US Environmental Protection Agency
Office of Pesticide Programs

NON-ANIMAL TESTING APPROACH TO EPA LABELING FOR
EYE IRRITATION

May 11, 2009


-------
VOLUNTARY PILOT PROGRAM TO EVALUATE USE OF A NON-ANIMAL
TESTING APPROACH TO EPA LABELING FOR EYE IRRITATION FOR CERTAIN
AN TIMICROBIAL PRODUCTS WITH CLEANING CLAIMS; 5/11/2009; John
Redden, MS; Mark J. Perry, MPH; Timothy Leighton, Jonathan Chen, Ph.D.; Tim
McMahon, Ph.D.

I. INTRODUCTION

The EPA Office of Pesticide Programs (OPP) currently uses the Draize rabbit eye test to
determine ocular hazards, and the required hazard labeling for pesticide products. This voluntary
pilot project is designed to evaluate the effectiveness of a specific alternative testing approach, as
a potential replacement for the Draize rabbit eye test, for labeling antimicrobial products with
cleaning claims. The proposed testing strategy uses three assays; the Bovine Corneal Opacity
and Permeability test (BCOP), the EpiOcular™ model (EO), and the Cytosensor
Microphysiometer assay (CM).

This approach is intended to allow OPP to differentiate among the four eye irritation
hazard categories used by the Agency. These categories and the associated label statements are
listed below. Along with the three alternative assays, OPP is asking participating registrants to
submit available consumer incident data and any existing Draize test results on similar or
structurally-related chemicals or products as further support for the testing approach.

Table 1. Eye Irritation-Triggered Label Statements and Eye Protection.

Toxicity
Category

Signal Word

Eye Protection and Label Precautionary Language

I

DANGER

Goggles, face shield, or safety glasses. Corrosive. Causes
irreversible eye damage.

II

WARNING

Goggles, face shield, or safety glasses. Causes substantial
but temporary eye injury.

III

CAUTION

Protective eyewear if appropriate. Causes moderate
irritation.

IV

CAUTION

No statements are required.

II. OBJECTIVE

The objective of this voluntary pilot project is to evaluate and gain experience with
certain non-animal testing methods {i.e., ex vivo and in vitro) to assess eye irritation and generate
labeling for certain antimicrobial products with cleaning claims. The Agency, which has formed
a workgroup to manage this pilot, also intends to use the resulting data set to further evaluate the
individual non-animal assays. In addition, this pilot incorporates the goals of the 3 R's of animal
testing:

•	Refinement alternative: New or modified test method/s that refine/s procedures to
lessen or eliminate pain or distress in animals or enhances animal well-being.

•	Reduction alternative: New or modified test method/s that reduce/s the number of
animals required for a test method, while remaining consistent with sound scientific
practices necessary to obtain valid results.

1


-------
•	Replacement alternative: New or modified test method/s that replace/s animals with
non-animal systems or replace/s an animal species with a phylogenetically lower
species.

The pilot test strategy and supporting science have been reviewed and approved by the
EPA's Office of Pesticide Programs Science Policy Council (SciPoC). The Science Policy
Council consists of senior staff from all of the science and regulatory divisions and serves as a
central forum that assists OPP in reviewing and transitioning new science policies and methods
into the pesticide program. As mentioned above, the assays that will serve as the basis of the
pilot approach for eye irritation labeling include the Bovine Corneal Opacity and Permeability
test, the EpiOcular model and the Cytosensor Microphysiometer assay. These non-animal
assays, when used in a tiered approach, are being considered as a potential replacement to the
Draize rabbit test. A brief description of these assays follows:

•	Bovine Corneal Opacity and Permeability Assay (BCOP) - is an assay that uses
bovine (cow) eyes which are received shortly after the slaughter of the animal so the
cells are still viable. The corneas are excised and treated with a chemical to determine
its potential to damage the eye. The BCOP model is a model with endpoints similar to
many human corneal responses.

•	EpiOcular ™ (EO) Model - this test is an in vitro model of the human corneal
epithelium composed of normal human-derived epidermal keratinocytes and is used
to evaluate the eye irritation potential of chemicals, particularly surfactants.

•	Cytosensor Microphysiometer (CM) Assay - this assay evaluates the potential eye
toxicity of a chemical by measuring the dose required to reduce the metabolic rate in
treated cells in vitro. A very sensitive instrument called a microphysiometer is used to
electronically measure the metabolic rate of cell populations through small changes in
acidic metabolites in the medium. The rate is constant in an undamaged cell
population and if the cells are injured, an altered metabolic rate is found.

It is intended during the pilot phase that labeling decisions will be made using data
derived from these non-animal tests if the testing methods and testing results are deemed by the
Agency to be adequate and appropriate to support such regulatory decisions. Antimicrobial
products with cleaning claims may include formulations of different composition (water or
surfactant-based; oxidizing chemistry; solvent-containing; or non-aqueous soluble formulations).
Proposed testing strategies for these various types of formulations in this pilot may differ.

III. PILOT DETAILS
A. Scope and Duration

As mentioned previously, this pilot project is limited to antimicrobial products with
cleaning claims. Additionally, because this subset of pesticide products typically accounts for
over 100 registrations per year, for practical purposes the pilot will be limited initially to a
timeframe of 18 months. After 18 months a decision will be made to determine if: 1) the non-
animal testing strategy is valid and adequate, or 2) the pilot needs to be extended to gather more
data. Also, since one of the goals of the pilot is to collect non-animal testing data for further

2


-------
evaluation, OPP does not expect to generally allow data submitted under this pilot to be cited or
bridged from one product to another.

B.	Review Process

The review process for the pilot project will involve determining the effectiveness of the
decision tree (see Section IV), as well as whether submitted assays were conducted according to
the published guidance and are acceptable. Initially, packages submitted to the Agency under the
pilot will be reviewed by the workgroup. One member of the workgroup will perform the
primary review, while secondary review will be performed by one or more of the remaining
workgroup members. Any issues which arise during this initial phase of the pilot will be
discussed by the workgroup as a whole. Toxicity reviewers from the Antimicrobials Division
(AD), who will ultimately be responsible for assessing these packages, will participate in these
discussions. The assays will be reviewed for compliance with the following protocols (see
Appendix I, Annex I, II and III):

•	Ocular Irritation Assay for Antimicrobial products with cleaning claims using the
Bovine Corneal Opacity and Permeability Assay and Histology

•	Ocular Irritation Assay for Antimicrobial products with cleaning claims using the
EpiOcular™ Human Cell Construct

•	Ocular Irritation Assay for Antimicrobial products with cleaning claims using the
Cytosensor Microphysiometer Bioassay

Once the Workgroup has gained experience reviewing the studies, the primary and
secondary review responsibility will be turned over to the Antimicrobials Division. The
workgroup will meet regularly with the AD reviewers to discuss any issues related to the
assessment of these packages. If needed, the workgroup will be available at any point in the
process to perform secondary reviews of submitted data packages, or to discuss study-related
issues with registrants and/or performing laboratories.

C.	Pilot Project Assessment:

Eighteen months after the start of the pilot project, the workgroup will compile and
analyze the study data. This process will involve an evaluation of the resulting eye irritation
categories, the types of products received for review, and a comparison of the study findings
with submitted incident data and/or existing Draize data. Consumer incident data and Draize
data will be considered from the subject product undergoing registration, as well as from other
pesticidal and non-pesticidal products deemed similar to the subject product. Product similarity
will be verified by the workgroup prior to inclusion of incident or Draize data in the assessment
document. The final assessment document will be presented to OPP's Science Policy Council
upon completion for their recommendations on the way forward with this non-animal approach.

3


-------
IV. DECISION TREE

Yes



Yes

f BCOP











Category

I. 5, '

To separate
m from IV,
Go to A

Category
III IV,

To separate
II from I,
Go toB

4


-------
The first step in the decision tree is to evaluate the existing information (including the
chemical and physical properties) on the active ingredient/formulation, and existing Draize
results or in vitro data on related compounds. If the formulation is based on oxidizing (reactive)
chemistries or the components fall in a class of chemicals that are severe irritants, it should be
tested in the BCOP assay. If the formulation is not based on oxidizing (reactive) chemistries, a
scientific judgment should be made (based on the type or concentration of formulation
ingredients, past registration of similar products, in-use information from similar, non anti-
microbial products, etc.) as to the expected ocular hazard category of the formulation. This
decision does not affect the outcome of the evaluation. If the initial judgment is incorrect,
that fact will be revealed in the testing strategy and a second test may have to be used. For
example, if EpiOcular indicates Category I, then BCOP will need to be performed. The
initial judgment is only made for efficiency; if correct, only one assay will have to be
conducted.

If the material is a "high solvent" formulation, it should be tested with a 3 minute
exposure. All other materials are tested with a ten-minute exposure. A high solvent formulation
is considered to have equal to or greater than 5% concentrations of organic solvents (e.g.
alcohols, glycol ethers). Because of earlier indications that some solvent-containing materials
might be over predicted, IIVS - for the last several years - has tested such materials in the BCOP
assay using two different exposure times: 3 minutes and 10 minutes. Based on this experience it
has been generally noticed that the three minute exposure gives a better prediction of the actual
irritancy potential than does the 10 minute exposure (extracted from: Background Review
Document of an In Vitro Approach for EPA Toxicity Labeling of Anti-Microbial Cleaning
Products see: http://iccvam.niehs.nih.gov/methods/ocutox/PeerPanel09.htm).

If the formulation receives a score of >75 in the BCOP assay it is labeled as a Category I.
Although such formulations are placed in Category I, performance of histopathology is strongly
encouraged. If it receives a score of <75, histopathology should be performed and the depth of
injury should be used to determine if the final Category should be I, II, or a lower. The data
presented by the AMCP supports the selection of >75 for Category I as well as <75 (with
histopathology) for differentiating Category I and II (see BRD: Background Review Document
of an In Vitro Approach for EPA Toxicity Labeling Anti-Microbial Cleaning Products; Rodger
D. Curren, Ph.D.; Jennifer; R. Nash, M.S.; Angela Sizemore, B.S.; John Harbell, Ph.D).

Although it should be noted that a standard scoring scheme forjudging histopathology results
has not been established.

If it is thought that the product might be a Category III or IV material, companies may
chose to use either the CM or EO test which can confirm the moderate Category III or Category
IV materials. If at the beginning of the testing scheme it is thought that the product is likely a
mild/moderate product, either the CM or EO assays should be chosen. The CM, however, could
only be used with completely water soluble materials (limitation noted in data submitted to
ICCVAM for CM). The results from either of these tests would determine whether the material
was a Category I, III, or IV. If the company thought the material might actually be a Category II,
then an additional BCOP assay with histopathology should be conducted to determine if this is
the case and to support a different classification.

It should be noted that the above strategy is self-correcting if the initial estimate of
irritation potential of a test substance is incorrect. If a highly irritating material is tested in
the Cytosensor or EpiOcular assays, it will receive a score indicating that it is a highly irritating
(Category I) material. If further resolution is desired (to determine if it is actually only a

5


-------
Category II material rather than a Category I material), the formulation can then be further tested
in the BCOP assay. Similarly a mild material will be identified as a Category III material by the
BCOP assay. If it is important to the company to distinguish between a Category III and IV for
labeling and marketing purposes, then an additional Cytosensor or EpiOcular assay may be
helpful in making that determination.

The proposed in vitro strategy is considered conservative and generally results in over
labeling of some products, especially many EPA category IV materials which could be over-
predicted to be EPA category III. The participating companies are aware of the potential for
over-predictions and have accepted it as a small consequence of adopting non-animal testing
strategy.

V. SUBMISSION PACKAGE GUIDANCE

A.	General Package

•	Raw data on the non-animal assays (BCOP, EO, and/or CM).

•	Available Draize rabbit test results for similar or structurally related compounds.

•	Available consumer incident data on the tested product or similar products/formulations.
Incident data should be provided on similar EPA-registered pesticide products, as well as
similar unregistered consumer products, if available. The following should also be
reported if applicable:

¦	EPA File Symbol/Registration Number

¦	Active ingredient(s)/PC Code(s) associated with each incident

¦	Symptoms/clinical effects

¦	Specific effects should be specified whenever possible/available

¦	Duration of symptoms

¦	Symptom reversibility

¦	Medical outcome/severity of incident

¦	Level of health care required (another indicator of severity/burden)

¦	Exposure route

¦	Exposure site

¦	Exposure reason

¦	Age

¦	Sex

¦	Year of incident

•	Any other useful existing knowledge (e.g., chemical physical properties, other data on
irritancy, Structural Activity Relationship (SAR) data on irritancy) for ocular hazard
labeling

B.	Data and Reporting

The study report should include a description of the test material, the methods and the
study results. At a minimum, all data should be detailed in tabular form for each individual
protocol and the following reported:

6


-------
•	physical nature, and, where appropriate, concentration and pH value for the test
substance;

•	description of any pre-test conditioning;

•	manufacturer, source, purity, and lot number of test substance;

•	Good Laboratory Standard (GLP) standards employed;

•	rationale for selection of test (BCOP, EpiOcular or Cytosensor);

•	identification, composition, and characteristics of any vehicles (e.g., diluents, suspending
agents, emulsifiers, and anesthetics) or other materials used in administering the test
substance;

•	a list of references cited in the body of the report, i.e., references to any published
literature used in developing the test protocol; performing the testing, making and
interpreting observations, and compiling and evaluating the results;

•	description of the method used to score the irritation;

•	description of any lesions observed (BCOP);

•	any effects other than ocular which were observed;

•	narrative description of the degree and nature of irritation or corrosion observed, and;

•	a tabular description irritant/corrosive* response data for each individual test.

* Eye corrosion is the production of irreversible tissue damage in the eye following application of a test substance to
the anterior surface of the eye. Eye irritation is the production of reversible changes in the eye following application
of a test substance to the anterior surface of the eye. (Note: Reversible changes can not be measured ex vivo.
However it is proposed that a score less than 75 in the BCOP with histopathology and the weight of evidence from
other sources will allow this determination.)

VI. SCIENTIFIC BASIS

A. Regulatory Background

As a result of discussions on the use of non-animal testing methods at the fall 2003
Pesticide Programs Dialogue Committee (PPDC), the Alternative Testing Working Group
(ATWG) was formed with the goal of developing a non-animal eye irritation testing approach for
antimicrobial products with cleaning claims. In 2004 the ATWG, which is comprised of industry
representatives from the PPDC, developed an approach which uses three tests (BCOP, EO and
CM) to determine an ocular irritation category. Subsequently, EPA's Office of Pesticide
Programs (OPP) requested the assistance of the National Toxicology Program's Interagency
Coordinating Committee on the Validation of Alternative Methods (ICCVAM) with the
evaluation of this approach. The ICCVAM formed the Ocular Toxicity Working Group
(OTWG) which has proposed an alternate strategy employing only the BCOP and EO assays to
assign materials to categories I and IV, respectively. A peer review of the OTWG approach is
scheduled for May 2009 at the National Institutes of Health.

The OTWG chose not to include the Cytosensor Microphysiometer in their proposed
strategy due to its reliance on Low Volume Eye Test (LVET) data for validation. The LVET,
which was developed as an alternative to Draize, involves application of 0.1 ml of test material
directly on the cornea of the test animal. Based on studies which indicate that LVET data under-
predicts severe irritants when compared to the Draize, the OTWG concluded that the LVET is
not an acceptable in vivo reference test method against which to compare in vitro test method
results. However, a recent European Centre for the Validation of Alternative Testing (ECVAM)
evaluation of Cytosensor Microphysiometer supports its use to classify category I and category
IV ocular irritants based on non-LVET reference data. Following the ECVAM evaluation, the
OTWG concurred with this decision.

7


-------
In addition, the OTWG concluded that the histopathology data for the BCOP did not give
the degree of sensitivity expected, and as a result, recommended that the BCOP only be used for
severe irritants (Toxicity Category I). It was noted that based on a limited data set (n=29) using
BCOP to identify corrosives followed by EpiOcular to identify non-irritants appears to be an
effective strategy for these two categories. The OTWG further concluded that there are
inadequate data to determine the usefulness of this strategy for identifying mild and
moderated irritants.

Additional data using BCOP, EpiOcular and Cytosensor Microphysiometer should be
submitted by the registrant community to the Agency. As noted above, there will be a May 2009
ICCVAM peer review to analyze the proposed strategy using BCOP, EpiOcular and Cytosensor.
Additional data from this pilot would strengthen the case for finalizing this pilot testing strategy
for certain antimicrobial products with cleaning claims. Before initiating this pilot, the Agency
has reviewed the data and concluded that the evidence available supports that this approach will
identify Toxicity Category I, II, III and IV labeling for eye irritation.

B.	Agency Conclusions on LVET Data

The LVET is not a guideline study, and the Agency does not consider it sufficient, by
itself, to satisfy the eye irritation data requirement for pesticides. The Agency concurs with the
OTWG that the LVET is not an acceptable in vivo reference test method against which to
compare in vitro Test Method results.

C.	Cytosensor

Because Cytosensor is considered, as reported at ICCVAM/OTWG, useful for identifying
ocular corrosives/severe irritants (top-down approach) and nonirritants (bottom-up approach)
specifically for surfactants and surfactant-based formulations, and because many antimicrobial
products with cleaning claims are surfactant containing formulations, SciPoc believes that
Cytosensor may be useful in a testing strategy for identifying ocular irritation.

As part of the testing scheme, chemical physical properties of the chemical of interest,
registrant gathered consumer incident data, and existing Draize test results on structurally related
compounds would be assessed to evaluate labeling decisions regarding product safety. This
weight of evidence approach should allow greater confidence in the results reported in the in
vitro and ex vivo studies.

D.	BCOP Decision Criteria

BCOP is not intended to characterize EPA Toxicity Category IV. The AMCP BRD (see
reference, Section VII) gives the following guidance for when to do histopathology for BCOP.
The histopathology endpoint has often been added to the BCOP test to allow the actual extent of
damage to the cornea to be visualized and assessed. The initial area and depth of injury to the
cornea has been hypothesized to be directly related to the reversibility of the injury. The AMCP
BRD concluded that the greater the depth of injury, the less likely that the lesion is reversible.

8


-------
Table 2: AMCP BRD recommendations

In Vitro Score

EPA Category

GHS Category

Histopathology

>75

I

1

No histopathology needs to be conducted

<75 and >25

II

2A

They should be further assessed with a
histopathological evaluation and given the final
categorization of whichever determination (in
vitro score or histological evaluation) is more
severe.

<25

III

2B

They should be assessed with a histological
evaluation and given the final categorization of
whichever determination (in vitro score or
histological evaluation) is more severe.

Table 3: AMCP BRD Histo

jathology Decision Criteria

Extent of Damage

Suggested EPA Category

Suggested GHS Category

Cell loss or damage
extending no further than
midway through the
epithelium

IV

NL

Cellular damage or collagen
matrix damage extending
no further than the upper
third of the stroma

III

2B

Cellular damage or collagen
matrix damage extending
no further than two-thirds of
the way through the stroma

II

2A

Cellular damage or collagen
matrix damage extending
into the lower third of the
stroma and/or causing
damage o the endothelial
cells

I

1

According to the BCOP BRD "The in vitro irritation classification schemes used for this
evaluation were based on two different predetermined ranges of in vitro scores. The differences
between the two ranges are attributed to two different criteria used to identify ocular corrosives
and severe irritants (i.e., EPA Category I, EU R41, and GHS Category 1)." One approach
(Table 4) included the ICCVAM recommended decision criteria for identifying an ocular
corrosive/severe irritant (i.e., IVIS > 55.1, ICCVAM 2006).

Table 4. In Vitro Ocular Irritancy Classification Scheme for the BCOP Test Method
(ICCVAM 2006)

In I nro Score Range
0-3.0
3.1 - 25
25.1 - 55
>55.1

In I nro Classification
Not Labeled
Mild irritant
Moderate irritant
Severe irritant

9


-------
The second approach (Table 5) included an alternative decision criteria used for identifying an
ocular corrosive/severe irritant in the AMCP BRD submission (i.e., IVIS > 75).

Table 5. In Vitro Ocular Irritancy Classification Scheme for the BCOP Test Method
(AMCP BRD Submission)

E.	OTWG/ICCVAM Conclusions Regarding BCOP Histopathology

For 10 out of 17 of the AMCP materials, the EPA category classification using only
BCOP was inconsistent with the EPA Category determined by Draize. For 11 out of 17 of the
AMCP materials, the final EPA Category (using BCOP with histopathology) was inconsistent
with the EPA Category by Draize. It appears that histopathology does not improve BCOP
accuracy, but this observation is based on very limited data (N=17). An important aspect of this
pilot project is to evaluate additional data to determine if modifying the histopathology decision
criteria could improve accuracy.

F.	ICCVAM-Proposed Alternative AMCP Strategy Using BCOP and EpiOcular

Two strategies have been evaluated by NICEATM/ICCVAM:

1.	Test in BCOP first ("top-down" strategy); Any substance classified as EPA Category I or II
based on BCOP would be classified as such and no further testing required. All other substances
would be subsequently tested in EpiOcular to classify as either EPA Category III or IV.

2.	Test in EpiOcular first ("bottom-up" strategy); Any substance classified as EPA Category III
or IV based on EpiOcular would be classified as such and no further testing required. All other
substances would be subsequently tested in BCOP to classify as either EPA Category I or II.

NOTE: Further Draft recommendation from OTWG: 1) overall database for BCOP and
EpiOcular indicate that BCOP is reliable in identifying Category I, while EpiOcular is reliable in
identifying Category IV; and 2) there are insufficient data with which to adequately demonstrate
that the ICCVAM-proposed alternative AMCP testing strategy can identify all four required EPA
hazard categories for ocular irritation/corrosion.

G.	Details on Determining Hazard Categories.

The proposed strategy is intended to identify Categories I-IV materials using a
combination of assays and a weight of the evidence approach, which is strengthened by available
consumer incident data and/or existing Draize test results on similar or structurally-related
chemicals. Using this approach, the EO and CM assays are intended to identify Category III and
IV materials, while the BCOP assay is intended to identify Category I and II materials. If the
Agency concludes that the data submitted is not convincing then the Agency will make the more
conservative decision.

In I nro Score Raiiiic

0-3.0
3.1 - 25
25.1 - 74.9

>75

In I nro Classification
Not Labeled
Mild irritant
Moderate irritant
Severe irritant

10


-------
Cytosensor

OPP believes that the CM can identify Category III materials. As can be seen in the
following graph of 105 materials, materials which scored between 2 mg/ml and 80 mg/ml are
designated as Category III materials (Figure 1). They are clearly less toxic than the Category I
and II materials which all scored below 2 mg/ml.

500-1

100-

ui

E

a
a.

10-:

1-.

0.1-

¦	Solvents

A	Surfactants

v	Bases

~	Acids

18
82
4
1

Cytosensor vs. EPA(LVET)

¦ 2 mg/mL Cutoff
- 80 mg/mL Cutoff

Predicted as
Category IV

N=105

¦ A

Dual animal trials:

•	Material 1022
~ Material 1056

~	Material 1079

Predicted as
Category III

VA

*A
¦ A

OB"V

A
A

ft*

A

K

"AA

A

~

A_

A*a
A

AaaaA

i

AA

$A



AAA

Predicted as
Category I

IV

EPA Category (LVET)

Figure 1. Cytosensor MRD50 values plotted against EPA toxicity categories determined by the
LVET. Suggested cut-off values with their predicted EPA categories are included. There are 105
unique materials; however, 3 materials are graphed with 2 different EPA categories since they
were tested twice in the animal trials with different results each time (BRD).

EpiOcular

OPP believes that the EO assay can identify Category III materials. As can be seen in
Figure 2, materials having a score between 4 and 70 minutes would be considered Category III.
Admittedly there were not an overwhelming number of materials tested, but an additional 25
materials which had LVET data showed the same pattern.

11


-------
EpiOcular vs EPA Category (LVET)

>240
100

j v Surfactants 12
¦---Cut-off = 70 min
. -—'Cut-off = 4 min

Predicted as
Category IV



N=16





~

a



Predicted as

V

*
V

-

Category III

V

V



V







-

Predicted as

V

~



Category I

V

~

10-

1-

0.1

IV

EPA Category (LVET)

Figure 2. EpiOcular ET50 values plotted against EPA categories determined by the LVET.
Oxidizers have been removed since they will be tested only in the BCOP assay. Suggested cut-off
values with their predicted EPA categories are included (BRD).

BCOP

OPP believes that the BCOP assay can identify Category II materials. As can be seen in
the Figure 3, all materials scoring between 25 and 75 are considered Category II materials. These
are clearly different than the Category I's (>75) or Category Ill's (<25). In order to assure
accurate classification for Category II and Category III products, EPA is also asking participating
registrants to submit the consumer incident data and existing Draize test results on related
compounds for EPA to further evaluate the reliability of this interim pilot project.

12


-------
600-i
500-
400-
300-
200-
100-L

BCOP Scores vs. EPA Category

All materials except High Solvents
(10 minute exposure)

~~

~

~

ABS
~

Predicted as

¦ Solvents - 10 min	6

~ Surfactants - 10 min	15

~	Bases - 10 min	10

•	Acids - 10 min	6
~ Other-10 min	1

*	Oxidizers - 10 min	15

	Cut-off = 25

N=53

T

Category 1



		 Cut-off = 75

±A

-



~



~

•

Predicted as
Category II

¦

~



~







A

¦







~ Predicted as
Category III

¦¦
¦?

~ BR
TT

•

1

IV

EPA Category (Draize)

Figure 3. BCOP in vitro scores for non-High Solvent materials plotted against EPA categories
determined by the Draize test. Proposed cut-off values with their predicted EPA categories are
included. The EPA categories for test materials BR and BS were determined by using the results
of an LVET assay (BRD).

13


-------
REFERENCES

AMCP BRD: Background Review Document of an In Vitro Approach for EPA Toxicity
Labeling Anti-Microbial Cleaning Products (AMCP); Rodger D. Curren, Ph.D.; Jennifer; R.
Nash, M.S.; Angela Sizemore, B.S.; John Harbell, Ph.D.

Berridge, M.V., Tan, A.S., McCoy, K.D., Wang, R. (1996) The Biochemical and Cellular Basis
of Cell Proliferation Assays That Use Tetrazolium Salts. Biochemica 4:14-19.

Botham, P.A., Osborne, R., Atkinson, K., Carr, G., Cottin, M., and Van Buskirk, R.G. (1997)
IRAG working group 3: cell function-based assays. In: Eye Irritation Testing: Practical
Applications of Non-Whole Animal Alternatives. Food And Chemical Toxicology 37:67-77.

Bruner, L.H., D.J. Kain, D.A. Roberts and Parker, R.D. (1991). Evaluation of seven in vitro
alternatives for ocular safety testing. Fundamental and Applied Toxicology 17: 136-149.

Buehler, E.V. and Newmann, E.A. A Comparison of Eye Irritation in Monkeys and Rabbits.

Toxicology and Applied Pharmacology 6:701-

710(1964).

Curren R., Ph.D.; 8/26/08 presentation to ICCVAM's Ocular Toxicity Working Group. Co-
author, Background Review Document of an In Vitro Approach for EPA Toxicity Labeling of
Anti-Microbial Cleaning Products.

Curren, R., Evans, M., Raabe, H., Dobson, T., and Harbell, J.(1999) Optimization of the bovine
corneal opacity and permeability assay: histopathology aids understanding of the EC/HO false
negative materials. ATLA 27:344.

Gautheron, P.D., Dukic, M., Alix, D., and Sina, J.F. (1992) Bovine Corneal Opacity and
Permeability Test: An in Vitro Assay of Ocular Irritancy. Fundamental and Applied Toxicology
18:442-449.

Harbell, J.W., Osborne, R., Carr, G.J., and Peterson, A. (1997) Assessment of the
Cytosensor microphysiometer assay in the COLIPA in vitro eye irritation validation study.
Submitted, Toxicology In Vitro.

Harbell, J.W., Raabe, H.A., Evans, M.G., and Curren, R.D. (1999) Histopathology associated
with opacity and permeability changes in bovine corneas in vitro. The Toxicologist 48:336-337.

McConnel, H.M., Owicki, J.C., Parce, J.W., Miller, D.L., Baxter, G.T., Wada, H.G., and
Pitchford, S. (1992) The Cytosensor microphysiometer: biological applications of silicon
technology. Science 257:1906-1912.

MTT Effective Time 50 (ET-50) Protocol, MatTek Corporation

Parce, J.W., Owicki, J.C., Kercso, K M., Sigal, G.B., Wada, H.G., Muir, V.C., Bousse,
L.J., Ross, K.L., Sikic, B.I, McConnell, H.M. (1989) Detection of cell-affecting agents with a
silicon biosensor. Science 246: 243-247.

Sina, J.F., Galer, D.M., Sussman, R.G., Gautheron, P.D., Sargent, E.V., Leong, B., Shah,
P. V., Curren, R.D., and Miller, K. (1995) A collaborative evaluation of seven alternatives to the

14


-------
Draize eye irritation test using pharmaceutical intermediates. Fundamental and Applied
Toxicology 26:20-31.

15


-------
APPENDIX I

Each of the three in vitro test methods has a slightly different area of strength. The CM and EO
assays are very sensitive to small amounts of damage, and can be useful in the less irritancy side
of the toxicity range. For example, they can separate EPA Category IV materials from EPA
Category III and higher materials. In contrast, the BCOP assay uses a much more robust tissue
and can be useful to correctly distinguish between EPA Category I and II materials. The CM
assay especially is not useful to differentiate between Category I and Category II materials since
virtually all of the tissue would be destroyed by just a Category II product; more toxicity caused
by a Category I material could not be measured.

As stated above, generally only one test should be necessary to determine the appropriate hazard
classification, but in some cases the registrant may fine tune the final hazard classification by
using a second test that is either more robust or more sensitive. For example, if BCOP assay
indicated that the AMCP should have only a Category III or less classification, the EO or CM
assay could then be employed to help make the final decision of whether the product was
actually a Category III material or a Category IV material.

Antimicrobial products with cleaning claims can be formulated in different ways depending on
the desired cleaning capacity for the product. For example, the formulation can rely on its
alkaline or acidic properties for cleaning, or on surfactants and solvents, or on oxidizing
(reactive) chemistries. Most of these classes react similarly in the in vitro assays, i.e. the hazard
categories of the various types of formulations are similarly predicted. However, formulations
with oxidizing (reactive) chemistries and those with a high solvent concentration (>5%) should
be treated somewhat differently from the others (see below). It is also useful to determine the
water solubility of the formulation since only fully water soluble materials can be tested in the
Cytosensor Microphysiometer (limitation noted in data submitted to ICCVAM for CM)..

In summary:

•	Three in vitro assays are included in the pilot program.

•	The in vitro assays were selected to address the commonly recognized modes of action
for eye irritants.

•	The assays complement each other to cover the range of irritancy potential.

•	The testing strategy is consistent with current practice (using an up/down approach) and
with the proposed ECVAM strategy developed with input from ICCVAM
representatives.

AREA OF APPLICATION

The testing strategy described in this document is proposed for use only with certain
antimicrobial products with cleaning claims. Antimicrobial products with cleaning claims are
defined as either water or surfactant-based; excluding solvents.

There are some differences in how some antimicrobial products with cleaning claims are tested,
depending on certain characteristics of their formulation. These are:

16


-------
1)	Oxidizing materials

Formulations containing specific reactive chemicals, e.g., hypochlorite, peroxide,
percarbonate, oxygen bleaches, etc.

These materials are best tested in the BCOP assay (i.e. usually severe ocular irritants).
Because of the very reactive nature of these formulations, they are often over predicted
by the EO and CM assays, where there is less substrate available to bind the reactive
materials than in the human or animal cornea. In addition there are some delayed effects
which can only be visualized in the longer-term BCOP assay.

2)	"High Solvent" formulations

Formulations having >5% concentrations of organic solvents, e.g. alcohols, glycol ethers,
etc.

These materials often can be over predicted in the BCOP assay when using the
conventional exposure time of 10 minutes (see data presented in Appendix B -
Background Review Document of an In Vitro Approach for EPA Toxicity Labeling of
Antimicrobial products with cleaning claims, and ICCVAM recommendations for the use
of the BCOP assay). When such materials are tested in the BCOP assay a shorter
exposure time of 3 minutes should be used (addressed earlier in document).

3)	Non-aqueous soluble formulations

These materials should only be tested in the EO or BCOP assays (i.e. usually a slight or
nonirritant). Because of physical constraints imposed by the pumps and tubing used to
circulate the test material, only fully water-soluble materials can be tested in the CM.

17


-------
ANNEX I: OCULAR IRRITATION ASSAY FOR CERTAIN ANTIMICROBIAL

PRODUCTS MAKING CLEANING CLAIMS USING THE BOVINE CORNEAL
OPACITY AND PERMEABILITY ASSAY AND HISTOLOGY (Courtesy of
Institute for In Vitro Sciences, Inc.)

1.0 PURPOSE

The purpose of this study is to evaluate the potential ocular irritancy/toxicity of a test
article as measured by the test article's ability to cause opacity and/or permeability in an
isolated bovine cornea.

2.0 SPONSOR

2.1	Name:

2.2	Address:

2.3	Representative:

3.0 IDENTIFICATION OF TEST AND CONTROL SUBSTANCES

3.1	Test Articles:	1

3.2	Controls:	Positive: Ethanol(CAS #64-17-5) Neat

Negative: Sterile deionized water

3.3	Determination of Strength, Purity, etc.

4.0 TESTING FACILITY AND KEY PERSONNEL

4.1	Name:

4.2	Address:

4.3	Study Director:

4.4	GLP: 40 CFR Part 160 Good Laboratory Practice Standards (GLP) apply to this
assay

TEST

SCHEDULE



5.1

Proposed Experimental

Initiation Date:

5.2

Proposed Experimental

Completion Date

5.3

Proposed Report Date:



18


-------
TEST SYSTEM

The test system (target tissue) is the isolated bovine cornea obtained as a by-product from
freshly slaughtered animals. The procedures for preparing and handling the test system
were developed by Gautheron et al. (1992). The assay measures three components which
are predictive of eye irritation: corneal opacity, permeability, and tissue architecture.

Each cornea holder will be uniquely identified with a number written in permanent
marker, on both the anterior and posterior chambers. The treatment of each cornea will be
identified with the test article number (or control) written in permanent marker on
colored tape, affixed to each holder. Furthermore, the depth and degree of injury is
assessed by histological evaluation.

EXPERIMENTAL DESIGN AND METHODOLOGY

Liquid test articles will be tested neat unless otherwise directed by the Sponsor. The pH
of each liquid test article (or diluted test article) will be determined, if possible, and
recorded. Two or three corneas treated with sterile deionized water will serve as the
negative control. Two or three corneas will be exposed to the positive control. Three
corneas will be treated with each test article at each exposure time.

7.1	Reagents:

7.1.1	Hanks' Balanced Salt Solution with Ca++ and Mg++ (HBSS)

(containing Penicillin/Streptomycin)

7.1.2	Fetal Bovine Serum (FBS)

7.1.3	Minimum Essential Medium (EMEM) without phenol red

7.1.4	Complete MEM: EMEM without phenol red containing 1%FBS and
2mM L-glutamine

7.1.5	Minimum Essential Medium (EMEM) with phenol red
(used for rinsing test substances from corneas only)

7.1.6	Complete MEM: EMEM with phenol red containing 1% FBS and 2mM
L-glutamine

7.1.7	Sodium Fluorescein - diluted in DPBS

7.1.8	Sterile Deionized Water

7.1.9	10% Buffered formalin solution

7.2	Bovine Eyes

Bovine eyes will be obtained from a local abattoir or other commercial supplier.
The eyes will be excised by an abattoir employee (as soon after slaughter as
possible) and held in HBSS on ice. Once the required number of eyes has been
obtained, the eyes will be transported to the testing facility. Immediately upon

19


-------
receipt of the eyes into the laboratory, preparation of the corneas will be initiated.

7.3	Preparation of Corneas

All eyes will be carefully examined for defects (opacity, scratches, pigmentation,
etc.) and those exhibiting defects discarded. The tissue surrounding the eyeball
will be carefully pulled away and the cornea will be excised leaving a 2 to 3 mm
rim of sclera. The isolated corneas will be stored in a petri dish containing HBSS
prior to mounting. Corneas will then be mounted in the corneal holders with the
endothelial side against the O-ring of the posterior chamber. The anterior
chamber will then be positioned on top of the cornea and tightened with screws.
The chambers of the corneal holder will then be filled with EMEM (without
phenol red) containing 1% FBS (Complete MEM). The posterior chamber will
always be filled first. The corneas will be incubated for the minimum of one hour
at 32±1°C.

7.4	Sample Preparations

Liquid test articles will be tested neat whenever possible. When appropriate, test
articles will be diluted or suspended in either sterile deionized water or other
Sponsor-directed solvent. Samples will be diluted on a w/v basis, unless
otherwise specified by the Sponsor.

7.5	Initial Opacity Reading

At the end of the one-hour incubation period, the medium will be removed from
both chambers and replaced with fresh Complete MEM. An initial opacity
measurement will be performed on each of the corneas. Two or three corneas
with opacity readings approximately equivalent to the median opacity of all
corneas will be selected as the negative control corneas. The opacity of each
cornea (including the negative control corneas) will be read against an air-filled
chamber and recorded. Corneas that have an initial opacity reading that is 10 or
more units greater or lower than the average opacity of all used corneas will not
be dosed. The medium will be removed from the anterior chamber and replaced
with the test article, negative control, or positive control.

7.6	Treatment of Corneas

Antimicrobial products making cleaning claims described by the Sponsor as being
a High Solvent (defined as having a solvent concentration of >5%) will be tested
as in 7.6.1 below, with the exception that the exposure time will be 3 minutes.

7.6.1 Method A: Liquids

Liquids will be tested undiluted, unless otherwise directed by the Sponsor.
At least three corneas will be dosed per material. Approximately seven
hundred and fifty |iL of test substance (test article, negative control or
positive control) will be introduced into the anterior chamber. Highly
viscous materials will be applied directly to the corneal surface. The
holder will be slightly rotated (with the corneas maintained in a horizontal
position) to ensure uniform distribution over the cornea. The test article

20


-------
treated corneas will be exposed for 10 minutes at 32±1°C. The negative
and positive controls will be tested for 10 minutes also. At the end of the
exposure time, the test substance will be removed and the epithelium will
be washed at least 3 times (or until no visual evidence of test substance
can be observed) with complete MEM (containing phenol red). Once the
media is free of test substance, the corneas will be given a final rinse with
complete MEM (without phenol red). If the test article cannot be removed
from the cornea a note will be documented in the raw data record. The
anterior chamber will then be refilled with fresh complete MEM without
phenol red and an opacity measurement will be performed. The corneas
will then be incubated for a total of approximately 2 hours at 32±1°C. At
the completion of the incubation period, a second measure of opacity will
be performed (final opacity). The values obtained at this second
measurement will be used in calculating the corneal opacity.

7.6.2 Method B: Solids

Solid materials will generally be tested as a 20% dilution (w/v) in sterile
deionized water (or Sponsor directed solvent). Different concentrations
may be evaluated at the Sponsor's request.

Seven hundred and fifty |iL of test substance (test article, negative control
or positive control) will be introduced into the anterior chamber. The
holder will be slightly rotated (with the corneas maintained in a horizontal
position) to ensure uniform distribution of the test substance over the
cornea. The corneas will be incubated in a horizontal position at 32±1°C
for approximately 4 hours or as specified by the Sponsor. The test
substance will then be removed and the epithelium washed at least 3 times
(or until no visual evidence of test substance can be observed) with
complete MEM (containing phenol red). Once the media is free of test
substance, the corneas will be given a final rinse with complete MEM
(without phenol red). If the test article cannot be removed from the cornea
a note will be recorded in the raw data record. The anterior and the
posterior chambers will then be refilled with fresh complete MEM without
phenol red, and an opacity measurement performed immediately (without
any further incubation)(final opacity).

7.7	Opacity Measurement

The opacitometer will determine the difference in the light transmission between
each treated or control cornea and an air-filled chamber, and a numerical opacity
value (arbitrary unit) will be displayed and recorded.

7.8	Permeability Determinations

Method A: Liquids

After the second opacity measurement is performed, the medium will be removed
from both chambers of the holder. The posterior chamber will be refilled with
fresh complete MEM without phenol red. One mL of a 4 mg/mL fluorescein
solution will be added to the anterior chamber.

21


-------
Method B: Solids

After the opacity measurement is performed, the medium will be removed from
the anterior chamber only and replaced with 1 mL of a 5 mg/mL fluorescein
solution.

After the addition of the fluorescein solution to the anterior chamber, the corneas
will be incubated in a horizontal position for approximately 90 minutes at 32±1°C.
The medium from the posterior chamber will be removed at the completion of the
incubation period, and 360 |iL will be transferred to the appropriate wells of a
prelabeled 96-well plate. Three hundred and sixty |iL of fresh Complete MEM
without phenol red will be added to the wells designated as blanks. The optical
density at 490 nm (OD490) will be determined using a spectrophotometer. Samples
reading 1.500 and above (OD490) will be diluted to bring the reading within the
linear range of the platereader and the plate read again.

7.9	Fixation of the Corneas

After the medium is removed for the fluorescein determination, each cornea will
be carefully removed from its holder and transferred to a prelabelled tissue
cassette. The endothelial surface will be placed on a sponge to protect it. The
cassettes will be placed in 10% neutral buffered formalin and fixed for a
minimum of 24 hours.

7.10	Hi stol ogi cal Evaluati on
7.10.1 Name of Evaluator

7.10.2 The fixed tissues will be transferred to the pathology laboratory for
embedding, sectioning, staining and histological evaluation. If the
histological evaluation is conducted off-site, a Principal Investigator will
be assigned by the sub-contractor. Each cornea will be bisected and a
section from each half will be cut, placed in a cassette and embedded in
paraffin to produce a single slide. Each slide will then be stained with
hematoxylin and eosin. Cornea sections will be examined for the presence
of changes in the epithelium, stromal, and endothelial areas of the tissue.
Particular emphasis will be placed on assessment of depth of injury into
the stromal elements (Harbell et al, 1999 , and Curren et al, 1999).

Treated tissues will be compared to the negative and positive control
tissues. Representative fields will be photographed for illustration of the
changes.

CRITERIA FOR DETERMINATION OF A VALID TEST

The test will be accepted if the positive control causes an in vitro score that falls within
two standard deviations of the historical mean.

22


-------
9.0 EVALUATION OF TEST RESULTS

The change in opacity for each cornea (including the negative control corneas) will be
calculated by subtracting the initial opacity reading from the final opacity reading. These
values will then be corrected by subtracting from each the average change in opacity
observed for the negative control corneas. The mean opacity value for each treatment will
be calculated by averaging the corrected opacity values of each cornea for a given
treatment.

The mean OD490 value of each treatment group will be calculated by averaging the OD490
values of the treated corneas (less the average negative control values) for each treatment
condition.

9.1	In Vitro Score Calculation

The following formula was used to determine the in vitro score:

In Vitro Score = Mean Opacity Value + (15 x Mean OD490 Value)

9.2	Data Interpretation

All antimicrobial products making cleaning claims having an In Vitro Score of
>75 will be classified as an EPA Category I. Antimicrobial products making
cleaning claims having an In Vitro Score <75 and >25 are given a preliminary
classification of EPA Category II, but will be further assessed with a
histopathological examination (as part of the weight of evidence) and given the
final categorization of whatever determination (In Vitro Score or histopathology)
is more severe. Antimicrobial products making cleaning claims having an In Vitro
Score <25 are given a preliminary classification of EPA Category III then either
an EpiOcualr of CM assay must be performed.

Histological changes will be reported for each treatment group of three corneas.
Injury to each tissue layer will be scored and representative photographs taken to
illustrate the degree of damage.

10.0 REPORT

A report of this study will be prepared by the Testing Laboratory and will
accurately describe all methods used for generation and analysis of the data. A
summary will be presented for each treatment group. The report will also include
a discussion of results. A copy of the protocol used for the study and any
significant deviation(s) from the protocol will appear as a part of the final report.

11.0 RECORDS AND ARCHIVES

A separate working notebook will be used to record the materials and procedures used to
perform this study. Upon completion of the final report, all raw data, reports and
specimens will be retained in the archives for a period of either a) 5 years, b) the length
of time specified in the contract terms and conditions, or c) as long as the quality of the
preparation allows evaluation, whichever is applicable.

23


-------
All data and materials generated by PAI will be shipped or delivered to the study director
at the Testing Facility upon finalization of the pathology report, or within three months of
the issuance of the draft pathology report, whichever occurs first.

12.0 REFERENCES

Curren, R., Evans, M., Raabe, H., Dobson, T., and Harbell, J.(1999) Optimization of the
bovine corneal opacity and permeability assay: histopathology aids understanding of the
EC/HO false negative materials. ATLA 27:344.

Gautheron, P.D., Dukic, M., Alix, D., and Sina, J.F. (1992) Bovine Corneal Opacity and
Permeability Test: An in Vitro Assay of Ocular Irritancy. Fundamental and Applied
Toxicology 18:442-449.

Harbell, J.W., Raabe, H.A., Evans, M.G., and Curren, R.D. (1999) Histopathology
associated with opacity and permeability changes in bovine corneas in vitro. The
Toxicologist 48:336-337.

Sina, J.F., Galer, D.M., Sussman, R.G., Gautheron, P.D., Sargent, E.V., Leong, B., Shah,
P. V., Curren, R.D., and Miller, K. (1995) A collaborative evaluation of seven alternatives
to the Draize eye irritation test using pharmaceutical intermediates. Fundamental and
Applied Toxicology 26:20-31.

13.0 APPROVAL

SPONSOR REPRESENTATIVE	DATE

(Print or Type Name)

STUDY

24


-------
ANNEX II: OCULAR IRRITATION ASSAY FOR CERTAIN ANTIMICROBIAL
PRODUCTS MAKING CLEANING CLAIMS USING THE EpiOcular™ HUMAN CELL
CONSTRUCT (Courtesy of Institute for In Vitro Sciences, Inc.)

1.0 PURPOSE

The purpose of this study is to evaluate the potential ocular irritation of the test article by
measuring 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) dye
conversion by the EpiOcular™ tissue construct after topical exposure to the test article.

2.0 SPONSOR

2.1	Name:

2.2	Address:

2.3	Representative:

3.0 IDENTIFICATION OF TEST AND CONTROL SUBSTANCES

3.1	Test Article(s):

3.2	Controls:	Positive:

Negative:

3.3	Determination of Strength, Purity, etc.

4.0 TESTING FACILITY AND KEY PERSONNEL

4.1	Name:

4.2	Address:

4.3	Study Director:

4.4	GLP: 40 CFR Part 160 Good Laboratory Practice Standards (GLP) apply to this

assay

5.0 TEST SCHEDULE

5.1	Proposed Experimental Initiation Date:

5.2	Proposed Experimental Completion Date:

5.3	Proposed Report Date:

6.0 TEST SYSTEM

0.3% Triton®-X-100
negative (Sterile deionized water or
other solvent as appropriate)
blank control (MTT reading only)

25


-------
The EpiOcular™ human cell construct, provided by the MatTek Corporation, will be used
in this study. The EpiOcular™ cultures offers features appropriate for a model for ocular
irritation. First, the model is composed of stratified human keratinocytes (neonatal foreskins)
in a three-dimensional structure. Secondly, test materials can be applied topically to the
model so that water insoluble materials may be tested. Prior to use, each plate (6, 12, and
24-well) will be uniquely identified with a number written in permanent marker on the
plate and its cover, the test article number, and the exposure time.

EXPERIMENTAL DESIGN AND METHODOLOGY

The experimental design of this study consists of the determination of the pH of the neat
liquid test article (and/or dosing solution as appropriate), if possible, and a single
definitive assay. The toxicity of the test article will be evaluated by the exposure time
required to reduce tissue viability to 50% of controls (ET50). Viability will be determined by
the NAD(P)H-dependent microsomal enzyme reduction of MTT (and to a lesser extent, by
the succinate dehydrogenase reduction of MTT) in control and test article-treated cultures
(Berridge, et al., 1996). Data will be presented in the form of relative survival (relative
MTT conversion) versus test article exposure time.

The standard exposure time range extends up to 90 minutes and is used for most materials to
be tested. In general, a standard exposure range of 2, 15, 45 and 90 minutes will be used,
unless the Sponsor specifies an alternative exposure time range or if the Study Director
determines that the class of test articles warrants the use of an alternative exposure time
range.

7.1 Media and Reagents

7.1.1	Assay Medium: supplied by MatTek Corporation

7.1.2	EpiOcular™ Tissue: OCL-200 supplied by MatTek Corporation

7.1.3	Dulbecco's Modified Eagle's Medium (DMEM) containing 2mM
L-glutamine by Quality Biological (or equivalent) (MTT Addition Medium)

7.1.4	Sterile deionized water by Quality Biological (or equivalent)

7.1.5	3-[4,5 - dimethylthiazol-2-yl] - 2,5 - diphenyltetrazolium bromide (MTT)
Solution: 1 mg/mL MTT in MTT Addition Medium

7.1.6	Ca++ and Mg++ Free Dulbecco's Phosphate Buffered Saline
(Ca++Mg++Free-DPB S)

7.1.7	Extraction Medium: Isopropanol

26


-------
7.2	Preparation and Delivery of Test Article.

Test articles will be tested neat. End use concentrations or other forms may be used
as directed by the Sponsor. One hundred |iL of pipettable substances, such as
liquids, gels, creams, and foams, will be applied directly on the tissue so as to cover
the upper surface. To aid in filling the pipet for pipettable materials that are viscous,
the test article may first be transferred to a syringe. The pipet tip of the positive
displacement pipet will be inserted into the dispensing tip of the syringe so that the
material can be loaded into the displacement tip under pressure. Simultaneously, the
syringe plunger is depressed as the pipet piston is drawn upwards. If air bubbles
appear in the pipet tip, the test article should be removed (expelled) and the process
repeated until the tip is filled without air bubbles. This method should be used for
any materials that cannot be easily drawn into the pipet such as gels, and solid test
articles that are creamed. A dosing device (a flat headed cylinder of slightly less
diameter than the inner diameter of the tissue insert) may be placed over the test
article to assure even spreading, if required. Dry powders will be ground with a
mortar and pestle and passed through a #40 copper sieve, if needed. Powders will be
placed directly onto the culture at approximately 30 mg/culture. Materials that are
too viscous to spread over the tissue will first be spread onto the flat end of a dosing
device. The dosing device will then be placed into the Millicell® to bring the test
article in contact with the tissue. When the test article must first be applied to a
dosing device, approximately 30 |iL or 30 mg of material will be applied to the
dosing device so as to cover the dosing surface. The sample should be spread to
form a relatively smooth even layer on the surface of the dosing device to maximize
uniform tissue. All exposure conditions will be documented in the study workbook.

The stability of the test article under the actual experimental conditions will not be
determined by the testing facility.

7.3	Route of Administration

The test article(s) will be administered by topical application to the construct.

7.4	pH Determination

The pH of the neat liquid test article (and/or dosing solution as appropriate) will
be determined, if possible. The pH will be determined using pH paper (for
example, with a pH range of 0 - 14 to estimate, and/or a pH range of 5 - 10 to
determine a more precise value). The typical increments on the pH paper used to
report the pH are approximately 0.3 to 0.5 pH units. The maximum increment on
the pH paper is 1.0 pH units.

7.5	Controls

Generally, at least two negative control exposure times will be used. One
negative control exposure time will be selected to fit the range of the shortest test article
or positive control exposure times (the minimum negative control exposure time will be
15 minutes). The second negative control exposure time will be selected to match the
longest test article or positive control exposure time (whichever is longer, up to 90
minutes). On occasion, the second negative control exposure time may be selected to fit
the longest test article exposure time of a test article run concurrently, but from an

27


-------
independent study. If all exposure times are one hour and less, a single negative control
exposure time may be used. Additional negative control exposure times may be selected
at the discretion of the Study Director. Positive control cultures are treated with 0.3%
(3 mg/mL) Triton®-X-100 prepared in sterile deionized water and are exposed for 15 and
45 minutes. At least two cultures will be used for each negative and positive control
exposure time.

7.6	Assessment of Direct Test Article Reduction of MTT

It is necessary to assess the ability of each test article to directly reduce MTT. A
1.0 mg/mL MTT solution will be prepared in warm MTT Addition Medium as described
in §7.8. Approximately 100 |iL (liquid test articles) or 30 mg (solid test articles) will be
added to 1 mL of the MTT solution and the mixture incubated in the dark at 37±1°C in a
humidified atmosphere of 5+1% CO2 in air (standard culture conditions) for
approximately one hour. The negative control (100 |iL) will be run concurrently. If the
MTT solution color turns blue/purple, the test article is presumed to have reduced the
MTT. Water insoluble test materials may show direct reduction (darkening) only at the
interface between the test article and the medium.

7.7	Receipt of the EpiOcular™ model

Upon receipt of the EpiOcular™ assay materials, the solutions will be stored as
indicated by the manufacturer. The tissue will be stored at 2-8°C until used.

On the day of dosing, EpiOcular™ Assay Medium will be warmed to approximately
37°C. Nine tenths (0.9) mL of Assay Medium will be aliquoted into the appropriate
wells of prelabeled 6-well plates. The 6-well plates will be labeled with the test
article(s) and exposure time(s). Each tissue will be inspected for air bubbles
between the agarose gel and Millicell® insert prior to opening the sealed package.
Cultures with air bubbles under greater than 50% of the Millicell® area will not be
used. Each 24-well shipping container will be removed from its plastic bag and its
surface disinfected by wiping with 70% ethanol-soaked tissue paper. An appropriate
number of tissues will be transferred aseptically from the 24-well shipping
containers into the 6-well plates. The EpiOcular™ tissues will be incubated at
standard culture conditions for at least one hour. The medium will be aspirated and
0.9 mL of fresh Assay Medium will be aliquoted into each assay well below the
tissue. Upon opening the bag, any unused tissues remaining on the shipping agar
at the time of tissue transfer will be briefly gassed with an atmosphere of 5%
C02/95%) air, and the bag will be sealed and stored at 2-8°C for subsequent use.

7.8	Definitive MTT Assay

Four to five exposure times will be tested for each test article. The exposure times
will generally be 2, 15, 45 and 90 minutes, although other exposure times may be
suggested by the Sponsor, or selected by the Study Director. In the short term exposure
assay, if the expected range of toxic response is unknown, a 20 minute exposure time
may be performed first to determine the remaining exposure durations.

Each test article and control exposure time will be tested by treating two tissues. The
dosing procedure will be determined as indicated in §7.2. Generally, exposure times
of ten minutes or greater will be incubated at standard culture conditions.

28


-------
The positive control will be exposed for 15 and 45 minutes. A second negative
control will be exposed for the longest exposure time used for the test or control
articles up to 240 minutes.

At the end of the treatment time, the test article will be removed by extensively
rinsing both sides of the culture with room temperature Ca++ and Mg++-Free
Dulbecco's Phosphate Buffered Saline (Ca++Mg++Free-DPBS). The process will be
performed until the culture appears free from test article. If it is not possible to
remove all of the visible test material, this will be noted in the workbook.

After rinsing, the tissue will be transferred to 5 mL of Assay Medium for a 10 to 20
minute incubation at room temperature. This rinse is intended to remove any test
article absorbed into the tissue.

A 10X stock of MTT prepared in PBS (filtered at time of batch preparation) will
be thawed and diluted in warm MTT Addition Medium to produce the 1.0 mg/mL
solution no more than two hours before use. Alternatively, a 1.0 mg/mL MTT
solution will be prepared in warm MTT Addition Medium and filtered through a
0.45 |im filter to remove undissolved crystals. Three hundred |iL of the MTT
solution will be added to each designated well of a prelabeled 24-well plate. The
tissue will be transferred to the appropriate wells after rinsing, and the plates
incubated for 3 + 0.1 hours at standard culture conditions.

After 3 + 0.1 hours, the bottom of the EpiOcular™ tissue constructs will be blotted
on absorbent paper, cleared of excess liquid, and transferred to a prelabeled 24-well
plate containing 2.0 mL of isopropanol in each designated well. The plates will be
sealed with parafilm and stored in the refrigerator (2-8°C) until the last exposure
time is harvested. The plates, then, will be shaken for at least 2 hours at room
temperature. At the end of the extraction period, the liquid within each Millicell®
insert will be decanted into the well from which it was taken. The extract solution
will be mixed and 200 |iL transferred to the appropriate wells of a prelabeled
96-well plate(s). Two hundred |iL of isopropanol will be added to the wells
designated as blanks. The absorbance at 550 nm (OD550) of each well will be
measured with a Molecular Devices Vmax plate reader (or equivalent).

Freeze Killed Controls for Assessment of Residual Test Article Reduction of MTT

In cases where the test article is shown to reduce MTT, only test articles that
remain bound to the tissue after rinsing, resulting in a false MTT reduction signal,
present a problem. To demonstrate that residual test article is not acting to directly
reduce the MTT, a functional check is performed in the definitive assay to show
that the test material is not binding to the tissue and leading to a false MTT
reduction signal.

To determine whether residual test article is acting to directly reduce the MTT, a
freeze-killed control tissue is used. Freeze killed tissue is prepared by placing
untreated EpiOcular™ constructs in the -20°C freezer at least overnight, thawing
to room temperature, and then refreezing. Once refrozen, the tissue may be stored
indefinitely in the freezer. To test for residual test article reduction, killed tissues
are treated with the test article in the normal fashion. Generally, each test article

29


-------
will be evaluated for at least the shortest and longest exposure times (or longest
exposure time if all exposures are 1 hour or less) in single replicate killed tissues.
All assay procedures will be performed as for the viable tissue. A killed control
treated with sterile deionized water (negative killed control) will be tested in
parallel since a small amount of MTT reduction is expected from the residual
NADH and associated enzymes within the killed tissue.

If little or no MTT reduction is observed in the test article-treated killed control,
the MTT reduction observed in the test article-treated viable tissue may be
ascribed to the viable cells. If there is appreciable MTT reduction in the treated
killed control (relative to the amount in the treated viable tissue), additional steps
must be taken to account for the chemical reduction or the test article may be
considered untestable in this system. The OD550 values from the killed controls
will be analyzed as described in §7.10.

7.10 Presentation of Data

The raw absorbance values will be captured, and the following calculations made:

The mean OD550 of the blank control wells will be calculated. The corrected mean
OD550 of the exposure time control(s) will be determined by subtracting the mean
OD550 of the blank control from their mean OD550s. The corrected OD550 of the
individual test article exposure times and the positive control exposure times will
be determined by subtracting the mean OD550 of the blank control from their
respective OD550s. When applicable, corrected OD550 values will be calculated for
the control and test article-treated killed controls, as well. Generally, all
calculations will be performed using Microsoft Excel.

Corr. test article exposure time OD550 = Test article exp. time OD550 - Blank

mean OD550

If killed controls (KC) are used, the following additional calculations will be
performed to correct for the amount of MTT reduced directly by test article
residues. The OD550 value for the negative control killed control will be
subtracted from the OD550 values for each of the test article-treated killed controls
(at each exposure time), to determine the net OD550 values of the test article-
treated killed controls.

Net OD550 for each test article KC = Raw OD550 test article KC - Raw OD550
negative control KC

The net OD550 values represent the amount of reduced MTT due to direct
reduction by test article residues at specific exposure times. In general, if the net
OD550 value is greater than 0.150, the net amount of MTT reduction will be
subtracted from the corrected OD550 values of the viable treated tissues, at each
corresponding exposure time, to obtain a final corrected OD550 value. These final
corrected OD550 values will be used to determine the % of Control viabilities at
each exposure time.

30


-------
Final Corrected OD550 = Corrected test article OD550 (viable) - Net OD550test
article (KC)

Finally, the following % of Control calculations will be made:

corrected OD550 of each Test Article or Positive Control exposure time

% of Control = 	 X 100

corrected mean OD550 of Negative Control

The individual % of Control values are then averaged to calculate the mean % of
Control per exposure time. Viability calculations for test articles treated in the
long exposure time assay may be performed by comparing the corrected OD550s
of each test article exposure time to the appropriate exposure time control(s).

Exposure time response curves may be plotted with the % of control on the ordinate
and the test article exposure time on the abscissa. Other plot forms may be used as
requested by the Sponsor. The ET50 will be interpolated from each plot. To
determine the ET50, two adjacent points will be selected, one that shows greater
than 50% survival and one that shows less than 50% survival. The two selected
points will be used to determine the slope and the y-intercept for the equation
y = m(x) + b. Finally, to determine the ET50, the equation will be solved for
y = 50. If all of the exposure time points show greater than 50% survival, the
ET50 will be listed as greater than the longest exposure time. If all of the exposure
times show less than 50% survival, the ET50 will be presented as less than the
shortest exposure time. At the Study Director's option, additional assays may be
performed to produce the final ET50 value.

8.0 CRITERIA FOR DETERMINATION OF A VALID TEST

The assay will be accepted if the positive control, 0.3% Triton®-X-100, causes an ET50
within two standard deviations of the historical mean. The historical mean is updated every
three months. The corrected mean OD550 value for the minimum negative control exposure
time must be within 20% of the corrected mean OD550 value for the maximum negative
control exposure time (up to 240 minutes).

9.0 EVALUATION OF TEST RESULTS

If the antimicrobial product making a cleaning claim has an ET50 score of <4 minutes, it is
classified as an EPA Category I. However, a BCOP must be done to confirm this result. If
the antimicrobial product making a cleaning claim has an ET50 score of >4 minutes, but <70
minutes, it is classified as an EPA Category III. If the antimicrobial product making a
cleaning claim has an ET50 score of >70 minutes, it is classified as an EPA Category IV.

31


-------
10.0 REPORT

A report of the results of this study will be prepared by the Testing Laboratory and will
accurately describe all methods used for generation and analysis of the data. A summary
will be prepared reporting the ET50 values for each test article as well as the positive control
data. A copy of the protocol used for the study and any significant deviation(s) from the
protocol will appear as a part of the final report.

11.0 RECORDS AND ARCHIVES

A separate working notebook will be used to record the materials and procedures used to
perform this study. Upon completion of the final report, all raw data, reports and
specimens will be retained in the archives for a period of either a) 5 years, b) the length of
time specified in the contract terms and conditions, or c) as long as the quality of the
preparation allows evaluation, whichever is applicable.

12.0 REFERENCES

MTT Effective Time 50 (ET-50) Protocol, MatTek Corporation

Berridge, M.V., Tan, A.S., McCoy, K.D., Wang, R. (1996) The Biochemical and Cellular
Basis of Cell Proliferation Assays That Use Tetrazolium Salts. Biochemica 4:14-19.

13.0 APPROVAL

SPONSOR REPRESENTATIVE	DATE

(Print or Type Name)

STUDY DIRECTOR	DATE

32


-------
ANNEX III: OCULAR IRRITATION ASSAY FOR CERTAIN ANTIMICROBIAL
PRODUCTS MAKING CLEANING CLAIMS USING THE CYTOSENSOR
MICROPHYSIOMETER BIOASSAY (Courtesy of Institute for In Vitro Sciences, Inc.)

1.0 PURPOSE

The purpose of this study is to evaluate the potential ocular toxicity of a test article by
measuring the test material-induced reduction in the metabolic rate of treated L929 cells.
Changes in metabolic rate are measured indirectly as a function of changes in the
extracellular acidification rate. The dose which induces a 50% decrease in metabolic rate
(the MRD50 value [in units of mg/mL]) is the endpoint of the assay.

2.0 SPONSOR

2.1	Name:

2.2	Address:

2.3	Representative:

3.0 IDENTIFICATION OF TEST AND CONTROL SUBSTANCES

3.1	Test Article(s):

3.2	Controls:	Positive: sodium lauryl sulfate (SLS)

Solvent: solvent (when other than Low-
Buffered DMEM is used)

3.3	Determination of Strength, Purity, etc.

4.0 TESTING FACILITY AND KEY PERSONNEL

4.1	Name:

4.2	Address:

4.3	Study Director:

4.4	GLP: 40 CFR Part 160 Good Laboratory Practice Standards (GLP) apply to this

assay

5.0 TEST SCHEDULE

5.1	Proposed Experimental Initiation Date:

5.2	Proposed Experimental Completion Date:

5.3	Proposed Report Date:

33


-------
6.0 TEST SYSTEM

L929 cells obtained from ATCC, Manassass, VA, will be used in the study. An isolated
population of L929 cells will be exposed to increasingly concentrated doses of a test
article starting at the lowest concentration. The concentration of test article that causes a
50% decrease in the acidification rate (MRD50) will be determined.

7.0 EQUIPMENT : CYTOSENSOR MICROPHYSIOMETER

The Cytosensor Microphysiometer manufactured by Molecular Devices Corporation,
Menlo Park, CA., measures the extracellular acidification rate of cell cultures. The
Cytosensor Microphysiometer consists of a variety of components which may include: 1)
two Cytosensor Microphysiometer units which include eight built-in peristaltic pumps for
each channel; 2) a computer which runs the Cytosensor Microphysiometer and collects
the data; 3) a printer; and 4) sensor chambers. Various adherent cell types can be seeded
in the capsule cup. Each cell culture-containing cell capsule (capsule cup and spacer
assembly) is loaded into the sensor chamber. The capsule insert will not be included in
the assembly. The bottom of the sensor chamber is made of the silicon sensor chip. This
chip is capable of detecting very small changes in pH. Low-buffered medium is perused
across the cells in a stop/flow manner. When the flow is stopped, the change in pH due
to acidic metabolites (e.g., lactate and CO2) build-up is detected by the silicon sensor.
The acidification of the medium occurs at a reproducible rate in the presence of a normal,
undamaged cell population. Cells which have received a toxic insult will produce an
altered acidification rate.

8.0 EXPERIMENTAL DESIGN AND METHODOLOGY

The experimental design of this study consists of a solubility or miscibility test to confirm
the solubility/workability of the test article in Low-Buffered DMEM (unless otherwise
specified by the Sponsor or the Study Director), the determination of the pH of the neat
test article if possible, the determination of the pH at the highest concentration of test
article in the medium if possible, a dose range finding assay and at least two definitive
assay trials. At the Study Director's discretion, additional definitive assay trials may be
performed. Activity in the Cytosensor Microphysiometer assay is evaluated on the basis
of reduction of the acidification rate of the individual cell population after the exposure to
and subsequent washout of a series of test article concentrations. The concentration of
test article which causes a 50% reduction in the acidification rate is calculated and termed
the MRD50 (Metabolic Rate Decrement 50%). The MRD50 will be expressed in mg/mL.

The methods for conducting the Cytosensor Microphysiometer assay are modifications of
procedures described in the Operator's Manual supplied by Molecular Devices
Corporation. Additional background information is given by Parce et al. (1989).

8.1 Media and Reagents

8.1.1	Growth Medium: Dulbecco's Modified Eagle's Medium with 1.0 mM
sodium pyruvate (DMEM) containing 10% Fetal Bovine Serum and
2.0 mM L-glutamine (Complete DMEM).

8.1.2	Seeding Medium: DMEM containing 1% Fetal Bovine Serum, 50 |ag/mL
gentamicin, 2.0 mM L-glutamine (Diluted DMEM).

34


-------
8.1.3	Low-Buffered Medium: Serum-free, Sodium Bicarbonate-free, DMEM
containing 50 |ag/m L gentamicin, 2.0 mM L-glutamine, and additional
NaCl for consistent osmolality (Low-Buffered DMEM).

8.1.4	Ca++Mg++-Free Phosphate Buffered Saline (PBS)

8.1.5	0.05% Trypsin in Ca++Mg++Free- Hanks' Balanced Salts Solution

8.1.6	Positive control - SLS 10% in water (stock)

8.2	Preparation and Delivery of Test Article

The test article will be dissolved in Low-Buffered DMEM. Other solvent systems
will be used only after consultation with the Sponsor but should generally be
avoided. If extraction of the test article is required, the extraction procedure will
be determined in consultation with the Sponsor. It is essential that the test
material be in a single phase solution/suspension in the highest dose used to
prepare the subsequent dilutions (see section 8.7).

8.3	Route of Administration

The test article dosing solutions will be administered directly to the cells. Cells
will be exposed to each concentration of test article for approximately 810 sec,
after which time the test article is rinsed out of the sensor chamber with fresh
medium. The acidification rate is immediately measured after washout of the
sample. Dosing is generally conducted by testing lower concentrations first and
gradually increasing the dose (the same cell chamber is used for each dose) until
the MRD50 point has been surpassed or until the highest concentration has been
dosed.

8.4	pH Determination

The pH of the neat liquid test article (and/or dosing solution as appropriate) will
be determined, if possible. The pH will be determined using pH paper (for
example, with a pH range of 0 - 14 to estimate, and/or a pH range of 5 - 10 to
determine a more precise value). The typical pH increments on the pH paper used
to report the pH are approximately 0.3 to 0.5 pH units. The maximum increment
on the pH paper is 1.0 pH units.

8.5	Controls

The baseline acidification rate will serve as the internal control for each cell
culture. For each sensor chamber used, baseline rates will fall between 50 and
200 microvolts/sec after a stabilization period of approximately 1 hour. The cell
capsule in any chamber which fails to achieve these ranges will be replaced, or
the channel will not be used in the assay, unless the Study Director determines the
chamber to be acceptable.

Each assay will include a concurrent solvent control (when a solvent other than
Low-Buffered DMEM is used) and a positive control. The positive control will

35


-------
be tested like a test article except that the dose range will be set based on
historical data.

At the beginning of each assay, at least four to five stable rates are taken as the
baseline rate. For each sensor chamber, these baseline data points should vary
from their mean by no more than 10%, and will be determined just prior to
introduction of the first sample dilutions. If the baseline data contain one out of
five outlying points that can be explained (e.g., caused by a bubble), it is
permissible to delete that data point and use only four for calculations.

8.6	Cell Maintenance and Preparation of the Capsule Cups

Stock cultures of L929 cells will be maintained and passaged in Growth Medium
and incubated at 37 ± 1°C and 5 ± 1% C02 in air. L929 cells will be seeded onto
capsule cups at approximately 6.0 x 105 cells per capsule cup in Seeding Medium
as described below.

Flasks of L929 cells to be passaged or seeded are selected at or near confluency.
The size of flasks used will depend on the number of cells needed. The Growth
Medium is decanted and the cell sheet washed twice with approximately 10 mL of
PBS for each 75cm2 of growth surface. The cells are trypsinized with
approximately 3 mL of trypsin (for each 75cm2 of growth surface) for 15 to 30
seconds. The trypsin solution is aspirated and the cells are incubated at room
temperature for approximately 2 to 5 minutes, until the cells begin to round. The
cells are dislodged by tapping the flask and approximately 5mL of Seeding
Medium are for each 75cm2 of growth surface. The cells are triturated using a
pipet in order to break up clumps and are transferred by pipet to a conical
centrifuge tube. If more than one flask is used, the contents of each are pooled.

Cell counts are performed as required. The L929 cells will be seeded with
approximately 6.0 X 105 cells per each capsule cup (0.5 mL of a 1.2 X 106 cell
suspension) with 1.5 mL of Seeding Medium added to each outside well. The
plate will be labeled with cell type, seeding density, and date. The plate will then
be incubated at 37 ± 1°C and 5 ± 1% CO2 in air for 16 to 32 hours. Prior to the
start of the assay, the medium in capsule cups will be switched to Low-Buffered
DMEM and a spacer will be added to each capsule cup and gently tapped down to
the bottom. The cell capsules will be placed into the sensor chambers and exposed
to Low-Buffered DMEM at 37 ± 1°C.

For routine passaging, the stock cultures are trypsinized as described above, but
are dislodged and resuspended using warm (approximately 37°C) Growth
Medium, seeded into a culture flask(s), and returned to the humidified incubator
maintained at 37 ± 1°C and 5 ± 1% CO2 in air.

8.7	Dose Range Finding Assay

A dose range finding assay will be performed to establish an appropriate test
article dose range for the definitive Cytosensor Microphysiometer assay. Dosing
solutions will be prepared by serial three-fold dilutions (producing the same
concentrations suggested in the following table) in sterile, Low-Buffered DMEM
that has been allowed to equilibrate to room temperature.

36


-------
IMPORTANT: Do not attempt to use preparations that separate into more than
one phase in the Cytosensor. Similarly, do not attempt to use such preparations to
make dilutions. At the discretion of the Study Director, a suspension that
maintains a single phase may be assayed and used to prepare further dilutions.

If the sample does not go into a single phase with the medium at 10.0 mg/mL
(maintaining a ratio of 100 mg/10 mL), prepare dilutions 2 or 3 as required. If a
single phase test article/medium mixture is not achieved, the Study Director and
Sponsor are to be consulted.

DILUTION #

CONCENTRATION

1

10 mg/mL

2

3.33 mg/mL

3

1.11 mg/mL

4

0.370 mg/mL

5

0.123 mg/mL

6

0.0412 mg/mL

7

0.0137 mg/mL

The test article will be evaluated by exposure to L929 cells contained in sensor
chambers. The injection port for each sensor chamber will be labeled with the
designated test article or positive control prior to exposure. After the baseline
data points have been taken, the exposure cycle will begin with the lowest test
article concentration. From these baseline data points, the spreadsheet will
compute the mean baseline value used in the MRD50 calculation. Each exposure
cycle will take 20 minutes.

The maximum solvent concentration (other than Low-Buffered DMEM) will be
10% unless otherwise specified by the Sponsor or Study Director.

There will be three phases in the exposure cycle, with the following parameters
selected within the Cytosensor Microphysiometer software (Cytosoft): First, a test
article concentration will be introduced into the sensor chamber for 13 minutes and 30
seconds. The nominal rate of flow will be 100 |iL per minute for the first minute, and 20
|iL per minute for the next 12 minutes and 30 seconds. The second phase will be the
wash-out phase which will be six minutes at a nominal rate of 100 |iL per minute. The
test article will be washed out of the sensor chamber during this phase. Finally, the third
phase will be the measurement of the acidification rate. For 25 seconds, there will be no
flow and the rate of pH change will be measured.

The exposure cycle will repeat with increasing test article concentrations until
either the highest test article concentration is reached or until the MRD50 value
has been surpassed. Each test article concentration will be tested on a single set
of cells. Positive control materials and solvent controls (for solvents other than
Low-Buffered DMEM) will be tested in the same fashion. If possible, an MRD50
value will be calculated from the dose range finding assay.

37


-------
The test article doses for the definitive assay will be chosen so that generally
seven doses (spaced as three-fold dilutions) will be available for the determination
of the MRDsq. Generally, three concentrations will be chosen to result in
expected survivals lower than 50%, one concentration will be chosen to result in
an expected survival of approximately 50%, and three or more concentrations will
be chosen to result in expected survivals greater than 50%. If a test article fails to
cause 50% toxicity in the dose range finding Cytosensor Microphysiometer assay,
the maximum dose will generally be 270 mg/mL, or less based on its
solubility/workability.

8.8	Definitive Assay

The definitive assay will be performed in the same manner as the dose range
finding assay, with the exception that if the MRD50 value from the dose range
finding assay is > 10 mg/mL, higher doses of test article will be prepared and
tested in the definitive assay. At least seven doses, spaced at three-fold dilution
intervals, up to a maximum of 270 mg/mL will be prepared. The determination of
the final MRD50 will be based upon the results of at least two definitive assays
and will generally also include the results of the dose range finding assay, if an
MRD50 could be determined. At the Study Director's option, the results from
additional definitive assays may also be incorporated into the calculation of the
final MRD50.

8.9	Data Analysis

The acidification rates which occur after exposure to each test article
concentration are calculated by the Cytosoft program and compared to the mean
acidification rate (basal acidification rate) of the same cells prior to exposure to a
test material to determine the percent of control acidification rate for each dose.
The dose response curve will be plotted with the percent of control acidification
rates on the ordinate and the test article concentrations on the abscissa. The
concentration of test material which results in a fifty percent reduction in
acidification rate is interpolated from the curve and referred to as the MRD50.
MRD50 data will be expressed in mg/mL.

9.0 CRITERIA FOR DETERMINATION OF A VALID TEST

The Cytosensor Microphysiometer assay will be accepted if the positive control MRD50
falls within two standard deviations of the historical mean. The historical mean will be
updated every three months.

10.0 EVALUATION OF TEST RESULTS

If the antimicrobial product making a cleaning claim has an MRD50 score of <2 mg/mL, it is
classified as an EPA Category I. However, a BCOP must be conducted to confirm this
result. If the antimicrobial product making a cleaning claim has an MRD50 score of >2
mg/mL, but <80 mg/mL, it is classified as an EPA Category III. If the antimicrobial product
making a cleaning claim has an MRD50 score of >80 mg/mL, it is classified as an EPA
Category IV.

38


-------
39


-------
11.0 REPORT

A report of the results of this study will be prepared by the Testing Facility and will
accurately describe all methods used for the generation and analysis of the data. For each
test article, the individual MRD50 values from each assay trial, and the average MRD50
value from at least two valid definitive trials will be presented. The MRD50 value from
the dose range finding assay will be included in the calculation of the average MRD50, if
one can be determined. A separate summary will be prepared reporting the MRD50
values for each assay with each test article as well as the positive control data. A copy of
the protocol used for the study and any significant deviation(s) from the protocol and the
SOPs of the Testing Facility will appear as a part of the final report.

12.0 RECORDS AND ARCHIVES

A separate working notebook will be used to record the materials and procedures used to
perform this study. Upon completion of the final report, all raw data, reports and
specimens will be retained in the archives for a period of either a) 5 years, b) the length
of time specified in the contract terms and conditions, or c) as long as the quality of the
preparation allows evaluation, whichever is applicable.

13.0 REFERENCES

Botham, P.A., Osborne, R., Atkinson, K., Carr, G., Cottin, M., and Van Buskirk, R.G.
(1997) IRAG working group 3: cell function-based assays. In: Eye Irritation Testing:
Practical Applications of Non-Whole Animal Alternatives. Food And Chemical
Toxicology 37:67-77.

Bruner, L.H., D.J. Kain, D.A. Roberts and Parker, R.D. (1991). Evaluation of seven in
vitro alternatives for ocular safety testing. Fundamental and Applied Toxicology 17:
136-149.

Harbell, J.W., Osborne, R., Carr, G.J., and Peterson, A. (1997) Assessment of the
Cytosensor microphysiometer assay in the COLIPA in vitro eye irritation validation
study. Submitted, Toxicology In Vitro.

McConnel, H.M., Owicki, J.C., Parce, J.W., Miller, D.L., Baxter, G.T., Wada, H.G., and
Pitchford, S. (1992) The Cytosensor microphysiometer: biological applications of silicon
technology. Science 257:1906-1912.

Parce, J.W., Owicki, J.C., Kercso, K M., Sigal, G.B., Wada, H.G., Muir, V.C., Bousse,
L.J., Ross, K.L., Sikic, B.I, McConnell, H.M. (1989) Detection of cell-affecting agents
with a silicon biosensor. Science 246: 243-247.

40


-------
14.0 APPROVAL

SPONSOR REPRESENTATIVE

IIVS STUDY DIRECTOR

DATE

DATE

41


-------