2,2,3,3,4,4,5,5-octafluoropentyl methacrylate

Short-term toxicity to fish

Link to relevant study record(s)

Reference

Reference 1

Endpoint:

fish embryo acute toxicity (FET)

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

February 1, 2017 to September 29, 2017

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

comparable to guideline study with acceptable restrictions

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 236 (Fish embryo acute toxicity (FET) test)

Version / remarks:

2013

Deviations:

yes

Remarks:

Species used was FATHEAD MINNOW (Pimephales promelas); no positive control

GLP compliance:

yes

Specific details on test material used for the study:

Purity: 99.7%

Analytical monitoring:

yes

Details on sampling:

- Concentrations: All test concentrations
- Sampling method: Water samples were collected from the newly prepared batch of test solutions of each treatment and control group at the beginning of the test on Day 0 and from old test solutions pooled from individual wells of the surrogate test chambers of each treatment and control group at 24 hours (±1 hour) to represent concentrations in the test system during each 24-hour renewal interval.
- Sample storage conditions before analysis: Samples were placed in 20 mL glass scintillation vials or 40 mL glass vials and processed immediately for analysis or stored under refrigerated conditions prior to analysis.

Vehicle:

no

Details on test solutions:

Each day a primary stock solution was prepared at a nominal concentration of 100 mg/L in pH adjusted UV sterilized well water. One day prior to the preparation of the test solutions, the pH of 8-L of UV sterilized well water was adjusted to a final pH of approximately 6.5 to 6.6 using concentrated phosphoric acid, aeration was added to the pH adjusted well water and allowed to stand overnight prior to use for the preparation of the test solutions. The final pH of the acidified dilution water was measured prior to use and ranged from 6.7 to 6.8.

A glass aspirator bottle was filled with test substance and dilution water with minimal headspace to achieve a final stock concertation of 100 mg/L. The bottle mouth was wrapped with Parafilm™, stirred for approximately 12 to 19 hours using a stir bar and a stir plate, and appeared clear and colorless with oil globules at the bottom of the bottle at the beginning of the mixing and appeared all clear and colorless with no visible precipitate at the end of the mixing period. At the end of the stirring, the solution was filtered through a 0.2-micron membrane filter. The filtrate was used as the highest test solution and as a primary stock solution to prepare the four additional test solutions. Proportional dilutions of the primary stock solution were made in dilution water to prepare nominal test. All test solutions appeared clear and colorless at test initiation, at the beginning of the test, at each renewal and at the end of the test with no visible precipitate noted.

Test organisms (species):

Pimephales promelas

Details on test organisms:

- Common name: fathead minnow;
- Justification for species: This species is representative of an important group of aquatic vertebrates and was selected for use in the test based upon past history of use in the laboratory. A non-GLP exploratory range-finding toxicity test was conducted prior to the definitive test to evaluate the effect of OFPMA with the embryos of zebrafish (Danio rerio). However, the species of the test organisms was changed from zebrafish to fathead minnow during the definitive study after unacceptable percent survival of the control in initial test trials.
- Source: Osage Catfisheries, Inc. of Osage Beach, Missouri
- Age of breeding stock: approximately 10 months of age and were sexually dimorphic adults that are reproductively mature and actively spawning.
- Breeding conditions: Twenty-four breeding groups of two males and four females were held under flow-through conditions for at least 14 days prior to collection of embryos.
- Collection of embryos: Prior to the onset of light on the day of exposure initiation, clean spawning substrates (inverted semi-circular sections of aged PVC pipe) were placed in 24 breeding tanks each containing 2 males and 4 females. After allowing sufficient time for egg deposition and fertilization, tiles containing newly-deposited eggs were collected from six breeding tanks. The eggs were gently removed from the tiles and pooled. The fertilization of the embryos was confirmed microscopically and the fertilization rate was approximately 90%. Embryos no older than 16-celled blastodisc developmental stage were selected and placed in the transfer chambers (Petri dishes) containing appropriate test solutions.

Test type:

semi-static

Water media type:

freshwater

Limit test:

no

Total exposure duration:

120 h

Remarks on exposure duration:

While development of fathead minnow embryos is essentially the same as zebrafish embryos, the time schedule of development is slightly slower. Therefore, exposure duration was extended to 5 days.

Hardness:

Negative control: 120-128 mg/L as CaCO3
Highest Treatment group: 16-32 mg/L as CaCO3

Test temperature:

24.2 to 26.0 °C

pH:

6.7 to 8.6

Dissolved oxygen:

≥ 75% saturation; no aeration

Conductivity:

Negative control: 309 to 312 µS/cm
Highest Treatment Group: 294 to 302 µS/cm

Nominal and measured concentrations:

Nominal: 6.3, 13, 25, 50, 100 mg/L (Proportional dilutions of primary 100 mg/L stock solution)
Nominal target concentrations: 1.1, 2.1, 4.3, 8.5, 17 mg/L
Initial measured concentration: 0.703, 1.55, 3.12, 5.95, 12.6 mg/L
Time-Weighted mean measured concentrations: 0.17, 0.32, 0.56, 0.96, 4.69 mg/L

Details on test conditions:

TEST SYSTEM
- Test vessel: clear polystyrene 24-well well plates (125 mm x 85 mm) with approximately 2 to 5 mL filling capacity per well and each well plate was equipped with a clear polystyrene lid.
- Type: Test chambers were loosely closed with lids following filling
- Material, size, headspace, fill volume: 2.0 mL of newly prepared test solutions and dilution water were added to the appropriate wells of each well plate using a pipette
- Renewal rate of test solution (frequency/flow rate): At each renewal at approximately 24-hour intervals, old solutions of each treatment and control group were removed from each well using a pipette. A small amount of old solution was left in each well to ensure that embryos remain covered with old test solutions to avoid drying of embryos. New solution was then added to each respective well.
- No. of organisms per vessel: 20 embryos per well plate
- No. of vessels per concentration: total of 20 replicates per group (i.e., each well is a replicate)
- No. of vessels per control: one 24-well plate for acidified well water control and one plate for negative control group; 4 additional embryos per plate maintained in dilution water served as internal plate controls

TEST MEDIUM / WATER PARAMETERS
Freshwater obtained from a well approximately 40 meters deep located on the EAG Laboratories-Easton site. The well water was passed through a sand filter to remove particles greater than approximately 25 µm, and pumped into a 37,800-L storage tank where the water was aerated with spray nozzles. Prior to use in the test system, the water was filtered to 0.45 µm to remove fine particles and was passed through an ultraviolet (UV) sterilizer.

OTHER TEST CONDITIONS
- Adjustment of pH: UV sterilized well water was adjusted to a final pH of approximately 6.5 to 6.6 using concentrated phosphoric acid
- Photoperiod: 16 hours of light and 8 hours of darkness with a 30-minute transition period
- Light intensity: fluorescent tubes that emit wavelengths similar to natural sunlight

EFFECT PARAMETERS MEASURED:
Observations of mortality, abnormal development or other observed effects, and hatching were recorded for each tested embryo at approximately 24-hour intervals during the exposure period. Apical observations performed on each tested embryo included coagulation of the embryo, lack of somite formation, non-detachment of the tail and lack of heartbeat.

These four apical observations were used for the determination of lethality in each tested embryo. Any positive outcome in one of these observations means that the embryo was dead. In addition, hatching was recorded in each treatment and control group daily beginning at 48 hours. Although hatching is not an endpoint used for the calculation of the LC50 value, hatching, if it occurs by Day 5, ensured the exposure of the embryo without a potential barrier function of the chorion.

RANGE FINDING
Embryos (no older than 16-celled blastodisc) were exposed to five test concentrations of 0.81, 2.7, 9.0, 30 and 100 mg/L nominal test concentration, a negative control and internal control, with one replicate of 20 embryos maintained in each treatment and control together with four additional internal controls for each treatment and negative control group under static-renewal conditions for 96 hours.

Embryos were observed daily for mortality, abnormal development and hatching at approximately 24-hour intervals during the exposure period. Based on the results of the preliminary range-finding test, nominal test concentrations of 6.3, 13, 25, 50 and 100 mg/L (1.1, 2.1, 4.3, 8.5 and 17 mg/L target nominal OFPMA concentrations), were selected for the definitive test.

Reference substance (positive control):

no

Duration:

48 h

Dose descriptor:

LC50

Effect conc.:

1.4 mg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: 95% confidence interval (1.1 - 2.1 mg/L)

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

1.1 mg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: 95% Confidence interval (0.96 - 4.7 mg/L)

Duration:

120 h

Dose descriptor:

LC0

Effect conc.:

0.56 mg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Duration:

120 h

Dose descriptor:

LC100

Effect conc.:

4.7 mg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Duration:

120 h

Dose descriptor:

LC50

Effect conc.:

1.1 mg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

test mat.

Basis for effect:

other: hatching success

Remarks on result:

other: 95% Confidence interval: (0.80 - 1.8 mg/L)

Details on results:

All fathead minnow embryos in the negative and internal control groups appeared normal throughout the test with no mortality noted.

The overall fertilization rate of all eggs collected was 90%.

The water temperatures were maintained at 25 ± 1ºC in test chambers throughout the test.

Overall survival of embryos in the negative (dilution water) control was 100%.

Hatching rate in the negative control was 100%.

The dissolved oxygen concentration in the negative (dilution water) control and highest test concentration were ≥75% (6.1 mg/L). However, the guideline is specifically for the test with zebrafish (Danio rerio), rather than fathead minnow which require different environmental conditions. In addition, the validity criteria for the early life-stage toxicity test recommended that the dissolved oxygen measurements should be >60% of the air saturation value throughout the test. Therefore, dissolved oxygen concentrations slightly below the guideline recommendation for zebrafish will not result in an adverse impact on the embryos of the fathead minnow.

Test solutions in the test chambers appeared clear and colorless during the test, with no evidence of precipitation observed. Measured concentrations of the samples collected from the newly prepared test solutions on Day 0 ranged from 63.9 to 74.2% of nominal target concentration. Measured concentrations of the samples collected from the Day 1 old test solutions ranged from
The measured concentration of the abiotic replicate at 17 mg/L nominal target concentration collected on Day 1 of the test was 16 mg/L, with the percent recovery of 94.5% of nominal. The result of the abiotic replicate confirmed that the preparation of the test solution was appropriate. The reason for the slight decline in the measured concentration in the abiotic chamber is not known but may be due to hydrolysis and/or through volatilization. The result from the abiotic replicate was excluded from the calculation of the time-weighted mean measured test concentration. The results of the study were based on the mean measured concentrations.

Validity criteria fulfilled:

no

Remarks:

A positive control was not included

Conclusions:

Following a methodology similar to OECD Guideline 236, fathead minnow (Pimephales promelas) embryos (no older than 16-celled blastodisc) were exposed to OFPMA. The 96 hour LC50 for mortality was 1.1 mg/L, with a 95% confidence interval of 0.96 to 4.7 mg/L based on the time-weighted mean measured concentration.

Executive summary:

A study was performed to assess the acute toxicity of OFPMA on embryonic fathead minnows (Pimephales promelas) under semi-static conditions for 120 hours. The method followed was similar to OECD Guideline 236, “Fish Embryo Acute Toxicity (FET) Test”.

 

Each day a primary stock solution was prepared at a nominal concentration of 100 mg/L in pH adjusted UV sterilized well water. The dilution bottle was nearly filled with the dilution water to minimize headspace to reduce a possibility of loss of material through volatility. The bottle mouth was wrapped with Parafilm, stirred for approximately 12 to 19 hours using a stir bar and a stir plate. At the end of the stirring, the solution was filtered through a 0.2-micron membrane filter. Proportional dilutions of the primary stock solution were made in dilution water to prepare the nominal test solutions of 6.3, 13, 25 and 50 mg/L. Based on available solubility information, the target concentration in the test solution were 1.1, 2.1, 4.3, 8.5, and 17 mg/L.

 

Preliminary range-finding tests with zebrafish (Danio rerio) embryos prompted a change in species due to the unacceptable percent survival of the control zebrafish embryos in the initial trials. Following range-finding tests (conducted with fathead minnow (Pimephales promelas) embryos using nominal concentrations of 0.81, 2.7, 9.0, 30, and 100 mg/L), fathead minnow embryos were exposed to nominal test concentrations of 6.3, 13, 25, 50 and 100 mg/L (1.1, 2.1, 4.3, 8.5 and 17 mg/L target nominal OFPMA concentrations) for the definitive test. For the definitive test, one 24-well well plate per each treatment, acidified well water control and negative control group. Each well plate containing 20 embryos per treatment, acidified water control or negative control and 4 additional embryos per plate maintained in dilution water serving as internal plate controls. The study was conducted for 120 hours, with renewal of test solutions at 24-hour internals. Test chambers were clear polystyrene 24-well well plates (125 mm x 85 mm) with approximately 2 to 5 mL filling capacity per well and each well plate was equipped with a clear polystyrene lid. The individual wells in each plate were considered independent replicates.

 

In this study, the highest test concentration causing no mortality at test end and the lowest test concentration causing 100% mortality at test end were 0.56 and 4.7 mg/L, respectively. Based on the time-weighted mean measured concentrations, the 48-hour LC50 value based on mortality was 1.4 mg/L, with a 95% confidence interval of 1.1 to 2.1 mg/L, while the 96-hour LC50 value was 1.1 mg/L, with a 95% confidence interval of 0.96 to 4.7 mg/L.

 

ECHA has stated that OECD Guideline 236 cannot be used to directly address REACH Annex VIII, 9.1.3 endpoint of short-term toxicity on fish, rather it may be used within a weight of evidence approach. Within this dossier, several QSARs predictions have been calculated which support the acute toxicity findings of this study. Therefore, while no positive control was included with this study, the results are still acceptable as part of the weight of evidence approach.

Based on a weight of evidence approach which includes this FET study, the median fish LC50 prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Reference 2

Endpoint:

short-term toxicity to fish

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

results derived from a (Q)SAR model, with limited documentation / justification, but validity of model and reliability of prediction considered adequate based on a generally acknowledged source

Justification for type of information:

1. SOFTWARE: Danish QSAR Database (DTU)

2. MODEL: Fathead minnow 96h LC50 (mg/L) Battery prediction based on Leadscope and SciQSAR

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL: C(F)(F)(C(F)(F)C(F)(F)C(F)F)COC(=O)C(=C)C

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: Fathead minnow 96h LC50 (mg/L)
- Unambiguous algorithm: Details not provided, a battery prediction based on the QSAR predictions from Leadscope and SciQSAR are reported
- Defined domain of applicability:
Leadscope: Domain is not provided
SciQSAR: Domain is not provided
- Appropriate measures of goodness-of-fit and robustness and predictivity:
Leadscope: R² = 0.75, Q² = 0.73
SciQSAR: R² = 0.74, Q² = 0.72
- Mechanistic interpretation:
Information not provided

5. APPLICABILITY DOMAIN
- Descriptor domain: Descriptor domains were not provided. The DTU output states the predictions are “IN” the domain.
- Structural and mechanistic domains / Similarity with analogues in the training set: As the training sets were not provided, it is not possible to confirm the target substance is within the domain.

6. ADEQUACY OF THE RESULT
The prediction provides a discrete LC50 value for acute fish toxicity. While it is not possible to confirm the applicability of the prediction, this prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Guideline:

other: Danish QSAR Database (DTU)

Principles of method if other than guideline:

- Software tool(s) used including version: Danish QSAR Database (DTU)
- Model(s) used: Fathead minnow 96h LC50 (mg/L) Battery prediction based on Leadscope and SciQSAR
- Justification of QSAR prediction: see field 'Justification for type of information'

Specific details on test material used for the study:

SMILES: C(F)(F)(C(F)(F)C(F)(F)C(F)F)COC(=O)C(=C)C

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

43.4 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value (Battery value)

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

0.11 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value (Leadscope value)

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

86.6 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value (SciQSAR value)

Details on results:

The Battery prediction of the Leadscope and SciQSAR values is made using a so-called battery algorithm. The battery approach can give more reliable predictions and can also expand the applicability domain. For a given effect, QSAR predictions are made in each of the independent QSAR model systems and combined into a battery prediction. With the battery approach, it is in many cases possible to reduce "noise" from the individual model estimates and thereby improve the accuracy and/or broaden the applicability domain.

Validity criteria fulfilled:

no

Remarks:

it was not possible to confirm the applicability domain due to the lack of details provided on the training sets and domain descriptors

Conclusions:

The Danish QSAR Database (DTU) Fathead minnow 96h LC50 (mg/L) Battery prediction, based on Leadscope and SciQSAR, predicted an LC50 of 43.4 mg/L. With the battery approach, it is in many cases possible to reduce "noise" from the individual model estimates and thereby improve the accuracy and/or broaden the applicability domain.

While it was not possible to confirm the applicability domain due to the lack of details provided on the training sets and domain descriptors, this prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Based on a weight of evidence approach, the median fish LC50 prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Reference 3

Endpoint:

short-term toxicity to fish

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model, but not (completely) falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE: ECOSAR (US EPA) Version 1.11

2. MODEL: Fish, 96-hr, LC50

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL: FC(F)C(F)(F)C(F)(F)C(F)(F)COC(=O)C(C)=C

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: Fish, 96-hr, LC50
- Unambiguous algorithm:
Class-specific estimations: Methacrylates - Log 96-h LC50 (mmol/L) = –0.7088 (log Kow) + 0.6751
Baseline toxicity: Neutral Organic - Log LC50 (mmol/L) = -0.8981 log Kow + 1.7108
- Defined domain of applicability:
Hydrolytically stable chemicals, i.e., half-life greater than 14 days.
Maximum Log Kow = 5
Maximum molecular weight 1000 Daltons
- Appropriate measures of goodness-of-fit and robustness and predictivity: R² = 0.7299
- Mechanistic interpretation:
The methacrylates are alkyl carbon, olefinic carbon, acetylenic carbon, or aromatic carbon methacrylates.
Neutral Organics have no specific structure; if there are no positive class-specific identifications located, this class is assigned.

5. APPLICABILITY DOMAIN
- Descriptor domain:
Hydrolytically stable chemicals: the target chemical is outside the domain with a hydrolysis ½ life of less than 14 days
Maximum Log Kow = 5: the target chemical is within the domain with a log Kow of 3.03
Maximum molecular weight 1000 Daltons: the target chemical is within the domain at 300.15 Daltons
- Structural and mechanistic domains / Similarity with analogues in the training set:
While the target chemical is accurately described as a methacrylate, the alkyl fluoride portion of the chemical was not represented within the methacrylate training set. While there are alkyl fluoride compounds within the Neutral Organics training set, none of these alkyl fluorides are acrylates.

6. ADEQUACY OF THE RESULT
The predictions provides discrete LC50 values for acute fish toxicity. However, the Methacrylates ECOSAR class does not include any short-chain polyfluoroalkyl esters, and the Neutral Organic SAR ECOSAR class does not include any methacrylates within their training sets. While not adequate on their own, the predictions form part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Guideline:

other: ECOSAR (US EPA) Version 1.11

Principles of method if other than guideline:

- Software tool(s) used including version: ECOSAR (US EPA) Version 1.11
- Model(s) used: Fish, 96-hr, LC50
- Justification of QSAR prediction: see field 'Justification for type of information'

Specific details on test material used for the study:

SMILES: FC(F)C(F)(F)C(F)(F)C(F)(F)COC(=O)C(C)=C

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

10.11 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value (ECOSAR Class: Methacrylates)

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

29.301 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value (ECOSAR Class: Neutral Organic SAR)

Validity criteria fulfilled:

no

Remarks:

The polyfluoroalkyl portion is not represented within the Methacrylates training set, and the methacrylate portion is not represented within the Neutral Organics training set.

Conclusions:

Fish, 96-hr, LC50 by the US EPA ECOSAR Version 1.11 predicted an LC50 of 10.110 mg/L based on the Methacrylates ECOSAR class and 29.301 mg/L based on the Neutral Organic SAR ECOSAR class.

The target chemical is a short-chain polyfluoroalkyl methacrylate, but the polyfluoroalkyl portion is not represented within the Methacrylates training set, and while there are short-chain polyfluoroalkyl compounds within the Neutral Organics training set, none of them is a methacrylate. The predictions form part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Based on a weight of evidence approach, the median fish LC50 prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Reference 4

Endpoint:

short-term toxicity to fish

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model, but not (completely) falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE: US EPA Toxicity Estimation Software Tool (T.E.S.T.) Version 4.2

2. MODEL: Predicted Fathead minnow LC50 (96 hr) for 355-93-1 from Consensus method

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL: CAS Number 355-93-1

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: Fathead minnow LC50 (96 hr)
- Unambiguous algorithm:
In the consensus method, the predicted toxicity is simply the average of the predicted toxicities from the other QSAR methodologies, specifically hierarchical clustering and single model, (taking into account the applicability domain of each method). If only a single QSAR methodology can make a prediction, the predicted value is deemed unreliable and not used. This method typically provides the highest prediction accuracy since errant predictions are dampened by the predictions from the other methods. In addition, this method provides the highest prediction coverage because several methods with slightly different applicability domains are used to make a prediction.
- Defined domain of applicability / Mechanistic interpretation:
The descriptors used are the following:
- SsssN: Sum of >N- E-States
- SdssS_acnt: Count of =S<
- MDEC33: Molecular distance edge between all tertiary carbons
- MDEO11: Molecular distance edge between all primary oxygens
- BEHm2: Highest eigenvalue n. 2 of Burden matrix / weighted by atomic masses
- nDB: Number of double bonds
- nS: Number of Sulfur atoms
- nR09: Number of 9-membered rings
- ATS5p: Broto-Moreau autocorrelation of a topological structure - lag 5 / weighted by atomic
polarizabilities
- MATS7e: Moran autocorrelation - lag 7 / weighted by atomic Sanderson electronegativities
- GATS3e: Geary autocorrelation - lag 3 / weighted by atomic Sanderson electronegativities
- SRW03: Self-returning walk count of order 3
- ALOGP: Ghose-Crippen octanol water coefficient
- NO2 [aromatic attach]: -NO2 [aromatic attach] fragment count
- CHO [aliphatic attach]: -CHO [aliphatic attach] fragment count
- OH [phosphorus attach]: -OH [phosphorus attach] fragment count
- >NN=O: >NN=O fragment count
- C(=O)- [2 nitrogen attach]: -C(=O)- [2 nitrogen attach] fragment count
- C#N [aliphatic attach]: -C#N [aliphatic attach] fragment count
- Cl [olefinic attach]: -Cl [olefinic attach] fragment count
- CHO [aromatic attach]: -CHO [aromatic attach] fragment count
- Appropriate measures of goodness-of-fit and robustness and predictivity:
Mean absolute error in -Log10(mol/L): Entire set 0.48
Mean absolute error in -Log10(mol/L): Similarity coefficient (≥ 0.5) 0.17

5. APPLICABILITY DOMAIN
- No chemicals in the test set exceed a minimum similarity coefficient of 0.5 for comparison purposes
- Only two analogues formed the dataset

6. ADEQUACY OF THE RESULT
The prediction provides a discrete LC50 value for acute fish toxicity. However, only two analogues formed the dataset for this prediction, and although both contained polyfluorinated alkyl groups, neither was a methacrylate. While not adequate on its own, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Guideline:

other: US EPA Toxicity Estimation Software Tool (T.E.S.T.) Version 4.2

Principles of method if other than guideline:

- Software tool(s) used including version: US EPA Toxicity Estimation Software Tool (T.E.S.T.) Version 4.2
- Model(s) used: Predicted Fathead minnow LC50 (96 hr) for 355-93-1 from Consensus method
- Justification of QSAR prediction: see field 'Justification for type of information'

Specific details on test material used for the study:

CAS Number: 355-93-1

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

3.59 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value

Validity criteria fulfilled:

no

Remarks:

The model lacked the necessary level of robustness as only 2 analogues formed the data set

Conclusions:

The US EPA Toxicity Estimation Software Tool (T.E.S.T.) Version 4.2 US EPA Toxicity Estimation Software Tool (T.E.S.T.) Version 4.2 model predicted an LC50 of 3.59 mg/L.

Only two analogues formed the dataset for this prediction, and although both containted polyfluorinated alkyl groups, neither was a methacrylate. While not adequate on its own, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Based on a weight of evidence approach, the median fiah LC50 prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Reference 5

Endpoint:

short-term toxicity to fish

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model, but not (completely) falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE: QSAR Toolbox 4.1

2. MODEL: Trend analysis, executed via AW "Ecotoxicological Endpoint"

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL: CC(=C)C(=O)OCC(F)(F)
C(F)(F)C(F)(F)C(F)F

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: EC50, LC50; Mortality; Pimephales promelas; 96h; No guideline specified
- Unambiguous algorithm: Linear approximation: EC50, LC50 = 1.77 (±0.733) + 1.06 (±0.253) * log Kow, log (1/mol/L)
- Defined domain of applicability:
log Kow: from 0.27 to 7.12
EC50 LC50: from 2.44 to 9.38 log(1/mol/L) (from 0.000195 to 582 mg/L)
Organic functional groups, Norbert Haider (checkmol) (primary grouping)
Substance type (subcategorization)
Protein binding by OASIS (subcategorization)
- Appropriate measures of goodness-of-fit and robustness and predictivity:
The prediction is based on 10 values within range 0.000195 - 582 mg/L
Prediction confidence range (0.95%): ± 1.5 log(1/mol/L)
R2 = 0.922
- Mechanistic interpretation:
Organic functional groups, Norbert Haider (checkmol) (primary grouping)
Halogen derivative;
Alkyl fluoride;
Alkyl halide
Substance type (subcategorization)
Discrete chemical;
Mono constituent (predefined);
Organic
Protein binding by OASIS (subcategorization)
No alert found

5. APPLICABILITY DOMAIN
- Descriptor domain:
log Kow: from 0.27 to 7.12 – target chemical is in domain with a log Kow of 3.03
- Structural and mechanistic domains / Similarity with analogues in the training set:
While the target chemical is accurately described as an alkyl fluoride, the methacrylate portion of the chemical was not represented within the 10 analogues the prediction is based upon.

6. ADEQUACY OF THE RESULT
The prediction provides a discrete LC50 value for acute fish toxicity, the methacrylate portion of the chemical was not represented within the dataset used for the prediction. While not adequate on its own, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Guideline:

other: QSAR Toolbox 4.1

Principles of method if other than guideline:

- Software tool(s) used including version: QSAR Toolbox 4.1
- Model(s) used:Trend analysis, executed via AW "Ecotoxicological Endpoint"
- Justification of QSAR prediction: see field 'Justification for type of information'

Specific details on test material used for the study:

SMILES: CC(=C)C(=O)OCC(F)(F)C(F)(F)C(F)(F)C(F)F

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

0.381 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value (from 0.0122 to 11.9)

Validity criteria fulfilled:

no

Remarks:

The methacrylate portion of the chemical was not represented within the 10 analogues the prediction is based upon

Conclusions:

Trend analysis, executed via AW "Ecotoxicological Endpoint" by the QSAR Toolbox Version 4.1 predicted an LC50 of 0.381 mg/L (from 0.0122 to 11.9 mg/L).

The target chemical is a short-chain polyfluoroalkyl methacrylate, but the methacrylate portion was not represented within the 10 analogues the prediction is based upon. The prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Based on a weight of evidence approach, the median QSAR prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Reference 6

Endpoint:

short-term toxicity to fish

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model, but not (completely) falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE: VEGA 1.1.4

2. MODEL: Fathead Minnow LC50 96h (EPA) Version 1.0.7

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL: O=C(OCC(F)(F)C(F)(F)C(F)(F)C(F)F)C(=C)C

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: Fathead Minnow LC50 96h
- Unambiguous algorithm: linear regression made on 21 molecular descriptors
- Defined domain of applicability / Mechanistic interpretation:
The descriptors used are the following:
- SsssN: Sum of >N- E-States
- SdssS_acnt: Count of =S<
- MDEC33: Molecular distance edge between all tertiary carbons
- MDEO11: Molecular distance edge between all primary oxygens
- BEHm2: Highest eigenvalue n. 2 of Burden matrix / weighted by atomic masses
- nDB: Number of double bonds
- nS: Number of Sulfur atoms
- nR09: Number of 9-membered rings
- ATS5p: Broto-Moreau autocorrelation of a topological structure - lag 5 / weighted by atomic
polarizabilities
- MATS7e: Moran autocorrelation - lag 7 / weighted by atomic Sanderson electronegativities
- GATS3e: Geary autocorrelation - lag 3 / weighted by atomic Sanderson electronegativities
- SRW03: Self-returning walk count of order 3
- ALOGP: Ghose-Crippen octanol water coefficient
- NO2 [aromatic attach]: -NO2 [aromatic attach] fragment count
- CHO [aliphatic attach]: -CHO [aliphatic attach] fragment count
- OH [phosphorus attach]: -OH [phosphorus attach] fragment count
- >NN=O: >NN=O fragment count
- C(=O)- [2 nitrogen attach]: -C(=O)- [2 nitrogen attach] fragment count
- C#N [aliphatic attach]: -C#N [aliphatic attach] fragment count
- Cl [olefinic attach]: -Cl [olefinic attach] fragment count
- CHO [aromatic attach]: -CHO [aromatic attach] fragment count
- Appropriate measures of goodness-of-fit and robustness and predictivity:
Training set: n = 652; R² = 0.69; RMSE = 0.83
Test set: n = 164; R² = 0.64; RMSE = 0.89

5. APPLICABILITY DOMAIN
- no similar compounds with known experimental value in the training set have been found
- similar molecules found in the training set have experimental values that disagree with the predicted value
- some atom centered fragments of the compound have not been found in the compounds of the training set or are rare fragments (2 infrequent fragments found)

6. ADEQUACY OF THE RESULT
The prediction provides a discrete LC50 value for acute fish toxicity. However, the target short-chain polyfluoroalkyl methacrylate is not within the applicability domain of the model. While not adequate on its own, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Guideline:

other: VEGA 1.1.4 Fathead Minnow LC50 96h (EPA) Version 1.0.7

Principles of method if other than guideline:

- Software tool(s) used including version: VEGA 1.1.4
- Model(s) used: Fathead Minnow LC50 96h (EPA) Version 1.0.7
- Justification of QSAR prediction: see field 'Justification for type of information'

Specific details on test material used for the study:

SMILES: O=C(OCC(F)(F)C(F)(F)C(F)(F)C(F)F)C(=C)C

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

9.98 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value

Validity criteria fulfilled:

no

Remarks:

the predicted compound is outside the Applicability Domain of the model

Conclusions:

The VEGA 1.1.4 Fathead Minnow LC50 96h (EPA) Version 1.0.7 model predicted an LC50 of 9.98 mg/L.

While the target short-chain polyfluoroalkyl methacrylate is not within the applicability domain of the model, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Based on a weight of evidence approach, the median fish LC50 prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Reference 7

Endpoint:

short-term toxicity to fish

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model, but not (completely) falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE: VEGA 1.1.4

2. MODEL: Fish Acute (LC50) Toxicity model (KNN/Read-Across) 1.0.0

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
O=C(OCC(F)(F)C(F)(F)C(F)(F)C(F)F)C(=C)C

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: Fish Acute (LC50) Toxicity
- Unambiguous algorithm:
The model performs a read-across similarity index based on structural aspects of the compounds.
- Defined domain of applicability:
The index takes into account several structural aspects of the compounds, such as their fingerprint, the number of atoms, of cycles, of heteroatoms, of halogen atoms, and of particular fragments (such as nitro groups).
- Appropriate measures of goodness-of-fit and robustness and predictivity:
n = 936; R² = 0.65; RMSE = 0.71
Non predicted compounds: n = 36

5. APPLICABILITY DOMAIN
- some atom centered fragments of the compound have not been found in the compounds of the training set or are rare fragments (2 infrequent fragments found; FC(F)(C)C and FC(F)C)

6. ADEQUACY OF THE RESULT
The model performs a read-across and provides a quantitative prediction of acute toxicity in fish. While the target short-chain polyfluoroalkyl methacrylate is not within the applicability domain of the model, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Guideline:

other: VEGA 1.1.4 Fish Acute (LC50) Toxicity model (KNN/Read-Across) 1.0.0

Principles of method if other than guideline:

- Software tool(s) used including version: VEGA 1.1.4
- Model(s) used: Fish Acute (LC50) Toxicity model (KNN/Read-Across) 1.0.0
- Justification of QSAR prediction: see field 'Justification for type of information'

Specific details on test material used for the study:

SMILES: O=C(OCC(F)(F)C(F)(F)C(F)(F)C(F)F)C(=C)C

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

8.28 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value

Validity criteria fulfilled:

no

Remarks:

the predicted compound is outside the Applicability Domain of the model

Conclusions:

The VEGA 1.1.4 Fish Acute (LC50) Toxicity model (KNN/Read-Across) 1.0.0 model predicted an LC50 of 8.28 mg/L.
 
While the target short-chain polyfluoroalkyl methacrylate is not within the applicability domain of the model, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Based on a weight of evidence approach, the median fish LC50 prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Reference 8

Endpoint:

short-term toxicity to fish

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model, but not (completely) falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE: VEGA 1.1.4

2. MODEL: Fish Acute (LC50) Toxicity model (NIC) 1.0.0

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
O=C(OCC(F)(F)C(F)(F)C(F)(F)C(F)F)C(=C)C

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: Fish Acute (LC50) Toxicity
- Unambiguous algorithm:
The model has been built as a Counter Propagation Artificial Neural Network (CP ANN) by the
National Institute of Chemistry Slovenia (NIC)..
- Defined domain of applicability:
The descriptors used by the CPANN are the following:
- nCIR: number of circuits
- nTB: number of Triple Bonds
- nP: number of P atoms
- nCL: number of Cl atoms
- nR10: number of 10-membered rings
- TI1: first Mohar index from Laplace matrix
- S2K: 2-path Kier alpha-modified shape index
- T(N..P): sum of topological distances between N..P
- T(P..Cl): sum of topological distances between P..Cl
- T(Cl..Cl): sum of topological distances between Cl..Cl
- piPC09: molecular multiple path count of order 9
- PCR: ratio of multiple path count over path count
- Xindex: Balaban X index
- MATS1e: Moran autocorrelation - lag 1/weighted by Sanderson electronegativity
- GATS7m: Geary autocorrelation - lag 7/weighted by mass
- EEig14x: eigenvalue n. 14 from edge adjacency matrix weighted by edge degree
- EEig14d: eigenvalue n. 14 from edge adjacency matrix weighted by dipole moment
- ESpm01d: spectral moment of order 1 rom edge adjacency matrix weighted by dipole moment
- GGI4: topological charge index of order 4
- Seigv: spectral moment of order 1 from Brysz matrix weighted by van der Waals volume
- Appropriate measures of goodness-of-fit and robustness and predictivity:
Training set: n = 564; R² = 0.89; RMSE = 0.42
Test+Evaluation set: n = 382; R² = 0.54; RMSE = 0.90

5. APPLICABILITY DOMAIN
- only moderately similar compounds with known experimental value in the training set have been found
- 1 descriptor(s) for this compound have values outside the descriptor range of the compounds of the training set.
- some atom centered fragments of the compound have not been found in the compounds of the training set or are rare fragments (2 infrequent fragments found: FC(F)(C)C and FC(F)C)

6. ADEQUACY OF THE RESULT
The model provides a quantitative prediction of acute toxicity in fish (LC50). While the target short-chain polyfluoroalkyl methacrylate is not within the applicability domain of the model, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Guideline:

other: VEGA 1.1.4 Fish Acute (LC50) Toxicity model (NIC) 1.0.0

Principles of method if other than guideline:

- Software tool(s) used including version: VEGA 1.1.4
- Model(s) used: Fish Acute (LC50) Toxicity model (NIC) 1.0.0
- Justification of QSAR prediction: see field 'Justification for type of information'

Specific details on test material used for the study:

SMILES: O=C(OCC(F)(F)C(F)(F)C(F)(F)C(F)F)C(=C)C

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

61.25 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value

Validity criteria fulfilled:

no

Remarks:

the predicted compound is outside the Applicability Domain of the model

Conclusions:

The VEGA 1.1.4 Fish Acute (LC50) Toxicity model (NIC) 1.0.0 model predicted an LC50 of 61.25 mg/L.

While the target short-chain polyfluoroalkyl methacrylate is not within the applicability domain of the model, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Based on a weight of evidence approach, the median fish LC50 prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Reference 9

Endpoint:

short-term toxicity to fish

Type of information:

(Q)SAR

Adequacy of study:

weight of evidence

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

results derived from a valid (Q)SAR model, but not (completely) falling into its applicability domain, with adequate and reliable documentation / justification

Justification for type of information:

1. SOFTWARE: VEGA 1.1.4

2. MODEL: Fish Acute (LC50) Toxicity classification (SarPy/IRFMN) 1.0.2

3. SMILES OR OTHER IDENTIFIERS USED AS INPUT FOR THE MODEL
O=C(OCC(F)(F)C(F)(F)C(F)(F)C(F)F)C(=C)C

4. SCIENTIFIC VALIDITY OF THE (Q)SAR MODEL
- Defined endpoint: Fish Acute (LC50) Toxicity classification
- Unambiguous algorithm:
The model has been built as three sets of rules. If no rules apply to the given compound, the
prediction will be class 4.
- Defined domain of applicability:
Following, the list of the alerts, as SMARTS strings, related to class 1 toxicity:
C(OCCCC)c1cccc(c1)
c1cc(c(O)c(c1)C(C)C)
Oc1ccc(cc1C)Cl
O=Cc1cccc(Oc2ccc(cc2))c1
Nc1ccc(cc1)CCCCCCCC
O=[N+]([O-])c1c(cc(c(c1)[N+](=O)[O-])Cl)Cl
NCCCCCCCCCCCC
O(CC)P(=S)(OCC)SCS
CC[Sn](CC)(CC)CC
c1cc(ccc1CCl)
O=CC=CC(=O)
N#CCC#N
[c]([Cl,Br,F])[c]([Cl,Br,F])[c]([Cl,Br,F])
CCC(=O)OC[a]
Following, the list of the alerts, as SMARTS strings, related to class 2 toxicity:
Oc1ccc(cc1)Cl
C(C)CCCCCCCCC
c1ccc(Oc2ccccc2)cc1
C(=O)Oc1ccccc1
C(OCC=C)CC
c1c(cccc1CC)CC
O(CC)P(=S)(O)
C=CC=C
C(=O)OCCCC
O=[N+]([O-])c1cc(cc(c1)C)
c1cc(c(cc1C))C
SCC
O(c1ccccc1)CCCC
c1cccc2ccccc12
c1cc(ccc1O)Br
c1ccccc1c2ccccc2
O=Cc1c(F)cccc1
I
O=C(OC)CC
[*;D1]#C[C;!D4][!C;D1]
*[C;D2][C;D2][C;D2][C;D2][C;D2][C;D2]*
[s;R]
[$([c]([Cl,Br,F])[c]([Cl,Br,F])),$([c]([Cl,Br,F])[c][c]([Cl,Br,F]))]
Following, the list of the alerts, as SMARTS strings, related to class 1 toxicity:
C(OC)c1ccc(cc1)
NCCCCCC
c1cc(c(cc1)C)C
Fc1ccccc1
O=C(OCCC)C
c1ccc(cc1)Br
O=[N+]([O-])c1cc(N)ccc1
c1ccc(cc1)CCCC
C=CCCC
c1ccc(cc1)Cl
Oc1ccc(cc1)C
c1cc(ccc1N(C)C)
c1cc(N)ccc1O
CCCCCCCC
C(C(Cl))Cl
O=P(OCC)(OCC)OCC
N(CCC)(CCC)C
o1c(ccc1)
C#CCC(O)
C(CCCl)C
OCC#CC
O=[C;D2][C;D2]
C=C
c1ccccc1
- Appropriate measures of goodness-of-fit and robustness and predictivity:
Training set: n = 567; Accuracy = 0.69

5. APPLICABILITY DOMAIN
- only moderately similar compounds with known experimental value in the training set have been found
- some atom centered fragments of the compound have not been found in the compounds of the training set or are rare fragments (2 infrequent fragments found)
The following relevant fragments have been found: Toxicity class 2 alert no. 5; Toxicity class 2 alert no. 9; Toxicity class 2 alert no. 19; Toxicity class 3 alert no. 5; Toxicity class 3 alert no. 23

6. ADEQUACY OF THE RESULT
The model provides a classification for toxicity to fish provided by Directive 92/32/EEC of the EU for dangerous substances. While the target short-chain polyfluoroalkyl methacrylate may not be within the applicability domain of the model, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Reason / purpose for cross-reference:

reference to other study

Guideline:

other: VEGA 1.1.4 Fish Acute (LC50) Toxicity classification (SarPy/IRFMN) 1.0.2

Principles of method if other than guideline:

- Software tool(s) used including version: VEGA 1.1.4
- Model(s) used: Fish Acute (LC50) Toxicity classification (SarPy/IRFMN) 1.0.2
- Justification of QSAR prediction: see field 'Justification for type of information'

Specific details on test material used for the study:

SMILES: O=C(OCC(F)(F)C(F)(F)C(F)(F)C(F)F)C(=C)C

Duration:

96 h

Dose descriptor:

LC50

Effect conc.:

>= 1 - < 10 mg/L

Conc. based on:

test mat.

Basis for effect:

mortality (fish)

Remarks on result:

other: QSAR predicted value

Validity criteria fulfilled:

no

Remarks:

the predicted compound could be out of the Applicability Domain of the model

Conclusions:

The VEGA 1.1.4 Fish Acute (LC50) Toxicity classification (SarPy/IRFMN) 1.0.2 model predicted an a toxicity class of Toxic-2 (between 1 and 10 mg/l).

While the target short-chain polyfluoroalkyl methacrylate may not be within the applicability domain of the model, the prediction forms part of a weight-of-evidence approach which combines various QSAR results and the results of a Fish Embryo Toxicity (FET) study.

Based on a weight of evidence approach, the median fish LC50 prediction is 10 mg/L. Based on the measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae appears to be the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

Description of key information

A GLP-compliant study similar to OECD Guideline 236 was conducted to assess the acute toxicity of octafluoropentyl methacrylate (OFPMA) to fathead minnow (Pimephales promelas) embryos. The study was initiated with less than 24 hours old fathead minnow (Pimephales promelas) embryos at nominal concentrations of 6.3, 13, 25, 50, and 100 mg/L (corresponding to time-weighted mean (TWM) measured concentrations of 0.17, 0.32, 0.56, 0.96 and 4.7 mg/L) under static renewal conditions for 5 days. The 96-hour LC50 for mortality was 1.1 mg/L (TWM measured concentration). The highest test concentration causing no mortality at test end and the lowest test concentration causing 100% mortality at test end were 0.56 and 4.7 mg/L (TWM measured concentrations), respectively.

ECHA (2016a) has stated OECD Guideline 236 may be used within a weight of evidence approach to address the REACH Annex VIII, 9.1.3 endpoint of short-term toxicity on fish. According to ECHA (2016b), correlation of zebrafish embryo test results with acute fish toxicity of zebrafish, fathead minnow, rainbow trout and bluegill indicate an overall high concordance of zebrafish embryo and acute fish toxicity data. However, further analysis of the available data indicates considerable variability between the measured fish embryo toxicity (FET) LC50 values and the measured acute fish toxicity (AFT) LC50 values. The data assessed by ECHA (2016b) was for zebrafish embryo test results and this dossier presents data for fathead minnow embryos; however, a similar trend is anticipated. To support the weight of evidence approach, the experimental fathead minnow FET results are considered alongside nine separate QSAR predictions.

Table 1 summaries the QSAR models’ predictions and identifies three general groupings, (1) those that considered the alkyl fluoride portion of the compound, (2) those that considered the methacrylate portion of the compound and (3) the Danish model where the applicable training sets and structural fragment domain were not publicly available. None of the individual models address all structural fragments of OFPMA nor do all the models include short-chain polyfluorinated alkyl methacrylates in their training sets; however, when considered as a whole, the combined predictions do estimate the toxicity of the whole OFPMA structure. Therefore, while the reliability of individual predictions is poor, as a whole, the overall reliability of the predictions is sufficient to support the FET results to address regulatory acute fish toxicity concerns.

Table 1: Summary of QSAR results
Model	96-hour LC50 prediction (mg/L)
Alkyl fluoride compounds
QSAR Toolbox 4.1 Trend analysis, executed via AW "Ecotoxicological Endpoint"	0.381 (0.0122 to 11.9)
US EPA Toxicity Estimation Software Tool (T.E.S.T.) Version 4.2 Predicted Fathead minnow LC50 (96 hr) for 355-93-1 (Consensus method)	3.59
ECOSAR (US EPA) Version 1.11 Fish, 96-hr, LC50 (Class: Neutral Organic SAR)	29.30
Methacrylate compounds
VEGA 1.1.4 Fish Acute (LC50) Toxicity classification (SarPy/IRFMN) 1.0.2	1-10
VEGA 1.1.4 Fish Acute (LC50) Toxicity model (KNN/Read-Across) 1.0.0	8.28
VEGA 1.1.4 Fathead Minnow LC50 96h (EPA) Version 1.0.7	9.98
ECOSAR (US EPA) Version 1.11 Fish, 96-hr, LC50 (Class: Methacrylates)	10.11
VEGA 1.1.4 Fish Acute (LC50) Toxicity model (NIC) 1.0.0	61.25
Training sets and structural fragment domain unknown
Danish QSAR Database (DTU) Fathead minnow 96h LC50 (mg/L) Battery prediction for Leadscope and SciQSAR	43.4

 

The 96-hour acute fish toxicity QSAR predictions span a range of LC50 0.4 to 61.3 mg/L. This range spans several regulatory thresholds, therefore careful assessment and analysis of the reliability of these values is required. The most conservative 96-hour LC50 prediction is the QSAR Toolbox 4.1 prediction of 0.381 mg/L with a 95% confidence interval which spans from 0.0122 to 11.9 mg/L. The upper range of this prediction agrees with the remaining eight QSAR predictions and the experimental FET value, specifically that the 96 hour LC50 value is greater than 1 mg/L.  As the predicted value from QSAR Toolbox 4.1 is quantitatively different from the remaining QSAR predictions and all of the predictions are poorly reliable, it is inappropriate to support the experimental FET value with this most conservative prediction and all QSAR values must be considered with equal weighting.

Based on a weight of evidence, the model predictions indicate the 96-hour acute fish toxicity will be greater than 1 mg/L. Statistical analysis of the eight endpoint-specific QSAR predictions (i.e., prediction of a specific value rather than a range as provided by the SarPy/IRFMN prediction) and the FET TWM measured 96-hour LC50 value indicate that the results are highly skewed towards the lower values with a Skewness of 1.3 and the distribution has a Kurtosis of 0.5 indicating most of the data is in the tails of the distribution and very little data surrounds the mathematical mean of the values.  As the data is highly skewed towards the lower values and very little data surrounds the mathematical mean, it is appropriate to consider the median LC50 value of 10 mg/L as a representative acute fish toxicity value for this weight of evidence approach. This median LC50 value of 10 mg/L agrees with the SarPy/IRFMN toxicity classification predicted range of 1-10 mg/L and is within the 95% confidence interval of the most conservative QSAR Toolbox 4.1 prediction.

The 96-hour acute fish toxicity value based on a weight of evidence approach which considered the measured value from a fish embryo toxicity study and the predicted values of 9 QSAR models is 10 mg/L. While this value is at a UN GSH (revision 7.0) regulatory threshold for classification, further investigation is not required as this value supports the hypothesis that fish are not the most sensitive species. Based on the experimentally measured 72-hour algae EC50 value of 4.4 mg/L, reported elsewhere within this dossier, algae are the most sensitive aquatic species and further aquatic vertebrate testing of OFPMA is not necessary to address the regulatory requirement.

References:

ECHA (2016a). Aquatic: Update Of The Test Guidelines. Annankatu 18, P.O. Box 400, FI-00121 Helsinki, Finland. Available at: https://echa.europa.eu/documents/10162/21650280/oecd_test_guidelines_aquatic_en.pdf/2548af92-ffe1-4e38-a42a-463103b1586f

ECHA (2016b). Analysis of the relevance and adequateness of using Fish Embryo Acute Toxicity test (FET) Test Guideline (OECD 236) to fulfil the information requirements and addressing concerns under REACH. Report ECHA-UFZ contract ECHA/2014/341. Available at: https://echa.europa.eu/documents/10162/13639/fet_report_en.pdf