indeed toxic in an in vivo model, thus confirming the usefulness of this method.

Most importantly the strength of the method is demonstrated by the very nice correlation of these in vitro data with the in vivo data. In conclusion, the in vitro data demonstrated the simplicity of use of the method and its usefulness in the assessment of the toxicity of different eye drop preparations.