Palm oil, epoxidized

Developmental toxicity / teratogenicity

Administrative data

Endpoint:

developmental toxicity

Type of information:

migrated information: read-across from supporting substance (structural analogue or surrogate)

Adequacy of study:

key study

Study period:

28th October 1992 to 18th November 1992

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

other: see 'Remark'

Remarks:

Current test guidelines indicates that the test material should be administered up to the day prior to the scheduled Caesarean section. In this study the rats were sacrificed on Day 20 and the fetuses were removed by Caesarean section. The test material was administered daily from day 6 to day 15 and therefore not up to the day prior to the Caesarean section. This read-across is based on the hypothesis that source and target substances have similar toxicological properties because of their structural similarities and they are assumed to have similar toxicokinetic profiles i.e. they are expected to be metabolised in a similar fashion. The target substance Epoxidized Palm Oil (EPO) and source substance Epoxidized Soybean Oil (ESBO) are derived respectively from their raw materials, palm oil and soybean oil, and are of variable composition consisting of fatty acid triglycerides. Both EPO and ESBO are organic UVCB sub-type 1: substances of biological nature that have been modified in chemical processing. Both are manufactured by the reaction of the respective oils with an epoxidizing agent (50-60% hydrogen peroxide at 60-75°C). The olefinic bonds of the oils are converted to epoxy oxirane groups. As Palm oil has lower unsaturated bonds than Soybean Oil, less of the epoxidizing agent hydrogen peroxide is required and thus EPO has a lower epoxidized adduct content than ESBO and is therefore expected to be less chemically reactive. The % epoxidation in ESBO is 6-8% while the % epoxidation in EPO is 2.5-3.5%. The target substance (EPO) and source substance (ESBO) have a structurally similar backbone which is an epoxidized triglyceride structure derived from one glycerol molecule and three fatty acid molecules. Therefore, the source and the target substances share structural similarities with common functional groups and side chains varying in their length and the amount of epoxide groups. The target substance contains 8 fatty acids with the largest components being C16:0 palmitic acid (44%), C18:1 oleic acid (39.2%) and C18:2 linoleic acid (10.1%). The source substance contains 5 fatty acids with the largest components being C16:0 palmitic acid (11.3%), C18:1 oleic acid (39.2%) and C18:2 linoleic acid (55.8%). ESBO does not contain lauric (C12), myristic (C14) and arachidic (C20) acids while they are present in very low amounts in EPO (0.2, 1.1 and 0.3% respectively). Stearic acid (C18:0) is present in both substances at similar levels (4.5% in EPO and 3.4% in ESBO) while α-linolenic acid (C18:3) is present at 0.4% in EPO and 6.4% in ESBO. So, the main component of the triglyceride structure of both EPO and ESBO is C16 (44%; 11.3%) and C18 (54.2%; 88.7%). The data gap for the target substance EPO is a pre-natal developmental toxicity study (Annex IX, 8.7.2). No reliable data on the pre-natal developmental toxicity study of EPO is available. Therefore, read-across from an existing pre-natal developmental toxicity study of the source substance, ESBO, is considered as an appropriate adaptation to the standard information requirements of Annex IX, 8.7.2 of the REACH Regulation for the target substance, in accordance with the provisions of Annex XI, 1.5 of the REACH Regulation.

Cross-reference

Reason / purpose for cross-reference:

reference to other study

Data source

Reference

Reference Type:

study report

Title:

Unnamed

Year:

1993

Report date:

1993

Materials and methods

Test guideline

Test guideline 1

Qualifier:

according to guideline

Guideline:

OECD Guideline 414 (Prenatal Developmental Toxicity Study)

Version / remarks:

12th May 1981

Test guideline 2

Qualifier:

according to guideline

Guideline:

other: E.E.C. Recommendation No. 87/302/E.E.C., 18th November 1987

GLP compliance:

yes

Limit test:

no

Test material

Test material information

Constituent 1

Reference substance name:

Epoxidised Soybean Oil

IUPAC Name:

Epoxidised Soybean Oil

Constituent 2

Reference substance name:

Soybean oil, epoxidized

EC Number:

232-391-0

EC Name:

Soybean oil, epoxidized

Cas Number:

8013-07-8

IUPAC Name:

8013-07-8

Details on test material:

- Name of test material (as cited in study report): Epoxidised Soybean Oil (ESBO)
- Alias name: Reoplast 39
- Physical state: clear yellowish liquid
- Lot/batch No.: 08380306
- Expiration date of the lot/batch: July 1994
- Stability under test conditions: stable
- Storage condition of test material: Room temperature

Test animals

Species:

rat

Strain:

Sprague-Dawley

Details on test animals or test system and environmental conditions:

TEST ANIMALS
Sprague-Dawley Rats: CRL CD (SD) BR
- Source: Charles River, 76410 Saint-Aubain-lès-Elbeuf, France
- Age at study initiation: At the beginning of the treatment period, the females were 9 weeks
- Weight at study initiation: approximately 200 g
- Housing: The rats were housed individually in polycarbonate U.A.R. cages (48.0 x 27.0x 20.0 cm). Each cage contained autoclaved sawdust and was equipped with a water bottle. Bacteriological analyses of the sawdust and detection of possible contaminants were periodically made. Bottles and cages were changed at least once a week.
- Diet (e.g. ad libitum): U.A.R. A04 C pelleted diet, batch No. 20721
- Water (e.g. ad libitum): drinking water filtered using 0.22 micron filter
- Acclimation period: 5 days before the beginning of the treatment period

ENVIRONMENTAL CONDITIONS
- Temperature (°C): 21 ± 2 °C
- Humidity (%): 50 ± 20 %
- Air changes (per hr): about 13 cycles/hour of filtered, non-recylced air
- Photoperiod (hrs dark / hrs light): 12 hours/12 hours

IN-LIFE DATES: From: 28/10/1992 To: 13/11/1992

Administration / exposure

Route of administration:

oral: gavage

Vehicle:

other: Soybean Oil

Details on exposure:

PREPARATION OF DOSING SOLUTIONS:
The test substance was diluted in the vehicle and homogenised using a magnetic stirrer.
The preparations were made by the C.I.T. Pharmacy for a maximum of 7 days of use according to the stability already demonstrated at the concentrations used in the study.

DIET PREPARATION
The animals had free access to U.A.R. A04 C pelleted diet, batch No. 20721 (U.A.R., 91360, Villemoisson-sur-Orge, France) and water (drinking water filtered using 0.22 micron filter). Each batch of food was analysed by the supplier.

Bacteriological and chemical analyses of the water and detection of possible contaiminants are made periodically. There were no known contaminants in the diet, water or sawdust at levels likely to influence the outcome of the study.

VEHICLE
- Justification for use and choice of vehicle (if other than water): Soybean Oil
- Lot/batch no. (if required): 08380327

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

Chemical Analysis of the Preparations:
On the first day of treatment of the first mated females and the last day of treatment of the last mated females, each preparation (control group included) was checked for achieved concentration of the test substance. Each preparation was sampled in duplicate for analysis.

Assay Method:

The solution was agitated with a magnetic stirrer. An intended volume of solution (15 mL for the 0 and 100 mg/kg bw/day groups, 5 mL for the 300 mg/kg bw/day group, 1.5 mL for the 1000 mg/kg bw/day group) was sampled in a flask using a calibrated glass pipette and diluted with 100 mL of CETAB (N-cetyl-N, N, N-trimethyl-ammonium bromide) before adding 4 drops of cristal violet indicator. The solution (agitated with a magnetic stirrer) was titrated by adding of perchloric acid solution (0.1 M perchloric acid in glacial acetic acid) with a volumetric distribution system (Dosimat 655, Methrom) until the colour of the indicator changed from violet to yellow.

Details on mating procedure:

The female rats were mated at the supplier's facilities. The day of mating was designated as day 0 of pregnancy.

Duration of treatment / exposure:

Dose levels were selected following the results of a previously conducted study (see other two studies presented in IUCLID Point 7.8.2). 1000 mg/kg bw/day was selected as the high dose level for this study; 100 and 300 mg/kg bw/day were selected as the low and intermediate dose levels respectively. Exposure was from day 6 to day 15 inclusive of pregnancy.

Frequency of treatment:

The test substance or the vehicle were given daily, at the same approximate daily time, 7 days a week from day 6 of pregnancy to day 15 inclusive.

Duration of test:

To day 20 of pregnancy

Doses / concentrations

Remarks:

Doses / Concentrations:
0, 100, 300, and 1000 mg/kg bw/day
Basis:
nominal conc.

No. of animals per sex per dose:

There were 4 groups of rats, 1 group per dose concentration, and there 25 female rats per concentration tested

Control animals:

yes, concurrent vehicle

Details on study design:

- Dose selection rationale: Dose levels were selected on basis of the results of a previously conducted study (CIT/Study No. 8707 RSR) performed at the dose levels of 150, 450 and 1000 mg/kg bw/day. The results showed that neither toxic effects, nor effects on reproduction and litters were noted at any of these dose levels. Therefore the 1000 mg/kg bw/day was selected as the high dose level for this study; 100 and 300 mg/kg bw/day were selected as the low and intermediate dose levels respectively. There was one group per dose concentration, with 25 animals per group, 0, 100, 300 and 1000 mg/kg bw/day.

Examinations

Maternal examinations:

CAGE SIDE OBSERVATIONS: Yes - animal were checked for mortaility or signs of morbidity
- Time schedule: twice a day

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: At least once daily

BODY WEIGHT: Yes
- Time schedule for examinations: on days 2, 6, 9, 12, 15 and 20 of pregnancy

FOOD CONSUMPTION : Yes
- Food consumption for each animal was determined at the following intervals: day 2 -day 6, day 6 - day 9, day 9 - day 12, day 12 - day 15 and day 15 - day 20

WATER CONSUMPTION AND COMPOUND INTAKE (if drinking water study): No data

POST-MORTEM EXAMINATIONS: Yes
- Sacrifice on gestation day # 20
- Organs examined: heart, lung, liver, kidneys, stomach, intestines

Ovaries and uterine content:

The ovaries and uterine content was examined after termination: Yes
Examinations included:
- Number of corpora lutea
- Number and distribution of live and dead foetuses
- Number and distribution of early and late resorptions
- Number of implantation sites

Fetal examinations:

Examination of Fetuses:
- Body weight - each live foetus was weighed
- External Examination: each foetus was submitted to an external examination (including palate) to check for the presence of malformations
- Soft Tissue Examination: the soft tissue findings were classified into malformations and anomalies.
- Skeletal Examination: The skeletal findings were classified into skeletal variations, anomalies and malformations
- Sex of Fetuses: The sex of foetuses was determined at the time of the evisceration after fixation in alcohol or at the time of Wilson's sections.

Statistics:

The mean values were compared by one-way analysis of variances and Dunnett's test. Percentage values were compared by Fisher's exact probability test.

Indices:

No data supplied

Historical control data:

C.I.T historical control data:
Fetal Soft Tissue Anomalies: Dilated renal pelvis, mean: 0.9 %, range of means: 0. - 2.8 %; ureteral dilatation: mean: 1.4 %, range of means: 0 to 6.1 % .

Fetal Skeletal Variations:
Reduced Ossification of the 6th Sternebra
Fetal incidenc mean: 30.3 % - range of means: 16.0 % to 39.1 %, litter incidence: mean: 75.7 % - range of means: 57.1 % to 100.0 %.

Unossification of the 5th Sternebra
Fetal incidence: mean: 15.1 % - range of means: 1.7 % to 31.8 %; litter incidence: mean: 47.3 % - range of means: 9.5 % to 100 %.

Fetal Skeletal Anomalies:
mean: 0.7 % - range of means: 0.0 % to 3.2 %

Results and discussion

Results: maternal animals

Maternal developmental toxicity

Details on maternal toxic effects:

Maternal toxic effects:no effects

Details on maternal toxic effects:
No treatment-related macroscopic changes were observed at necropsy of the females. No deaths occurred in the females of any group. No abortions were noted. The mean food consumption for the females with completed pregnancy was similar in the control and treated groups. The mean body weight gain of the females with completed pregnancy was similar in the control and treated groups.

Effect levels (maternal animals)

1

Dose descriptor:

NOAEL

Effect level:

1 000 mg/kg bw/day

Based on:

test mat.

Basis for effect level:

other: other:

2

Dose descriptor:

NOEL

Effect level:

1 000 mg/kg bw/day

Based on:

test mat.

Basis for effect level:

other: other:

Results (fetuses)

Details on embryotoxic / teratogenic effects:

Embryotoxic / teratogenic effects:yes

Details on embryotoxic / teratogenic effects:
LITTER DATA
Corpora lutea, implantation sites and pre-implantation loss
The mean number of corpora lutea and implantation sites were similar in the control and treated groups.
The pre-implantation loss was higher in each treated group when compared to the control group. (23 %; p < 0.05 in the 100 mg/kgday group - 20.4 % N.S. in the 300 mg/kg/day - 24.4 %; p < 0.01 in the 1000 mg/kg/day group vs. 16.6 % in the control group). As the treatment of the females began after the implantation of the ova, the increase of the pre-implantation loss is considered not to have a toxicological significance.

Total Post-Implantation Loss:
- Resorptions: The rate of resorptions was similar in the control (2.6 % ) and the 100 (2.7 %), 300 (1.6 %) and 1000 (1.1%) mg/kg bw/day groups.
- Dead Fetuses: No dead foetuses were noted in the control, 100 and 300 mg/kg bw/day groups. In the 1000 mg/kg bw/day group, 1 out of 269 foetuses died. This very low incidence of dead foetuses (0.4 %) was considered to be of no toxicological significance.
- Post-implantation Loss: The post-implantation loss was similar in the control and treated groups.

Live Fetuses:
- Mean Number: The mean number of live foetuses was similar in the control and treated groups.
- Body Weight: The mean fetal body weight was similar in the control and treated groups.
- Sex-ratio: No treatment-related effects were noted on the sex-ratio.

FETAL OBSERVATIONS
Fetal External Observation: No external malformations were observed in the fetuses of the control and treated groups.
Fetal Soft Tissue Observations:
- Anomalies: No soft tissue anomlies were noted in the fetuses of the control and 100 mg/kg bw/day groups.
In the 300 mg/kg bw/day group, 3 out of 122 fetuses (2.5 % ) had dilated renal pelvis associated for two of them (1.6 % ) with ureteral dilatation.
In the 1000 mg/kg bw/day group, 3 out of 130 fetuses (2.3 % ) had dilated renal pelvis associated for one of them (0.8 % ) with ureteral dilatation.
These two findings which are within the range of C.I.T. Historical control data are considered to be incidental.

- Malformations: No soft tissue malformations were noted in the fetuses of the control, 100 and 1000 mg/kg bw/day groups.
In the 300 mg/kg bw/day group, 1 out of 122 fetuses had a cerebral venticular dilatation. This malformation which was noted only in one fetus and which was not observed at a higher dose level is considered as incidental.

Fetal Skeletal Observations:
Fetal Skeletal Variations:
- The fetal and litter incidences of reduced ossification of the 6th sternebra were significantly different in the 1000 mg/kg bw/day group from those of the control group. As these incidences are lower than those of the control group and within the range of C.I.T. control historical data, they are considered to have no toxicological significance.
In the same way, the fetal incidence and the litter incidence of unossification of the 5th sternebra were significantly different from those of the control group in the 1000 mg/kg/day group but lower and within the range of C.I.T. historical data. Therefore they are considered not to have a toxicological significance. In the 300 mg/kg/day group, the fetal incidence was not significantly different from that of the control group, but the litter incidence was lower and within the range of the CIT historical data. The fetal incidence of unossification of the 4th metacarpal was significantly higher in the 1000 mg/kg/daya group., when compared to the control group. This incidence is very low and below the range of CIT control historical data and therefore is not considered related to treatment.
No other dose-related effects were noted on the incidence of the skeletal variations.

Fetal Skeletal Anomalies:
The fetal incidence of reduced ossification of thoracic vertebrae was higher in the 300 and 1000 mg/kg/day groups. As the differences from the control groups are very slight and within the range of CIT control historical data, a treatment-related effect is ruled out. The incidence of the few other skeletal anolmaies was similar in the control and treated groups.

Fetal Skeletal Malformations:
No skeletal malformations were observed in fetuses of any group.

Fetal abnormalities

Abnormalities:

not specified

Overall developmental toxicity

Developmental effects observed:

not specified

Any other information on results incl. tables

Throughout the study, a satisfactory concordance between obtained and nominal concentrations was found for the administered preparation.

Table 2       The pregnancy status is summarised in the table below:

Dose (mg/kg bw/day)	0	100	300	1000
Mated Females	25	25	25	25
Non-Pregnant Females	2	1	7	5
Pregnant Females	23	24	18	20
dead/sacrificed	0	0	0	0
aborted	0	0	0	0
total resorption	0	0	0	0
completed pregnancy	23	24	18	20

Read-Across Justification: Full report is attached in study summary

3 Analogue approach justification

3.1. Physicochemical properties

Physicochemical data shows that the physicochemical properties of the target and source substances are similar as outlined in the data matrix (Table 3). Both are liquids and the structural differences in the side chains do not significantly influence the physicochemical properties of both substances, i.e. vapour pressure, water solubility and partition co-efficient (log Pow). Both substances are highly insoluble in water; <0.01 mg/L at 30°C for EPO and <0.02 µg/L at 20°C (calculated) for ESBO. Neither of the substances is volatile, with a vapour pressure of 0.5 kPa at 25°C for EPO and 8.4 x 10-8 Pa at 25°C for ESBO. Both substances are highly lipophilic; mean Log Pow >6.2 at 25°C for ESBO and log Pow >10 (calculated) for EPO.

3.2. Toxicokinetics

No specific experimental data on absorption, distribution, metabolism or excretion is available for the source or target substance. Read-across was performed for all human health toxicity endpoints to Epoxidised Soybean Oil (ESBO, CAS No. 8013-07-8). An OECD SIDS report is available that concluded on a proposed metabolic pathway for epoxidised fatty acid esters, including ESBO (OECD, 2006). The toxicokinetic analysis is based on physicochemical data from EPO, read-across ESBO data from in vivo animal models and the OECD SIDS report in the literature (OECD, 2006).

Physicochemical data

The molecular weight of EPO and ESBO is > 500 g/mol and is not in the range for favourable oral absorption (<500 g/mol). The calculated log Pow of EPO (>10) and mean Log Pow >6.2 at 25°C for ESBO indicate they are highly lipophilic and water solubility (<0.01 mg/L at 30°C) for EPO and <0.02 µg/L at 20°C (calculated) for ESBO indicates they are both insoluble in water. These characteristics will not facilitate transport of EPO or ESBO via passive diffusion. Based on its high lipophilicity, absorption of EPO and ESBO via the lymphatic system through micellular solubilisation by bile salts is likely, similar to other vegetable oils. Insolubility in water of both EPO and ESBO indicates low dermal uptake while the high log Pow values for both are an indication for a high uptake into the stratum corneum but little or no penetration into the lower layers of the epidermis and dermis. Overall, the physical state, molecular weight, calculated log Pow and water insolubility indicate that dermal absorption of EPO and ESBO is unlikely. Due to the low vapour pressure of EPO (0.5 kPa at 25°C) and ESBO (8.4 x 10-8 Pa at 25°C) and physical state (liquid), exposure via the inhalation route of both is expected to be negligible. Based on the information available for the analogue ESBO in the OECD SIDS report (OECD, 2006; see ‘Other data in the literature’), during metabolism, breakdown products are produced that are more water soluble than the parent substance i.e. free fatty acids, so it is expected that any EPO metabolites will be excreted in the urine.

Other data in the literature

The OECD produced a report on Epoxidised Oils and Derivatives in 2006 (OECD, 2006), which included ESBO. The OECD SIDS concluded that epoxidised fatty acid esters, such as ESBO and therefore we assume EPO, produce metabolic products with similar primary constituents as other vegetable oils and are assumed to have similar metabolic pathways e.g. breakdown in the gastrointestinal tract by esterases (pancreatic lipase) to epoxidised fatty acids and glycerol which enter the normal nutritional pools (JECFA, 1974). Pancreatic lipase works at the oil/water interface since triglycerides are insoluble. During metabolism in the GI tract, pancreatic lipase preferentially hydrolyses triglycerides to release the free fatty acids from the SN-1 and SN-3 (terminal) positions of the glycerol backbone. The other products of metabolism are mono- and di-glycerides (OECD, 2006). The EFSA Panel on Contaminants in the Food Chain agreed with this assessment for ESBO in 2011 (EFSA, 2011). Overall, the proposed metabolic pathway for ESBO is enzymatic breakdown to epoxidised fatty acids and glycerol; a similar pathway is predicted for EPO. Based on the information available for the analogue ESBO in the OECD SIDS report, during metabolism, breakdown products are produced that are more water soluble than the parent substance i.e. free fatty acids, so it is expected that any EPO metabolites will be excreted in the urine.

Available in vivo toxicological data

The in vivo read-across data from ESBO indicate no adverse effects if oral absorption occurs (acute oral LD50 of >5,000 mg/kg (3), 2 year combined chronic/carcinogenicity toxicity study

NOEL (male) of 1000 mg/kg bw/day and NOEL (female) of 1400 mg/kg bw/day (13), pre-natal developmental toxicity maternal/developmental NOAEL of 1000 mg/kg bw/day (14). The in vivo read-across data from ESBO indicates is poorly absorbed via the dermal route (slightly irritating in the in vivo skin irritation study in rabbits (6) and non-sensitising in Guinea pig maximization test (8)). Any significant dermal absorption is unlikely.

3.3. Comparison of data from human health endpoints

3.3.1 Toxicity data of the target and source substances

There is no existing human health toxicity data for the target substance, EPO. As is presented in the data matrix (Table 3), the acute oral (LD 50 (male/female) >5,000 mg/kg bw) and acute dermal toxicity data (LD50 >20mL/kg bw) shows very low toxicity for the source chemical, ESBO, in rats and rabbits. The source chemical is slightly irritating to skin and eye in rabbits. The source substance is not a skin sensitizer in the guinea pig maximization test. In the in vitro bacterial reverse mutation study (Ames test), in vitro chromosomal aberration study and in vitro gene mutation study in mammalian cells, the source substance ESBO was negative in the presence and absence of metabolic activation. The source substance ESBO is not genotoxic. In a combined chronic toxicity/carcinogenicity study in rats for 104 weeks, a NOEL value of 1000 mg/kg bw/day (male) and 1400 mg/kg bw/day (female) was derived. In a pre-natal developmental toxicity study in rats, the NOAEL (maternotoxic, embryofetal) was 1000 mg/kg bw/day with no adverse effects noted. In accordance with Column 2 of ANNEX IX of the REACH Regulation, a two generation reproductive toxicity study does not need not to be conducted as the existing read across combined chronic toxicity/carcinogenicity study from ESBO is available and does not indicate clear adverse effects on reproductive organs or tissues.

The data gap for the target substance EPO is a pre-natal developmental toxicity study (Annex IX, 8.7.2). No reliable data on the pre-natal developmental toxicity study of EPO is available. Therefore, read-across from an existing pre-natal developmental toxicity study of the source substance, ESBO, is considered as an appropriate adaptation to the standard information requirements of Annex IX, 8.7.2 of the REACH Regulation for the target substance, in accordance with the provisions of Annex XI, 1.5 of the REACH Regulation.

The read-across pre-natal developmental toxicity study (RL2) was conducted according to OECD 414 and GLP. In this study, ESBO was administered to 4 groups of Sprague Dawley rats (25 females/group) by gavage at dose levels of 0, 100, 300 or 1000 mg/kg bw/day from days 6 through 15 of gestation. No treatment-related macroscopic changes were observed at necropsy of the females. No deaths occurred in the females of any group. No abortions were noted. The mean food consumption for the females with completed pregnancy was similar in the control and treated groups. The mean body weight gain of the females with completed pregnancy was similar in the control and treated groups. The maternotoxic NOAEL is 1000 mg/kg bw/day. A maternotoxic NOAEL of 1000 mg/kg bw/day is also predicted for EPO.

No dead foetuses were noted in the control, 100 and 300 mg/kg bw/day groups. In the 1000 mg/kg bw/day group, 1 out of 269 foetuses died. This very low incidence of dead foetuses (0.4 %) was considered to be of no toxicological significance. The mean number of live foetuses was similar in the control and treated groups. The mean fetal body weight was similar in the control and treated groups. No treatment-related effects were noted on the sex-ratio. The mean number of corpora lutea and implantation sites was similar in the control and treated groups. The pre-implantation loss was higher in each treated group when compared to the control group. (23 %; p < 0.05 in the 100 mg/kg bw/day group - 20.4 % N.S. in the 300 mg/kg bw/day - 24.4 %; p < 0.01 in the 1000 mg/kg bw/day group vs. 16.6 % in the control group). As the treatment of the females began after the implantation of the ova, the increase of the pre-implantation loss is considered not to have a toxicological significance. The rate of resorptions was similar in the control (2.6 %) and the 100 (2.7 %), 300 (1.6 %) and 1000 (1.1%) mg/kg bw/day groups. The post-implantation loss was similar in the control and treated groups.

No external malformations were observed in the fetuses of the control and treated groups. No soft tissue anomalies were noted in the fetuses of the control and 100 mg/kg bw/day groups. In the 300 mg/kg bw/day group, 3 out of 122 fetuses (2.5 % ) had dilated renal pelvis associated for two of them (1.6 % ) with ureteral dilatation. In the 1000 mg/kg bw/day group, 3 out of 130 fetuses (2.3 % ) had dilated renal pelvis associated for one of them (0.8 % ) with ureteral dilatation. These two findings which are within the range of the laboratory (C.I.T.) Historical control data are considered to be incidental. No soft tissue malformations were noted in the fetuses of the control, 100 and 1000 mg/kg bw/day groups. In the 300 mg/kg bw/day group, 1 out of 122 fetuses had a cerebral venticular dilatation. This malformation which was noted only in one fetus and which was not observed at a higher dose level is considered as incidental. The fetal and litter incidences of reduced ossification of the 6th sternebra were significantly different in the 1000 mg/kg bw/day group from those of the control group. As these incidences are lower than those of the control group and within the range of the laboratory (C.I.T.) control historical data, they are considered to have no toxicological significance. In the same way, the fetal incidence and the litter incidence of unossification of the 5th sternebra were significantly different from those of the control group in the 1000 mg/kg/day group but lower and within the range of the laboratory (C.I.T.) historical data. Therefore they are considered not to have a toxicological significance. In the 300 mg/kg/day group, the fetal incidence was not significantly different from that of the control group, but the litter incidence was lower and within the range of the laboratory (C.I.T.) historical data. The fetal incidence of unossification of the 4th metacarpal was significantly higher in the 1000 mg/kg bw/day group, when compared to the control group. These incidences are very low and below the range of the laboratory (C.I.T.) control historical data and therefore are not considered related to treatment. No other dose-related effects were noted on the incidence of the skeletal variations. The fetal incidence of reduced ossification of thoracic vertebrae was higher in the 300 and 1000 mg/kg/day groups. As the differences from the control groups are very slight and within the range of the laboratory (C.I.T.) control historical data, a treatment-related effect is ruled out. The incidence of the few other skeletal anomalies was similar in the control and treated groups. No skeletal malformations were observed in fetuses of any group. The embyrofetal NOAEL is 1000 mg/kg bw/day. An embyrofetal NOAEL of 1000 mg/kg bw/day is also predicted for EPO.

The dose descriptor obtained from the existing pre-natal developmental toxicity study performed on the source substance is considered as an appropriate starting point for deriving a DNEL for EPO. The remaining uncertainty associated with this read-across approach is accounted for by using the appropriate assessment factors. A qualitative uncertainty evaluation is completed on the sources of uncertainty in the following table.

Sources of uncertainty		Variability or uncertainty	Direction & Magnitude
Hazard assessment: Developmental toxicity	Overall database/Selection of data	UNC	++
	Extrapolation uncertainties (read across for similar substances)	UNC	+
	Adequacy of assessment factors associated with uncertainty	UNC	+
Overall effect on hazard estimate: The default assessment factor ‘quality of the whole database’ used in the hazard assessment of 1 is increased to 2 to account for uncertainty.

Key: +, ++, +++ = low, moderate and high overestimates; -, --, --- = low, moderate and high underestimates; VAR= variability, UNC= uncertainty

The predicted NOAEL (maternotoxic, embryofetal) value for EPO is 1000 mg/kg bw/day based on read-across from the ESBO study. The default assessment factor ‘quality of the whole database’ used in the hazard assessment of 1 is increased to 2 to account for uncertainty, based on the data provided in the available developmental study. This modification should be sufficient to account for any uncertainty in the read across approach.

3.3.2 Effect of structural differences between target and source chemical

The target substance consists of 8 fatty acids with the largest components being C16:0 palmitic acid (44%), C18:1 oleic acid (39.2%) and C18:2 linoleic acid (10.1%). The source substance consists of 5 fatty acids with the largest components being C16:0 palmitic acid (11.3%), C18:1 oleic acid (39.2%) and C18:2 linoleic acid (55.8%). ESBO does not contain lauric, myristic and arachidic acids while they are present in very low amounts in EPO (0.2, 1.1 and 0.3% respectively). Stearic acid is present in both substances at similar levels (4.5% in EPO and 3.4% in ESBO) while α-linolenic acid is present at 0.4% in EPO and 6.4% in ESBO. So, the main component of the triglyceride structure of both EPO and ESBO is C16 (44%; 11.3%) and C18 (54.2%; 88.7%). When these epoxidized fatty acid products are released during metabolism (see Section 3.2) they are not expected to have any adverse effects in the body as all are naturally occurring fatty acids and they are all ‘Not Classified’ in the ECHA Classification and Labelling (C&L) inventory (checked 24-08-15). The only exception is α-linolenic acid which is indicated as a skin sensitizer but this is not relevant to ESBO and EPO as the former is negative in the guinea pig maximization test and EPO is also predicted to be negative also. The main structural difference is that palm oil has lower unsaturated bonds than soybean oil therefore less of the epoxidizing agent hydrogen peroxide is required to produce the epoxidized derivatives. EPO has a lower epoxidized adduct content than ESBO and is therefore expected to be less chemically reactive. The % epoxidation in ESBO is 6-8% while the % epoxidation in EPO is 2.5-3.5%.

3.3.3 Classification and labelling

According to the ECHA Classification and Labelling (C&L) inventory, the source substance, ESBO, is ‘Not Classified’ (647 notifiers, joint entry; checked 24-08-15). The target substance, EPO, is not listed in the C&L inventory (checked 24-08-15). In accordance with Column 2 of ANNEX IX of the REACH Regulation, a two generation reproductive toxicity study does not need not to be conducted as the existing read across combined chronic toxicity/carcinogenicity study from ESBO is available and does not indicate clear adverse effects on reproductive organs or tissues. Based on the read-across oral pre-natal developmental toxicity presented for the source substance, EPO does not need to be classified for reproductive toxicity when the criteria outlined in Annex I of 1272/2008/EC are applied.

4. Conclusion

The structural similarities between the source and the target substances and estimated similar toxicokinetics presented above support the read-across hypothesis. The structural differences between the target and source substance are not expected to have an impact on the prediction. Adequate, reliable and available scientific information indicates that using the source substance for read across to the target substance is acceptable.

Therefore, based on the considerations above, it can be concluded that the oral pre-natal developmental toxicity conducted in rats with ESBO is likely to predict the oral pre-natal developmental toxicity of EPO and is considered as adequate to fulfill the information requirement of Annex IX, 8.7.2.

Applicant's summary and conclusion

Conclusions:

The test substance Epoxidised Soybean Oil (ESBO) when administered daily by oral gavage to pregnant female Sprague-Dawley rats during organogenesis at the dose levels of 100, 300 and 1000 mg/kg bw/day was well tolerated by the dams at all the dose levels and was neither embryotoxic or teratogenic.

The No Observable Adverse Effect Level (NOAEL) is defined as 1000 mg/kg bw/day in terms of maternotoxic effects and embryofetal development.

The No Observable Effect Level (NOEL) is also defined as 1000 mg/kg bw/day in terms of maternotoxic effects and embryofetal development.

Executive summary:

The objective of the study was to evaluate the potential toxic effects of the test substance Epoxidised Soybean Oil (ESBO) on embryonic and fetal development when administered by oral route (gavage) to pregnant female Sprague-Dawley rats during organogenesis (day 6 to day 15 of pregnancy inclusive).

The test substance Epoxidised Soybean Oil (ESBO) when administered daily by oral gavage to pregnant female Sprague-Dawley rats during organogenesis at the dose levels of 100, 300 and 1000 mg/kg bw/day was well tolerated by the dams at all the dose levels and was neither embryotoxic or teratogenic.

The No Observable Adverse Effect Level (NOAEL) is defined as 1000 mg/kg bw/day in terms of maternotoxic effects and embryofetal development.

The No Observable Effect Level (NOEL) is also defined as 1000 mg/kg bw/day in terms of maternotoxic effects and embryofetal development.