Cerium dioxide

Endpoint summary

Description of key information

Skin irritation / corrosion: not irritating

Eye irritation: not irritating

Key value for chemical safety assessment

Skin irritation / corrosion

Link to relevant study records

Reference

Reference 1

Endpoint:

skin irritation: in vitro / ex vivo

Type of information:

experimental study

Adequacy of study:

key study

Study period:

Not applicable

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

; however, the physico-chemical characterisation of the tested nano-CeO2 was of low quality but at the publication date the scientific community required less physico-chemical endpoints in such articles.

Reason / purpose for cross-reference:

reference to same study

Qualifier:

no guideline followed

Principles of method if other than guideline:

EpiDerm "EPI-200" human epidermal model skin irritation test (as developed in MatTek Corporation, Ashland, MA, USA).

GLP compliance:

not specified

Remarks:

The GLP compliance status was not specified in this article

Test system:

human skin model

Source species:

human

Cell type:

non-transformed keratinocytes

Vehicle:

not specified

Details on test system:

EpiDermTM "EPI-200" HUMAN EPIDERMAL MODEL
The EpiDerm™ EPI-200 human epidermal model from MatTek Corporation (USA) was used in this study for assessment of the safety of Envirox™ since the skin is a potential route of exposure during the formulation and use of the nano-CeO2 containing diesel fuel additive when accidental skin contamination might reasonably be expected to occur.

1% (w/v) Triton X-100 and 1% of 20% (w/v) SLS were used as positive control and standard reference material, respectively.

Control samples:

yes, concurrent positive control

Amount/concentration applied:

50 mg of test substance applied

Duration of treatment / exposure:

Exposure times of 960, 1200 and 1440 minutes

Irritation / corrosion parameter:

other: mean irritation potential (MIP)

Value:

< 0.01

Remarks on result:

other: for nano-CeO2

Remarks:

- Basis: mean calculated as ET50 (reference material) / ET50 (test material). - Reversibility: no data.

Irritation / corrosion parameter:

other: MIP

Value:

0.003

Remarks on result:

other: for bulk CeO2

Remarks:

- Reversibility: no data.

- EpiDermTM “EPI-200” HUMAN EPIDERMAL MODEL

Nano-CeO2 and bulk CeO2 did not directly reduce MTT. The ET50 values for nano-CeO2, 1% (w/v) TritonX-100 the positive control and 1% of 20% (w/v) SLS, a standard reference material, were 1517.18, 260.10 and 20.68 minutes respectively. The bulk CeO2 was assayed at a later time than nano-CeO2 and the corresponding ET50 values for bulk CeO2, the positive control and the reference material were > 1440, 410.60 and 46.22 minutes respectively.

A mean irritation potential (MIP) score was calculated as follows: MIP = ET50 (reference material) / ET50 (test material).

The MIP for both the nanometric and bulk CeO2 was calculated and determined as 0.01 for nano-CeO2 and 0.03 for bulk CeO2. Since both nanometric and bulk CeO2 had MIP values which were < 0.8 neither was considered to have the potential to be an in vivo skin irritant.

 

Test material	Exposure time (min)	Mean OD450	% viability	ET50 min	MIP
Negative control	960 1440	1.685	100
Nano-CeO2	960 1200 1440	2.010 1.138 1.089	119.29 67.54 64.53	1517.18	< 0.01
Bulk CeO2	960 1200 1440	1.994 1.975 1.729	92.87 97.36 86.23	> 1440	0.003
Positive control (1% TX100)	240 360 480	1.035 0.212 0.095	61.42 12.58 5.64	260.10	< 0.18
Reference material 20% SLS	15 30 60 120	0.983 0.710 0.273 0.100	58.34 42.14 16.20 5.93	20.68	1.00

Interpretation of results:

GHS criteria not met

Conclusions:

Since nano-CeO2 and bulk CeO2 had a mean irritation potential (MIP) value lower than 0.8, both materials were not considered to have the potential to be an in vivo skin irritant in this in vitro EpiDerm test.

Executive summary:

Park B et al. (2007, 2008) investigated the skin irritating potential of nanometric cerium dioxide (nano-CeO2) and micrometric CeO2 (bulk CeO2) in vitro.

A commercial nano-CeO2 from Energenics Europe Ltd. (EnviroxTM) was used in this study. The nanoparticles displayed a mean primary size of 9 nm and a specific surface area of 94.7 m²/g. A bulk CeO2 was also added to the study. This micrometric material provided by Aldrich had a mean primary size of 320 nm and a specific surface area of 2.64 m²/g. The surface chemistry of both crystalline materials was evaluated and demonstrated that the surface of nanometric and bulk CeO2 was composed of Ce and O.

The EpiDerm™ EPI-200 human epidermal model from MatTek Corporation (USA) was used in this study for assessment of skin irritation. Fifty milligrams of test substance was applied to the human epidermal model for 960, 1200 and 1440 minutes. The irritation assessment was based on the measurement of cellular viability using the MTT assay. A mean irritation potential (MIP) score was calculated as follows: MIP = ET50 (reference material) / ET50 (test material) where ET50 was the exposure time at which relative value of optical density was 50% of the negative control and the reference material was SLS (1% of 20% (w/v)).  

Nano-CeO2 exposure induced variations of cellular viability: 119.3% at 960 minutes, 67.5% at 1200 minutes, 64.5% at 1440 minutes. Exposure to bulk CeO2 caused no or only few changes in cell viability: 92.87% at 960 minutes, 97.36% at 1200 minutes, 86.23% at 1440 minutes. As nanoparticles and bulk material had a MIP value lower than 0.8 (i.e., < 0.01 for nano-CeO2 and 0.003 for bulk CeO2), both nano- and micro-CeO2 were not considered to have the potential to be an in vivo skin irritant based on this in vitro EpiDerm test.

Reference 2

Endpoint:

skin irritation: in vitro / ex vivo

Type of information:

experimental study

Adequacy of study:

supporting study

Study period:

Not applicable

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

The tested nano-CeO2 were poorly characterized.

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 439 (In Vitro Skin Irritation: Reconstructed Human Epidermis Test Method)

Principles of method if other than guideline:

EpiDerm (EPI-200) human epidermal model skin irritation test (as developed in MatTek Corporation, Ashland, MA, USA).

GLP compliance:

not specified

Remarks:

THe GLP status was not specified in the article.

Test system:

human skin model

Source species:

human

Cell type:

non-transformed keratinocytes

Justification for test system used:

Standardised human skin model validated by ECVAM and OECD TG 439

Vehicle:

other: culture medium containing 10% foetal bovine serum

Details on test system:

RECONSTRUCTED HUMAN EPIDERMIS (RHE) TISSUE
EpiDerm™ EPI-200 tissue inserts were received via overnight shipment from the supplier in 24-well plates at 4°C, on medium-supplemented agarose gel. Upon receipt, tissues were visually inspected to check for morphological defects. The tissue inserts were removed from the agarose gel and subsequently transferred into 6-well plates containing fresh 0.9 mL culture medium. The tissues were then preconditioned by incubation for 1 h at 37°C, 5% CO2 and 95% relative humidity (RH). The medium was then exchanged and the incubation was continued overnight for an additional 18 h to release debris and compounds associated with transport stress.

TEST CONDITIONS USED FOR TEST SYSTEM
Following overnight incubation, the tissues were topically exposed to the nanoparticles for 1 h. Each tissue insert was dosed with 1 mg/ml of nanoparticle test solution; 30 μl of the 1 mg/ml solution was applied directly to the apical surface of the EpiDermTM tissue at 1 min intervals. Three to four tissue inserts were used for each test substance, as well as for the negative and positive controls. Both controls were run in parallel and treated identically to dosed tissues.
After the last tissue was dosed, tissues were placed in the incubator (37°C, 5% CO2 and 95% RH) for 35 (± 1) min. At the end of 35-min incubation, tissues were taken out of the incubator and placed in a laminar flow hood until a 1-h exposure time had elapsed.

REMOVAL OF TEST MATERIAL AND CONTROLS
- Number of washing steps: Each tissue insert was thoroughly rinsed (~15 times) at 1-min intervals, in the same order as they were dosed, with DPBS to remove test materials from the tissue surface. After the last rinse, tissues were entirely submerged in 150 mL DPBS to remove any remaining test materials. Tissues were then gently blotted and dried with sterile cotton swabs to remove excess liquid from the surface. Tissue inserts were transferred into new 6-well plates containing fresh culture media and placed in an incubator (37°C, 5% CO2 and 95% RH) for 24 h to allow the development of cell damage. Tissue inserts were then transferred into new 6-well plates containing fresh culture media. Tissues were further incubated for another 18 h at 37 °C, 5% CO2 and 95% RH.

MTT DYE USED TO MEASURE TISSUE VIABILITY AFTER TREATMENT / EXPOSURE
- MTT concentration: 1 mg/mL
- Incubation time: 3 h
- Spectrophotometer: No data available
- Wavelength: 570 nm
- Filter, filter bandwidth, linear OD range of spectrophotometer: No data available
Approximately 42 h post-exposure, tissue inserts were removed from the 6-well plates, blotted on sterile absorbent pads, and transfered into 24-well plates pre-filled with 0.3 ml of MTT assay solution. Tissues were incubated for 3 h at 37 °C, 5% CO2 and 95% RH. Following this 3 h incubation, tissues were removed from the MTT solution, rinsed and transferred to a new 24-well plate. Tissues were then completely submerged on both sides by adding 2 ml of isopropanol to extract the formazan salt. For each tissue, two 200 μl aliquots of extraction solution were transferred into a 96-well microtiter plate for absorbance measurement and the plates were read at a wavelength of 570nm and optical density (OD) was measured using a spectrophotometer.

NUMBER OF REPLICATE TISSUES:
Three tissue inserts were used for each test substance, as well as for the negative and positive controls.

CONTROL TISSUES USED IN CASE OF MTT DIRECT INTERFERENCE
No data available

NUMBER OF INDEPENDENT TEST SEQUENCES / EXPERIMENTS TO DERIVE FINAL PREDICTION:
Only one experiment was performed.

PREDICTION MODEL / DECISION CRITERIA (choose relevant statement)
- The test substance is considered to be irritating to skin if the viability after 1-h exposure is less than 50%.
- Justification for the selection of the cut-off point(s) if different than recommended in TG 431 and 439: No data available

Control samples:

yes, concurrent negative control

yes, concurrent positive control

Amount/concentration applied:

TEST MATERIAL
- Amount(s) applied: 30 µg
- Concentration: 1 mg/mL
- Preparation: CeO2 nanoparticles received in dry powder form were weighed on an analytical balance, suspended in cell culture medium (Dulbecco’s Modified Eagle’s Medium (DMEM) provided by MatTek) containing 10% foetal bovine serum at a concentration of 1 mg/mL and subsequently dispersed using a probe sonicator with 4.5 watts output for 3 pulses, 2 seconds/pulse. For each experiment, a fresh stock suspension was prepared immediately prior to dosing the tissues. The CeO2 solutions were applied directly to the apical surface of the EpiDermTM tissue at 1-min intervals. Nylon mesh was placed on the surface of the tissue for the homogeneous, uniform distribution of test solution.

VEHICLE
- Amount(s) applied: 30 µL
- Lot/batch no., purity: No data available

NEGATIVE CONTROL
- Amount(s) applied, concentration: No data available

POSITIVE CONTROL
- Amount(s) applied, concentration: No data available

Duration of treatment / exposure:

1 h

Duration of post-treatment incubation (if applicable):

42 h

Number of replicates:

Three tissue inserts were used for each test substance, as well as for the negative and positive controls.

Irritation / corrosion parameter:

% tissue viability

Run / experiment:

#1

Value:

ca. 91

Vehicle controls validity:

other: not included

Negative controls validity:

other: 100 %

Positive controls validity:

other: 7.8%

Remarks on result:

other: for 58-nm CeO2

Irritation / corrosion parameter:

% tissue viability

Run / experiment:

#1

Value:

ca. 102

Vehicle controls validity:

other: not included

Negative controls validity:

other: 100%

Positive controls validity:

other: 7.8%

Remarks on result:

other: for 8-nm CeO2

Irritation / corrosion parameter:

% tissue viability

Run / experiment:

#1

Value:

ca. 100

Vehicle controls validity:

other: not included

Negative controls validity:

other: 100%

Positive controls validity:

other: 7.8%

Remarks on result:

other: for 40-nm CeO2

Interpretation of results:

GHS criteria not met

Conclusions:

Since the human skin equivalent model exposed to the nano-CeO2 tested displayed a viability higher than 50%, all three different CeO2 nanomaterials were not considered to be a dermal irritant in this in vitro EpiDerm test.

Executive summary:

Miyani VA and Hughes MF (2016) examined the potential of three different types of nanometric cerium dioxide (nano-CeO2) to elicit irritant contact dermatitis in vitro in a human skin equivalent model (HSEM) derived from keratinocytes.

Commercial nano-CeO2 from Alfa Aesar (58 nm), NanoAmor (8 nm) and Umicore (40 nm) were studied. The surface charge of all three nano-CeO2 was measured: -11.05 mV for 58-nm CeO2, -11.4 mV for 8-nm CeO2 and -9.67 mV for 40-nm CeO2. No further characterisation was performed in this study.

The CeO2 particles were suspended in media containing 10% foetal bovine serum. The particles (1 mg/mL) were applied to the epidermal surface of the EpiDerm(TM) EPI-200 human epidermal model (HSEM) from MatTek Corporation (USA). Positive (5% sodium dodecyl sulphate (SDS)) and negative controls (culture medium) were included. After 1 h exposure at 37°C, the HSEM was washed with saline to remove the nanoparticles. Following a 42-h incubation (37°C), HSEM viability was assessed using the MTT assay.

A test substance is considered a dermal irritant if the HSEM viability is < 50%. The mean viability for the SDS treated HSEM was 7.8%. The viabilities of the nanoparticle treated HSEM were 91% or greater. The CeO2 nanoparticles tested were not found to be dermal irritants under the conditions used in this study. According to the authors, the stratum corneum of the HSEM may limit penetration of metal nanoparticles to induce toxicity.

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed (not irritating)

Eye irritation

Link to relevant study records

Reference

Endpoint:

eye irritation: in vitro / ex vivo

Type of information:

experimental study

Adequacy of study:

key study

Study period:

Not applicable

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

other: :

Remarks:

the study was well documented and performed according to OECD guidelines 492 and 437. However, there was no mention to GLP. The study was designed to evaluate eye irritation potential of 20 nanosized materials and 3 micro-scale materials using a 2-tier non-animal testing strategy.

Reason / purpose for cross-reference:

reference to other study

Qualifier:

according to guideline

Guideline:

other: OECD Guideline 492 (Reconstructed human cornea-like epithelium test method for identifying chemicals not requiring classification and labelling for eye irritation or serious eye damage)

Deviations:

yes

Remarks:

the amounts of solid nano-CeO2 applied differed from those recommended in the guideline (i.e., 50 mg).

Qualifier:

according to guideline

Guideline:

OECD Guideline 437 (Bovine Corneal Opacity and Permeability Test Method for Identifying Ocular Corrosives and Severe Irritants)

Deviations:

no

Principles of method if other than guideline:

The test material was submitted to a 2-tier non-animal testing strategy composed of the EpiOcular™ Eye Irritation Test (EpiOcular™-EIT; OECD TG 492) and the Bovine Corneal Opacity and Permeability (BCOP; OECD TG 437) assay including histopathological evaluation of the cornea.

GLP compliance:

not specified

Remarks:

The GLP compliance was not specified in the published article.

Species:

other: not applicable - in vitro tests

Strain:

other: not applicable - in vitro tests

Details on test animals or tissues and environmental conditions:

EpiOcularTM-EIT HUMAN CORNEA-LIKE MODEL
The EpiOcular™-EIT model from MatTek Corp. (USA) and MatTek In Vitro Life Science Laboratories (Slovakia) use the cornea-like non-keratinized tissue construct EpiOcular™ OCL-200 kit that is composed of normal human epidermal keratinocytes obtained from individual donors. The model is cultured in proprietary serum-free culture medium (in Millicells® with 1-cm diameter) which induces corneal differentiation and formation of the organotypic cornea-like model. This three dimensional tissue consists of highly organized cell layers and exhibits barrier properties similar to the normal in vivo corneal epithelium.
On the day of arrival in the laboratory, the EpiOcular™ tissues (0.6 cm² surface area) were transferred to sterile 6-well plates with 1 mL DMEM and pre-conditioned at standard culture conditions (37°C, 5% CO2, 90–95% humidity) in the incubator for 16–24 h. After pre-incubation, the tissues were pre-treated with 20 μL PBS and further incubated at standard culture conditions for 30 min.

BOVINE CORNEA
Bovine corneas (source not given) were mounted in corneal holders. Both the anterior and posterior chambers of these holders were filled to excess with pre-warmed Eagle’s MEM (without phenol red). After equilibration in a vertical position at 32°C for at least 1 h, the medium in both chambers was replaced with fresh pre-warmed medium and the initial corneal opacity was measured.

Vehicle:

other: undiluted for EpiOcular™ assay or diluted in water for BCOP assay (see below in "Details on study design")

Controls:

other: see below in "Details on study design"

Amount / concentration applied:

- OECD 492 – EpiOcular™-EIT HUMAN CORNEA-LIKE MODEL
NM-211: 61 mg (corresponding to 50 μL bulk volume)
NM-212: 28 mg (corresponding to 50 μL bulk volume)

- OECD 437 - BCOP
NM-211 and NM-212: 20% (w/v) (corresponding to 750 μL bulk volume)

Duration of treatment / exposure:

- OECD 492 – EpiOcular™-EIT HUMAN CORNEA-LIKE MODEL
Exposure period: 90 min (variant 1), or 6 h (variant 2)
Post-exposure period: 12 min + 18 h (variant 1) or 25 min + 18 h (variant 2)

- OECD 437 - BCOP
Exposure period: 4 h

Observation period (in vivo):

Not applicable

Number of animals or in vitro replicates:

- OECD 492 – EpiOcular™-EIT HUMAN CORNEA-LIKE MODEL
2 tissues per treatment group

- OECD 437 - BCOP
3 corneas per treatment group

Details on study design:

In the 2-tier "bottom up" testing strategy suggested by the authors, the EpiOcular™-EIT was performed in Tier 1 to distinguish "Category 1 or 2" from "Non-category" substances. In Tier 2, the BCOP assay was conducted to identify "Category 1" substances from within the set of "Category 1 or 2" substances identified in Tier 1. All substances that were not identified as "Category 1" in the BCOP assay in Tier 2 were classified as "Category 2".

OECD 492 - EpiOcularTM-EIT HUMAN CORNEA-LIKE MODEL
Controls (see in Table 2): highly deionized water (50 µL) was used as negative control (NC) and methyl acetate (purity > 98 %; 50 µL) as positive control (PC).
Furthermore, 3 organic pigments that fall under the EU recommendation on the definition of nanomaterials representing three different chemical classes, i.e. Pigment Red 57:1, Pigment Yellow 95, and Pigment Black 32, were included (commercial grades from BASF SE). According to Kolle SN et al., these pigments were selected based upon availability of physico-chemical characterisation data and Good Laboratory Practice-compliant animal data. To date, no organic pigment has been classified as an eye irritant. In addition, non-nanosized talc (from a local retailer), the historical negative control for in vivo eye irritation testing, was added to the spectrum of test materials. The 3 organic pigments and talc were applied undiluted at amounts corresponding to 50 μL bulk volume: i.e., 12 mg of Pigment red 57:1, 9 mg of Pigment Yellow 95, 8 mg of Pigment Black 32, and 16 mg of talc.

The EpiOcular™-EIT was performed in two variants. Both NM-211 and NM-212 were submitted to protocol variant 1, i.e. as described by the supplier MatTek and Harbell et al. (2009). The three organic pigments and talc were assessed in accordance with variant 2, i.e. as described in the OECD TG 492. To ensure data comparability, the dry-powder NM-211 and NM-212 were additionally submitted to variant 2. The two test protocol variants differed in respect to the exposure and post-exposure immersion periods laid down for solid test materials.

The test materials were applied undiluted at amounts enabling to cover the entire tissue surfaces (i.e., 50 μL bulk volume). Using a sharp spoon or pipette, the dry-powder items were applied to cover the entire tissue surface. Two tissues were treated with either the test materials, the NC or the PC. After test material application, the tissues were placed into the incubator for the following exposure periods: 90 min (variant 1), or 6 h (variant 2). To remove the test materials, the tissues were washed with sterile PBS and immediately immersed into 12-well plates, pre-filled with 5 mL pre-warmed medium per well to remove test material residuals. After 12 min (variant 1) or 25 min (variant 2), each tissue was dried on absorbent paper and transferred to fresh 6-well plates filled with 1 mL pre-warmed medium per well (post-exposure immersion). Subsequently, the tissues were incubated at standard culture conditions (post-exposure incubation) for 18 h. During the post-exposure immersion and incubation periods, weak cytotoxic effects might reverse, and more pronounced effects might increase.

Tissue destruction was determined by formazan reduction after incubation with the tetrazolium salt 3-[4,5-dimethylthiazol- 2-yl]-2,5-diphenyltetrazolium bromide (MTT), which reflected impaired mitochondrial dehydrogenase activity. Upon completion of the post-exposure period, the assay medium was replaced by 0.3 mL MTT solution. After incubating the tissues for 3 h, the tissues were washed with PBS to terminate the MTT incubation. The produced formazan was extracted by incubating the tissues in isopropanol at room temperature overnight or on a plate shaker for at least 2 h. The optical density of the formazan extracts was determined spectrophotometrically at a wavelength of 570 nm (OD570). For each microtitre plate, blank values were established from 4 wells filled with isopropanol.

For all test materials, pre-tests as described in the OECD TG 492 were performed that precluded the test materials’ ability to directly reduce MTT.

Tissue OD570 values were calculated by subtracting the mean blank value of the respective microtitre plate from the measured tissue OD570 value, and mean OD570 values were calculated for the two tissues of each treatment group. The quotient of the mean OD570 values of the test material-treated tissues and those of the NC (i.e., the mean relative tissue viability) was determined to evaluate whether or not a test material was an irritant:
* Mean relative tissue viabilities ≤ 60 % indicated "irritancy to the eye";
* Mean relative tissue viabilities > 60 % indicated "no irritancy to the eye".

- Acceptance criteria: in case one of the following acceptance criteria (AC) as described in the OECD TG 492 was not met, repetition of the EpiOcular™-EIT was considered.
* AC for the NC: the OD570 of the NC reflected the laboratory-specific tissue viability under the specific conditions of the assay. It was considered acceptable if the mean OD570 of the NC was ≥ 0.8 and ≤ 2.5 and the historical in-house mean at the respective time of testing was met (variant 1: OD570 of NC = 1.361 ± 0.138; variant 2: OD570 of NC = 1.650 ± 0.159).
* AC for the PC: in-house, the PC methyl acetate usually elicited relative tissue viabilities of approx. 25% (historical in-house means at the time of testing in accordance with variant 1: OD570 of PC = 0.318 ± 0.119; variant 2: OD570 of PC = 0.396 ± 0.098). In addition to these historical means, all relative tissue viability values < 50 % were considered acceptable.
* AC for tissue variability: the relative inter-tissue variability (ITV%) between the two tissues of a treatment group was considered acceptable if it was ≤ 20 %.

OECD 437 - BCOP
Controls (see in Table 2): highly deionized water was used as NC. Imidazole 20 % (w/v) dissolved in highly deionized water was used as PC. Moreover, the 3 organic pigments, and talc were applied undiluted (120, 80, 48, 45, and 40 mg, for talc and the three organic pigments Pigment Red 57:1, Pigment Yellow 95, and Pigment Black 32, respectively).

The BCOP assay was conducted according to OECD TG 437.
Bovine corneas were mounted in corneal holders. Both the anterior and posterior chambers of these holders were filled to excess with pre-warmed Eagle’s MEM (without phenol red). After equilibration in a vertical position at 32 °C for at least 1 h, the medium in both chambers was replaced with fresh pre-warmed medium and the initial corneal opacity was measured.

NM-211 and NM-212 powders were suspended in highly deionized water to achieve final concentrations of 20% (w/v) suspension and then stirred with a magnetic stirrer. All test material suspensions were applied immediately after preparation, and also during test material application, the preparations were stirred with a magnetic stirrer to ensure continued homogeneity. Generally, each treatment group (NC, PC, or test material) consisted of 3 corneas. Before application of the test materials, the medium in the anterior chamber was removed:
* For the NC, the anterior chambers were filled with 750 μL highly deionized water and, for the PC, with 750 μL of the 20% (w/v) imidazole solution.
* For NM-211 and NM-212, 750 μL of the 20% (w/v) test material preparation was applied directly to the epithelial surface of the cornea (i.e., using the "open chamber method").
* Talc and the three organic pigments Pigment Red 57:1, Pigment Yellow 95, and Pigment Black 32 were applied undiluted.

The corneas were incubated in a horizontal position at 32°C for 4 h as prescribed for non-surfactant solids in the OECD TG. Upon completion of the incubation period, the NC, PC and test materials were removed from the anterior chamber with a syringe, and the respective epithelia were washed at least 3 times with Eagle’s MEM (containing phenol red) and once with Eagle’s MEM (without phenol red). Both chambers were then refilled with fresh Eagle’s MEM (without phenol red).

The final corneal opacity was measured, and the opacity change per cornea was calculated by subtracting the initial from the final opacity value. Subsequently, the mean opacity change of the NC was subtracted thereby providing the corrected opacity change. Test results were provided as means of all corrected opacity changes per treatment group.
To determine corneal permeability, the medium in the anterior chamber was replaced by 1 mL sodium fluorescein solution (5 mg/mL) and incubated in a horizontal position for 90 min at 32°C. The amount of sodium fluorescein that permeated through the corneas was measured spectrophotometrically. Three aliquots per cornea were transferred to a 96-well microtitre plate and the optical density value (OD490) was determined subtracting the mean blank OD490 (blank = Eagle’s MEM without phenol red) from the OD490 of each cornea. Corrected OD490 values were calculated by subtracting the mean OD490 values of the corresponding NC. Final test results were calculated as means of all corrected OD490 values per treatment group.
The In Vitro Irritancy Score (IVIS) was calculated per treated cornea and finally the mean IVIS per treatment group ± standard deviation (SD) was determined: IVIS = mean opacity value + (15 x mean permeability value). An IVIS > 55 indicated a risk of serious damage to the eyes.

- Acceptance criteria: in case one of the following ACs laid down in OECD TG 437 was not met, repetition of the BCOP assay was considered.
* AC for the NC: the NC responses should be lower than the established upper limits for background opacity and permeability values for the respective NC.
* AC for the PC: the IVIS calculated for the PC should not lie outside the two-fold range of the SDs of the historical mean (i.e., 88.0–147.1 for imidazole).
* AC for the treatment groups: at least 2 of the 3 corneas per treatment group should provide predictions that coincided with the mean of all 3 corneas, and none of the corneas should provide a discordant prediction of 10 IVIS units above or below the cut-off threshold of 55.

HISTOPATHOLOGICAL EVALUATION
For histopathological evaluation by light microscopy, the corneas were fixed in 10% neutral buffered formalin for at least 24 h and trimmed along the whole diameter (2 stripes of 3–4 mm width). They were histotechnically processed with a standard method for light microscopy and stained with Hematoxylin and Eosin. Histopathological findings were assessed in the epithelium based on the depth of injury by using a standard semi-quantitative grading system (from 1 to 5) that is related to the extent of affected cell layers beginning from the corneal surface (squamous cell layer) down to the basal cell layer. In the stroma, tissue swelling and keratocyte changes were evaluated. These findings were summarized in a so-called Histopathological Score of Irritation (HSI) assigned for each cornea ranging from 0 = no irritation to IV = severe irritation. HSI IV was assessed as "severe irritation"; HSI I, II and III were overall assessed as "non-severe irritation"; HSI 0 was regarded as no irritation.

Irritation parameter:

in vitro irritation score

Remarks:

EpiOcular test - Variant 1 - MN-211

Run / experiment:

Run 1

Value:

81

Negative controls validity:

valid

Positive controls validity:

valid

Irritation parameter:

in vitro irritation score

Remarks:

EpiOcular Test - Variant 1 - NM-212

Run / experiment:

Run 1

Value:

81

Negative controls validity:

valid

Positive controls validity:

valid

Irritation parameter:

in vitro irritation score

Remarks:

EpiOcular test - Variant 2 - NM-211

Run / experiment:

Run 1

Value:

129

Negative controls validity:

valid

Positive controls validity:

valid

Irritation parameter:

in vitro irritation score

Remarks:

EpiOcular Test - Variant 2 - NM-211

Run / experiment:

Run 2

Value:

109

Negative controls validity:

valid

Positive controls validity:

valid

Irritation parameter:

in vitro irritation score

Remarks:

EpiOcular test - Variant 2 - NM-212

Run / experiment:

Run 1

Value:

105

Negative controls validity:

valid

Positive controls validity:

valid

Irritation parameter:

cornea opacity score

Remarks:

BCOP - MN-211

Run / experiment:

Run 1

Value:

18

Negative controls validity:

valid

Positive controls validity:

valid

Irritation parameter:

cornea opacity score

Remarks:

BCOP - NM-212

Run / experiment:

Run 1

Value:

17.2

Negative controls validity:

valid

Positive controls validity:

valid

Other effects / acceptance of results:

ACCEPTANCE OF RESULTS:
- Acceptance criteria met for negative control: yes
- Acceptance criteria met for positive control: yes

OTHER INFORMATION :

> OECD 492 - EpiOcularTM-EIT HUMAN CORNEA-LIKE MODEL (see in Table 3)
The acceptance criteria (ACs) for the NC and the PC were always met in the EpiOcular™-EIT performed in accordance with either the protocol variants 1 and 2.

* Test results, variant 1 (90-min treatment)
After washing, test material residues were observed on the EpiOcular™ tissues treated with CeO2 NM-211 and NM-212. However, since none of the test materials was able to reduce MTT directly, it was concluded that these residues did not interfere with the MTT assay.
For both NM-211 and NM-212, mean tissue viabilities above 60% relative to the NC were calculated. Hence, none of the tested nano-CeO2 revealed eye irritation potential in the EpiOcular™-EIT under the chosen test conditions (i.e., indicating likelihood of "neither Category 1 nor 2").

* Test results, variant 2 (6-hour treatment)
After the 1st test run conducted in accordance with the EpiOcular™-EIT variant 2, mean relative tissue viabilities > 60% (indicating likelihood of "neither Category 1 nor 2") with concordant satisfactory ITV% were determined for CeO2 NM-212. Also for the three organic pigments and talc the mean relative tissue viabilities consistently exceeded 60%.
Due to failure to meet the AC for tissue variability or high optical density values recorded in single measurements of the 1st test run, two test runs each were conducted to evaluate the eye irritating potential of CeO2 NM-211. For NM-211, all individual and mean relative tissue variability values ranged above 90%. Therefore, all of these findings were assessed as indicating likelihood of "neither Category 1 nor 2".

> OECD 437 - BCOP (see in Table 4)
The AC for the NC was always met. Due to high opacity scores, one PC did not meet the AC for the PC. However, since all other ACs were met and the test material results were unambiguous, the study was assessed as being valid.

According to the authors, as determined by AUC and / or LD as relevant, NM-211 and NM-212 suspended in water (i.e., as prepared for the BCOP assay) were predominantly present as agglomerates around 1 μm diameter. Dispersed fractions of the total dose with diameters up to 1 μm that exceeded 50% of the total dose (i.e., indicating higher dispersibility) were recorded for CeO2 NM-212. Lower than 50% of the total dose, but still noteworthy fractions of small agglomerates below 100 nm were further recorded for CeO2 NM-211.

None of the tested nano-CeO2 induced serious eye damage in the BCOP as assessed by the respective IVIS that were far below the cut-off value of 55. Likewise, the three organic pigments and talc did not induce serious eye damage in the BCOP, with all IVIS being 0.

> HISTOPATHOLOGICAL EVALUATION (see in Table 5)
Minimal findings (mostly minimal multifocal or diffuse desquamation) were observed for both NM-211 and NM-212, and an HSI of I (minimal) was assigned.

Table 3: Mean tissue viability of the negative (NC), positive controls (PC), CeO2 NM-211, NM-212, three organic pigments, and talc applied in the EpiOcular™ MTT eye irritation assay

Test substance	Protocol "variant 1"	Protocol "variant 2"
	Run 1	Run 1	Run 2	Run 4
Mean OD570 of NC	1.411	1.680	-	-
Mean NC% (and ITV%)	100 (7)	100 (4)	-	-
Mean relative PC (and ITV%)	21 (11)	25 (4)	-	-
Mean relative NM-211 (and ITV%)	81 (22)a	129 (3)b	109 (8)	-
Mean relative NM-212 (and ITV%)	81 (0)	105 (16)	-	-
Mean relative talc (and ITV%)	-	-	-	98 (10)
Mean relative Pigment Red 57:1 (and ITV%)	-	-	-	98 (0)
Mean relative Pigment Yellow 95 (and ITV%)	-	-	-	91 (8)
Mean relative Pigment Black 32 (and ITV%)	-	-	-	88 (18)

The mean tissue viability of 2 tissues per test group was expressed relative to the corresponding negative control value further indicating (in brackets) the ITV%

Abbreviations: ITV%: Relative inter-tissue variability; OD: Optical density

a: Since the ITV of the two tissues was > 20 %, the corresponding acceptance criterion was not met. However, since all other acceptance criteria were met and due to the non-ambiguous result recorded for the test substance, the test was considered valid despite this deviation and therefore was not repeated.

b: Repeat of test run due to high optical density value (> 130 % relative to the NC) of one of the measurements

Footnote to Table 3:

Highly deionized water was used as NC and methyl acetate as PC.

In the tests using the protocol "variant 1", run 1 included the testing of CeO2 NM-211 and NM-212.

In the tests using the protocol "variant 2", run 1 included the tests of CeO2 NM-211 and NM-212, run 2 the 2nd testing of CeO2 NM-211, and run 4 the testing of the three organic pigments and talc.

Table 4: Mean opacity and permeability values and corresponding in vitro irritation score (IVIS) of the negative (NC) and positive controls (PC), CeO2 NM-211, NM-212, three organic pigments, and talc used in the BCOP assay

Test substance and parameter	Run 3	Run 7	Run 8
NC Mean opacity value ± SD Mean permeability value ± SD IVIS ± SD	9.1 ± 6.0 0.02 ± 0.01 9.5 ± 5.9	8.1 ± 4.9 0.01 ± 0.00 8.2 ± 4.9	7.3 ± 3.3 0.00 ± 0.00 7.3 ± 3.3
PC Mean opacity value ± SD Mean permeability value ± SD IVIS ± SD	67.8 ± 9.4 3.4 ± 0.9 118.7 ± 8.8	83.0 ± 11.1 1.3 ± 0.2 102.4 ± 10.1	70.2 ± 25.4 3.0 ± 0.5 115.1 ± 33.3
NM-211 Mean opacity value ± SD Mean permeability value ± SD IVIS ± SD	18.0 ± 9.9 0.03 ± 0.04 18.5 ± 10.5	- - -	- - -
NM-212 Mean opacity value ± SD Mean permeability value ± SD IVIS ± SD	17.2 ± 5.4 -0.02 ± 0.01 16.9 ± 5.3	- - -	- - -
Talc Mean opacity value ± SD Mean permeability value ± SD IVIS ± SD	- - -	0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0	- - -
Pigment Red 57:1 Mean opacity value ± SD Mean permeability value ± SD IVIS ± SD	- - -	0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0	- - -
Pigment Yellow 95 Mean opacity value ± SD Mean permeability value ± SD IVIS ± SD	- - -	0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0	- - -
Pigment Black 32 Mean opacity value ± SD Mean permeability value ± SD IVIS ± SD	- - -	0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0	0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0

Footnote to Table 4:

Highly deionized water was used as NC and imidazole (20% (w/v) highly deionized water) as PC.

Run 3 included the tests of CeO2 NM-211 and NM-212, run 7 the testing of the three organic pigments and talc, and run 8 the testing of Pigment Red 57:1.

Table 5: Histopathological evaluation of the bovine corneas incubated with NM-211 or NM-212

Test material	Test run	HSI	Histopathological findings
CeO2 NM-211	1	I	Multifocal vacuolation in the squamous and wing cell layers; multifocal desquamation and brown granules on the epithelial surface
CeO2 NM-212	1	I	Multifocal desquamation and brown granules on the epithelial surface

HSI histopathological score of irritation (HSI) of 0 = no findings, I = minimal, II = mild, III = moderate, IV = severe

Interpretation of results:

GHS criteria not met

Conclusions:

In conclusion, both tested nano-CeO2 did not elicit eye irritation in either the EpiOcular™-EIT or the BCOP assay pointing thus to the low eye irritation potential of either NM-211 or NM-212.

Executive summary:

Kolle SN et al. investigated the in vitro eye irritation potential of nanosized cerium dioxide (nano-CeO2). The study was conducted according to OECD guidelines 492 and 437. However, there was no mention to GLP. The study was designed to evaluate eye irritation potential of 20 nanosized materials and 3 micro-scale materials using a 2 -tier EpiOcular™ Eye Irritation Test (EpiOcular™-EIT) and BCOP testing strategy including histopathology of the bovine corneas. The study was performed, within the context of the OECD Sponsorship Program for the Testing of Manufactured Nanomaterials.

Two commercial nano-CeO2 (NM-211 and NM-212) provided by JRC (Italy) were used in this study. The nanomaterials were physico-chemically characterised by Kolle SN et al.:

 

Parameters	Results		Methods
	NM-211	NM-212
Supplier	JRC	JRC	-
Primary particle size	10.3 nm	33 nm	X-ray diffraction (XRD)
Particle size distribution	D50 of agglomerate fractions = 2.14 µm in water	D50 of agglomerate fractions = 0.776 µm in water	Laser diffraction (LD)
Stability	Agglomeration	Agglomeration	-
Specific surface area	66 m²/g	28 m²/g	Brunauer-Emmett-Teller (BET) technique
Surface charge (zeta potential at pH 7)	No data	No data	-
Isoelectric point	No data	No data	-
Shape	Cubic	Cubic	XRD
Crystallinity	No data	No data	-
Analytical purity	> 95%	> 99.5%	Supplier’s data
Impurities	No data	No data	-
Solubility	< 0.1 ppm in water and DMEM + FCS (< 0.001% dissolution) or 0.002 wt% (Wohlleben W et al., 2013; Keller J et al., 2014)	< 0.1 ppm in water and DMEM + FCS (< 0.001% dissolution) or up to 0.02% depending on the medium (Wohlleben W et al., 2013; Keller J et al., 2014)	Analytical ultracentrifugation (AUC) combined with inductively coupled plasma-mass spectrometry or atomic emission spectroscopy (ICP-MS / ICP-AES)
Oxidation degree	No data	No data	-
Surface properties (atom %)	Non-coated	Non-coated	Supplier’s data
Physical state	Yellowish powder	Yellowish powder	Supplier’s data

The EpiOcular™-EIT model from MatTek Corp. (USA) uses the cornea-like non-keratinized tissue construct EpiOcular™ that is composed of normal human epidermal keratinocytes obtained from individual donors. The test was performed in two variants. Both NM-211 and NM-212 were submitted to protocol variant 1, i.e. as described by the supplier MatTek and Harbell et al. (2009), and protocol variant 2, i.e. as described in the OECD TG 492.

The test materials were applied undiluted at amounts enabling to cover the entire tissue surfaces (i.e., 61 mg for NM-211 and 28 mg for NM-212, corresponding to 50 μL bulk volume). Two tissues were treated with either the test materials, the negative control (NC) highly deionized water (50 µL), or the positive control (PC) methyl acetate (purity > 98 %; 50 µL). After test material application, the tissues were placed into the incubator for 90 min (variant 1), or 6 h (variant 2). After being washed, the tissues were immersed for 12 min (variant 1) or 25 min (variant 2) into 12-well plates, pre-filled with pre-warmed medium per well to remove test material residuals. Then, each tissue was dried and transferred to fresh 6-well plates filled with pre-warmed medium per well for 18-h incubation at standard culture conditions.

Upon completion of the post-exposure period, the assay medium was replaced by MTT solution for 3 h, and then the tissues were washed with PBS and incubated with isopropanol for at least 2 h to extract the produced formazan. The optical density of the formazan extracts was determined spectrophotometrically at a wavelength of 570 nm (OD570). Mean relative tissue viabilities ≤ 60 % indicated "irritancy to the eye"; mean relative tissue viabilities > 60 % indicated "no irritancy to the eye".

For all test materials, pre-tests as described in the OECD TG 492 were performed that precluded the test materials’ ability to directly reduce MTT.

The BCOP assay was conducted according to OECD TG 437. Bovine corneas were mounted in corneal holders. Both the anterior and posterior chambers of these holders were filled with Eagle’s MEM (without phenol red). After equilibration in a vertical position at 32°C for at least 1 h, the medium in both chambers was replaced with fresh medium and the initial corneal opacity was measured.

NM-211 and NM-212 powders were suspended in highly deionized water to achieve final concentrations of 20% (w/v) suspension (corresponding to 750 μL bulk volume) and then stirred with a magnetic stirrer. All test material suspensions were applied immediately after preparation. Three corneas were treated with either the test materials, NC, or PC. The medium in the anterior chamber was removed. The anterior chambers were filled with 750 μL highly deionized water (NC), or 750 μL of the 20% (w/v) imidazole solution in deionized water (PC). For NM-211 and NM-212, 750 μL of the 20% (w/v) test material preparation was applied directly to the epithelial surface of the cornea (i.e., using the "open chamber method"). The corneas were incubated in a horizontal position at 32°C for 4 h as prescribed for non-surfactant solids in the OECD TG. Upon completion of the incubation period, the NC, PC and test materials were removed from the anterior chamber, the respective epithelia were washed with EMEM, and both chambers were then refilled with fresh EMEM.

The final corneal opacity was measured, and the opacity change per cornea was calculated. To determine corneal permeability, the medium in the anterior chamber was replaced by sodium fluorescein solution (5 mg/mL) and incubated in a horizontal position for 90 min at 32°C. The amount of sodium fluorescein that permeated through the corneas was measured spectrophotometrically on 3 aliquots per cornea. The In Vitro Irritancy Score (IVIS) was calculated per treated cornea and then per treatment group was determined: IVIS = mean opacity value + (15 x mean permeability value). An IVIS > 55 indicated a risk of serious damage to the eyes.

Then, a histopathological evaluation of fixed corneas was performed using a standard method for light microscopy and a staining with Hematoxylin and Eosin. Histopathological findings were assessed in the epithelium (based on the depth of injury), and in the stroma. These findings were summarized in a so-called Histopathological Score of Irritation (HSI) assigned for each cornea ranging from 0 = no irritation to IV = severe irritation. HSI IV was assessed as "severe irritation"; HSI I, II and III were overall assessed as "non-severe irritation"; HSI 0 was regarded as no irritation.

Using the protocol "variant 1", the authors described that test material residues were observed on the EpiOcular™ tissues treated with CeO2 NM-211 and NM-212 after washing. However, since none of the test materials was able to reduce MTT directly, it was concluded that these residues did not interfere with the MTT assay. For both NM-211 and NM-212, mean tissue viabilities above 60% relative to the NC were calculated. Hence, none of the tested nano-CeO2 revealed eye irritation potential in the EpiOcular™-EIT under the chosen test conditions (i.e., indicating likelihood of "neither Category 1 nor 2").

Applying protocol "variant 2", mean relative tissue viabilities above 60% were determined for CeO2 NM-212 in the 1st test run conducted. Due to failure to meet the acceptance criteria for tissue variability or high optical density values recorded in single measurements of the 1st test run, two test runs each were conducted to evaluate the eye irritating potential of CeO2 NM-211. All individual and mean relative tissue variability values ranged above 90%. Therefore, all of these findings were assessed as indicating likelihood of "neither Category 1 nor 2".

According to the authors, NM-211 and NM-212 suspended in water (i.e., as prepared for the BCOP assay) were predominantly present as agglomerates around 1 μm diameter. Dispersed fractions of the total dose with diameters up to 1 μm that exceeded 50% of the total dose (i.e., indicating higher dispersibility) were recorded for CeO2 NM-212. Lower than 50% of the total dose, but still noteworthy fractions of small agglomerates below 100 nm were further recorded for CeO2 NM-211.

None of the tested nano-CeO2 induced serious eye damage in the BCOP as assessed by the respective IVIS that were far below the cut-off value of 55 (i.e., 18.5 and 16.9 for NM-211 and NM-212, respectively).

In addition, minimal findings (mostly minimal multifocal or diffuse desquamation) were observed for both NM-211 and NM-212, and an HSI of I (minimal) was assigned.

In conclusion, both tested nano-CeO2 did not elicit eye irritation in either the EpiOcular™-EIT or the BCOP assay pointing thus to the low eye irritation potential of either NM-211 or NM-212. According to Kolle SN et al., these results were consistent with available in vivo data for bulk material of the same composition.

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed (not irritating)

Respiratory irritation

Endpoint conclusion

Endpoint conclusion:

no study available

Additional information

SKIN IRRITATION

Three published studies are available on nano-CeO2 (Miyani VA and Hughes MF, 2016; Park B et al., 2007 and 2008); the three publications are thus used to conclude on the skin irritation / corrosion potential of nano-CeO2.

First, Park B et al. (2007, 2008) investigated the skin irritating potential of nano-CeO2 and bulk CeO2 in vitro. The study was well documented and met generally accepted scientific principles. The study was awarded a reliability score of 2 (Klimisch, 1997) and was flagged as key study.

The commercial nano-CeO2 and bulk CeO2 used in this study were only slightly characterised (see below in Table 1); however at the publication date, the scientific community required less physico-chemical endpoints in such articles.

 

Table 1: Physico-chemical characteristics available on the tested nano-CeO2 and bulk CeO2

Parameters	Park B et al. (2007, 2008)	Park B et al. (2007, 2008)	Miyani VA and Hughes MF (2016)	Miyani VA and Hughes MF (2016)	Miyani VA and Hughes MF (2016)
	Nano-CeO2	Bulk CeO2	-	-	-
Supplier	Energenics Europe Ltd. (EnviroxTM)	Aldrich	Alfa Aesar	NanoAmor	Umicore
Synthesis method	No data	No data	No data	No data	No data
Primary particle size	9 nm	320 nm	58 nm	8 nm	40 nm
Particle size distribution	No data	No data	No data	No data	No data
Stability	No data	No data	Aggregation based on surface charge	Aggregation based on surface charge	Aggregation based on surface charge
Specific surface area	94.7 m²/g	2.64 m²/g	No data	No data	No data
Surface charge (zeta potential)	No data	No data	-11.05 mV	-11.4 mV	-9.67 mV
Isoelectric point	No data	No data	No data	No data	No data
Shape	No data	No data	No data	No data	No data
Crystallinity	Crystalline (cerianite)	Crystalline (cerianite)	No data	No data	No data
Purity and impurities	Chemical composition: Ce and O	Chemical composition: Ce and O	No data	No data	No data
Solubility	No data	No data	No data	No data	No data
Oxidation degree	No data	No data	No data	No data	No data
Surface properties	Surface chemistry: Ce and O	Surface chemistry: Ce and O	No data	No data	No data

The EpiDerm™ EPI-200 human epidermal model from MatTek Corporation (USA) was selected in order to assess skin irritation. Fifty milligrams of test substance was applied to the human epidermal model for 960, 1200 and 1440 minutes. The irritation assessment was based on the measurement of cellular viability using the MTT assay. A mean irritation potential (MIP) score was calculated as follows: MIP = ET50 (reference material) / ET50 (test material) where ET50 was the exposure time at which relative value of optical density was 50% of the negative control and the reference material was SLS (1% of 20% (w/v)).

Nano-CeO2 exposure induced variations of cellular viability: 119.3% at 960 minutes, 67.5% at 1200 minutes, 64.5% at 1440 minutes. Exposure to bulk CeO2 caused no or only few changes in cell viability: 92.87% at 960 minutes, 97.36% at 1200 minutes, 86.23% at 1440 minutes. As nanoparticles and bulk material had a MIP value lower than 0.8 (i.e., < 0.01 for nano-CeO2 and 0.003 for bulk CeO2), both nano- and micro-CeO2 were not considered to have the potential to be an in vivo skin irritant based on this in vitro EpiDerm test.

In addition, Miyani VA and Hughes MF (2016) examined the potential of three different nano-CeO2 to elicit irritant contact dermatitis in vitro using the same model as Park and collaborators (2007/2008), that is a human skin equivalent model (HSEM) derived from keratinocytes. The study met generally accepted scientific principles. The study was awarded a reliability score of 2 (Klimisch, 1997) and was flagged as supporting study due to low charactrisations of the tested nanoparticles.

The three commercial nano-CeO2 used in this study were only slightly characterised (see above in Table 1).

The CeO2 particles were suspended in media containing 10% foetal bovine serum. The particles (1 mg/mL) were applied to the epidermal surface of the HSEM, the EpiDerm™ EPI-200 human epidermal model from MatTek Corporation (USA). Positive (5% sodium dodecyl sulphate (SDS) and negative controls (culture media) were included. After 1 h exposure at 37°C, the HSEM was washed with saline to remove the nanoparticles. Following 42-h incubation (37°C), HSEM viability was assessed using the MTT assay.

A test substance is considered a dermal irritant if the HSEM viability is < 50%. The mean viability for the SDS treated HSEM was 7.8%. The viabilities of the nanoparticle treated HSEM were 91% or greater. The CeO2 nanoparticles tested were not found to be dermal irritants under the conditions used in this study.

Given the low solubility of nano-CeO2 (Andreescu D et al., 2014; Brunner TJ et al., 2006; He X et al., 2010; Keller J et al., 2014; Molina RM et al., 2014; Van Hoecke K et al., 2009; Yokel RA et al., 2014) and based on published data available on nano-CeO2, no classification of nano-CeO2 for skin irritation is warranted according to the criteria of Regulation (EC) No. 1272/2008 (CLP/EU GHS).

SKIN CORROSION

There is no data available on the skin corrosive potential of nano-CeO2. However, since nano-CeO2 is not classified as irritant to the skin, the nanosized substance is thus considered as non-corrosive to the skin. Therefore, no classification of nano-CeO2 for skin corrosion is warranted according to the criteria of Regulation (EC) No. 1272/2008 (CLP/EU GHS).

EYE IRRITATION

Kolle SN et al. investigated the in vitro eye irritation potential of nanosized cerium dioxide (nano-CeO2). The study was conducted according to OECD guidelines 492 and 437. However, there was no mention to GLP. The study was designed to evaluate eye irritation potential of 20 nanosized materials and 3 micro-scale materials using a 2 -tier EpiOcular™ Eye Irritation Test (EpiOcular™-EIT) and BCOP testing strategy including histopathology of the bovine corneas. The study was performed, within the context of the OECD Sponsorship Program for the Testing of Manufactured Nanomaterials. The publication was thus awarded a reliability score of 2 (Klimisch, 1997) and was flagged as key study.

Two commercial nano-CeO2 (NM-211 and NM-212) provided by JRC (Italy) were used in this study. The nanomaterials were physico-chemically characterised by Kolle SN et al. (see in Table 2 below).

Table 2: Physico-chemical characteristics available on the tested nano-CeO2

Parameters	Kolle SN et al. (2016)	Kolle SN et al. (2016)
	NM-211	NM-212
Supplier	JRC	JRC
Synthesis method	Precipitation	Precipitation
Primary particle size	10.3 nm	33 nm
Particle size distribution	D50 of agglomerate fractions = 2.14 µm in water	D50 of agglomerate fractions = 0.776 µm in water
Stability	Agglomeration	Agglomeration
Specific surface area	66 m²/g	28 m²/g
Surface charge	No data available	No data available
Isoelectric point	No data available	No data available
Shape	Cubic	Cubic
Crystallinity	No data available	No data available
Purity and impurities	> 95% purity	> 99.5% purity
Solubility	< 0.1 ppm in water and DMEM + FCS (< 0.001% dissolution) or 0.002 wt% (Wohlleben W et al., 2013; Keller J et al., 2014)	< 0.1 ppm in water and DMEM + FCS (< 0.001% dissolution) or up to 0.02% depending on the medium (Wohlleben W et al., 2013; Keller J et al., 2014)
Oxidation degree	No data available	No data available
Surface properties	Non-coated	Non-coated
Physical sate	Yellowish powder	Yellowish powder

The EpiOcular™-EIT model from MatTek Corp. (USA) uses the cornea-like non-keratinized tissue construct EpiOcular™ that is composed of normal human epidermal keratinocytes obtained from individual donors. The test was performed in two variants. Both NM-211 and NM-212 were submitted to protocol variant 1, i.e. as described by the supplier MatTek and Harbell et al. (2009), and protocol variant 2, i.e. as described in the OECD TG 492.

The test materials were applied undiluted at amounts enabling to cover the entire tissue surfaces (i.e., 61 mg for NM-211 and 28 mg for NM-212, corresponding to 50 μL bulk volume). Two tissues were treated with either the test materials, the negative control (NC) highly deionized water (50 µL), or the positive control (PC) methyl acetate (purity > 98 %; 50 µL). After test material application, the tissues were placed into the incubator for 90 min (variant 1), or 6 h (variant 2). After being washed, the tissues were immersed for 12 min (variant 1) or 25 min (variant 2) into 12-well plates, pre-filled with pre-warmed medium per well to remove test material residuals. Then, each tissue was dried and transferred to fresh 6-well plates filled with pre-warmed medium per well for 18-h incubation at standard culture conditions.

Upon completion of the 18-h post-exposure period, the assay medium was replaced by MTT solution. After 3-h incubation, the tissues were washed with PBS and then incubated with isopropanol for at least 2 h to extract the produced formazan. The optical density of the formazan extracts was determined spectrophotometrically at a wavelength of 570 nm (OD570). Mean relative tissue viabilities ≤ 60 % indicated "irritancy to the eye"; mean relative tissue viabilities > 60 % indicated "no irritancy to the eye".

For all test materials, pre-tests as described in the OECD TG 492 were performed that precluded the test materials’ ability to directly reduce MTT.

The BCOP assay was conducted according to OECD TG 437. Bovine corneas were mounted in corneal holders. Both the anterior and posterior chambers of these holders were filled with Eagle’s MEM (without phenol red). After equilibration in a vertical position at 32°C for at least 1 h, the medium in both chambers was replaced with fresh medium and the initial corneal opacity was measured.

NM-211 and NM-212 powders were suspended in highly deionized water to achieve final concentrations of 20% (w/v) suspension (corresponding to 750 μL bulk volume) and then stirred with a magnetic stirrer. All test material suspensions were applied immediately after preparation. Three corneas were treated with the test materials, NC, or PC. The medium in the anterior chamber was removed. The anterior chambers were filled with 750 μL highly deionized water (NC), or 750 μL of the 20% (w/v) imidazole solution in deionized water (PC). For NM-211 and NM-212, 750 μL of the 20% (w/v) test material preparation was applied directly to the epithelial surface of the cornea (i.e., using the "open chamber method"). The corneas were incubated in a horizontal position at 32°C for 4 h as prescribed for non-surfactant solids in the OECD TG. Upon completion of the incubation period, the NC, PC and test materials were removed from the anterior chamber, the respective epithelia were washed with EMEM, and both chambers were then refilled with fresh EMEM.

The final corneal opacity was measured, and the opacity change per cornea was calculated. To determine corneal permeability, the medium in the anterior chamber was replaced by sodium fluorescein solution (5 mg/mL) and incubated in a horizontal position for 90 min at 32°C. The amount of sodium fluorescein that permeated through the corneas was measured spectrophotometrically on 3 aliquots per cornea. The In Vitro Irritancy Score (IVIS) was calculated per treated cornea and then per treatment group was determined: IVIS = mean opacity value + (15 x mean permeability value). An IVIS > 55 indicated a risk of serious damage to the eyes.

Then, a histopathological evaluation of fixed corneas was performed using a standard method for light microscopy and a staining with Hematoxylin and Eosin. Histopathological findings were assessed in the epithelium (based on the depth of injury), and in the stroma. These findings were summarized in a so-called Histopathological Score of Irritation (HSI) assigned for each cornea ranging from 0 = no irritation to IV = severe irritation. HSI IV was assessed as "severe irritation"; HSI I, II and III were overall assessed as "non-severe irritation"; HSI 0 was regarded as no irritation.

Using the protocol "variant 1", the authors described that test material residues were observed on the EpiOcular™ tissues treated with CeO2 NM-211 and NM-212 after washing. However, since none of the test materials was able to reduce MTT directly, it was concluded that these residues did not interfere with the MTT assay. For both NM-211 and NM-212, mean tissue viabilities above 60% relative to the NC were calculated. Hence, none of the tested nano-CeO2 revealed eye irritation potential in the EpiOcular™-EIT under the chosen test conditions (i.e., indicating likelihood of "neither Category 1 nor 2").

Applying protocol "variant 2", mean relative tissue viabilities above 60% were determined for CeO2 NM-212 in the 1st test run conducted. Due to failure to meet the acceptance criteria for tissue variability or high optical density values recorded in single measurements of the 1st test run, two test runs each were conducted to evaluate the eye irritating potential of CeO2 NM-211. All individual and mean relative tissue variability values ranged above 90%. Therefore, all of these findings were assessed as indicating likelihood of "neither Category 1 nor 2".

According to the authors, NM-211 and NM-212 suspended in water (i.e., as prepared for the BCOP assay) were predominantly present as agglomerates around 1 μm diameter. Dispersed fractions of the total dose with diameters up to 1 μm that exceeded 50% of the total dose (i.e., indicating higher dispersibility) were recorded for CeO2 NM-212. Lower than 50% of the total dose, but still noteworthy fractions of small agglomerates below 100 nm were further recorded for CeO2 NM-211.

None of the tested nano-CeO2 induced serious eye damage in the BCOP as assessed by the respective IVIS that were far below the cut-off value of 55 (i.e., 18.5 and 16.9 for NM-211 and NM-212, respectively).

In addition, minimal findings (mostly minimal multifocal or diffuse desquamation) were observed for both NM-211 and NM-212, and an HSI of I (minimal) was assigned.

In conclusion, both tested nano-CeO2 did not elicit eye irritation in either the EpiOcular™-EIT or the BCOP assay pointing thus to the low eye irritation potential of either NM-211 or NM-212. According to Kolle SN et al., these results were consistent with available in vivo data for bulk material of the same composition.

RESPIRATORY IRRITATION

For the respiratory irritation potential of nanoCeO2, no reliable study were identified.

Justification for classification or non-classification

The available data on nanometric cerium dioxide (nano-CeO2) show that irritation / corrosion of nano-CeO2 to the skin and the eyes is not expected. Thus, nano-CeO2 is not classified as skin or eye irritant according to the classification criteria of Regulation (EC) No. 1272/2008 (CLP/EU GHS) or UN GHS.