Titanium dioxide

Endpoint summary

Key value for chemical safety assessment

Effects on fertility

Description of key information

Based on the weight of evidence from the available long-term toxicity/carcinogenicity studies in rodents and the relevant information on the toxicokinetic behaviour in rats it is concluded that TiO2 does not present a reproductive toxicity hazard.

Within the scope of the re-evaluation of titanium dioxide (E171) as a food additive by the European Food Safety Authority (EFSA) adopted on the 28th June 2016, the conduct of a “multigeneration or extended-one generation reproduction toxicity study according to the current OECD guidelines” was recommended. This study was requested in order to establish a health-based guidance value (ADI) for the food additive titanium dioxide (EFSA Journal 2016;14(9):4545).

The experimental phase for an extended one generation reproductive toxicity study is currently ongoing (status June 2020). An interim report is already available, which is added as robust study summary. It is highlighted that the conclusions are still preliminary and that these will be finalised, once the final report becomes available.

Link to relevant study records

Reference

Endpoint:

extended one-generation reproductive toxicity – with F2 generation and both developmental neuro- and immunotoxicity (Cohorts 1A, 1B with extension, 2A, 2B, and 3)

Type of information:

experimental study

Adequacy of study:

key study

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 443 (Extended One-Generation Reproductive Toxicity Study)

Version / remarks:

adopted June 25, 2018

Deviations:

no

Principles of method if other than guideline:

A non-guideline 14-day palatability study was conducted prior to the Extended one generation reproductive toxicity (EOGRTS) to obtain information on the palatability of several treatment levels of Titanium dioxide E171-E when given to rats daily via the diet. 5 animals/sex/dose were treated with 0, 300 and 1000 mg Titanium dioxide E171-E/kg bw/day for 14 days. Clinical signs, mortality, body weight and food and drinking water consumption were recorded. At the end of the study all animals were inspected macroscopically. The results of this study are reported in the results and discussion of the parental generation.

GLP compliance:

yes (incl. QA statement)

Remarks:

signed 08.05.2017

Limit test:

no

Justification for study design:

SPECIFICATION OF STUDY DESIGN FOR EXTENDED ONE-GENERATION REPRODUCTION TOXICITY STUDY WITH JUSTIFICATIONS:

- Premating exposure duration for parental (P0) animals : extended pre-mating period of 10 weeks since several publications reported testicular toxicity (impairment of spermatogenesis) or effects on sperm integrity and function as well as effects on the oestrus cycle.
- Basis for dose level selection : based on available toxicity data and the results of a 14-day palatability study conducted prior to the main study (as reported in this robust study summary).
- Inclusion of extension of Cohort 1B : Cohort 1B animals were mated to establish the F2 generation. This was done to investigate publications reporting several reproductive and developmental effects induced by titanium dioxide.
- Termination time for F2 : based on available data all F2 pups were dissected on PND 4-8; the date of dissection of F2 pups would have been postponed to PND 21 to allow further examinations. However, no adverse effects were observed, thus all F2 pups were dissected as planned (on PND 4-8).
- Inclusion of developmental neurotoxicity Cohorts 2A and 2B : the full range of neurodevelopmental toxicity was evaluated to investigate publications reporting several neurotoxic effects induced by titanium dioxide.
- Inclusion of developmental immunotoxicity Cohort 3 : the full range of developmental immunotoxicity was evaluated to investigate publications reporting several immunotoxic effects induced by titanium dioxide.
- Route of administration : Ingestion via the food is the most relevant route of exposure for a food additive such as Titanium dioxide E171-E.


Determination of aberrant crypt foci (ACF): the EFSA panel recommended that in order to investigate the potential for neoplastic lesions (observations made in a publication of Bettini et al., 2017), biomarkers for putative preneoplastic lesions in the colon should be examined as additional parameters in the requested extended one-generation reproductive toxicity study as recommended by EFSA in 2016.

Species:

rat

Strain:

other: CD/Crl:CD(SD)

Remarks:

used for the 14-day palatability study and EOGRTS

Details on species / strain selection:

The rat is a commonly used rodent species for such studies and required by the guideline.

Sex:

male/female

Details on test animals or test system and environmental conditions:

14-day palatability study:
The same strain and supplier of rats as decribed below were used for the 14-day palatability study. All animals had a adapation period of 5 days and were housed under the same conditions (e.g. diet, water, housing conditions) as decribed below.

EOGRTS:

TEST ANIMALS
- Source: Charles River Laboratories, Research Models and Servives; Germany GmbH, Sandhofer Weg 7, 97633 Sulzfeld
- Health statsus: healthy virgin animals
- Age at study initiation: males and females: 71 days
- Weight at study initiation: males: 353.5 - 446.9 g; females: 198.8 - 285.4 g
- Housing: With the exception of the mating period, the males and females were kept singly in MAKROLON cages (type III plus) with a basal surface of approx. 39 cm x 23 cm and a height of approx. 18 cm. Granulated textured wood released for animal bedding (Granulat A2, J. Brandenburg, 49424 Goldenstedt/Arkeburg, Germany) was used as bedding material in the cages. The cages were changed and cleaned once a week.
- Diet: certified commercial diet (ssniff® R/Z V1320, ssniff Spezialdiäten GmbH, 59494 Soest, Germany; ad libitum (with the exception of the night before the day of blood withdrawal for laboratory examination)
- The titanium content in the diet, in the bedding material and any enrichment material (rat huts, wooden chewing sticks) were analysed under GLP prior to initiation of the study using a validated ICP-OES method.
- Water: tap water; ad libitum
- Acclimation period: 6 days of adaptation

ENVIRONMENTAL CONDITIONS
- Temperature: 22°C±3°C (maximum range)
- Humidity: 55%±15% (maximum range)
- Air changes: 15-20 per hr
- Photoperiod: The rooms were alternately lit (about 150 lux at approximately 1.5 m room height) and darkened in a 12 hours dark/12 hours light cycle.

Route of administration:

oral: feed

Details on exposure:

DIET PREPARATION

14-day palatability study:
The test item formulations were freshly prepared once a week. To maintain a constant dose level in relation to the animal's body weight, the concentration in the diet was adjusted based on the mean group food consumption per sex. The test item concentration was adjusted weekly using the food consumption values from the previous week. The appropriate amount of test item was weighed into a glass container. Some of the test item and diet was mixed with an impact mill to produce a premix. This process was repeated until the whole quantity of test item was distributed in the diet. Afterwards, the premix was added to the diet, mixed for 15 minutes and then transferred to a closable bucket. Each bucket was labelled according to group, sex, and dose.

EOGRTS
- Rate of preparation of diet (frequency): once a week
- Preparation: The appropriate amount of test item was weighed into a glass container. A portion of the test item and diet was mixed with an impact mill (Grindomix GM300, Retsch GmbH, Haan, Germany) to produce a premix. Then the premix was added to the diet, mixed with a mixer (Röhnradmischer; Typ ELTE 650; J. Engelsmann AG, Ludwigshafen, Germany) for 15 minutes and then transferred to a closable bucket. Each bucket was labelled with group number and dose. To maintain a constant dose level in relation to the animals’ body weight, the concentration of the test item in the diet was adjusted based on the mean group food consumption per sex. The concentration was adjusted weekly using the food consumption values from the previous week.

Details on mating procedure:

- M/F ratio per cage: 1/1, Sexually mature male and female rats were paired randomly monogamously, i.e. 1 male and 1 female animal were placed in one cage during the dark period. The female was placed with the same male until evidence of mating was observed or two weeks had elapsed.
- Length of cohabitation: two weeks
- Proof of pregnancy: The females were examined each morning for the presence of sperm. The day of conception (day 0 of gestation) was considered to be the day on which sperm was found.

Additional information:
Females without a positive mating sign were separated from its male partner after 2 weeks without further opportunity for mating. After a pseudo gestation period of approx. 24 test days these females were laparotomized and their non-pregnancy status was confirmed by Salewski Staining.

For the establishment of the F2 Generation, males and females of the same dose group were paired (1:1 pairing), sibling mating was avoided.

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

14-day palatability study:
No analytical verification of test substance the test item-diet mixtures was performed.

EOGRTS:
For the determination of the test item concentration in the exposure diet, samples of approximately 10 g were taken and stored at - 20°C ±10 % or colder until
shipment for analysis. The samples were taken on the day of preparation of the test item-diet mixtures for analysis of concentration and homogeneity and 7 days after the preparation from the left over in the bucket for analysis of stability.

Sampling for F0 Generation
For the F0 Generation samples for analysis were taken from the weekly test item diet preparations at the start of dosing (test week 3) and from the test weeks when the animals were sacrificed (test week 20 for the male animals (sacrifice during test week 20) and test week 19 for the female animals (sacrifice during test week 19 and 20)).

Sampling for F1 Generation
For the F1 Generation samples for analysis were taken from the weekly test item diet preparations from test day 1 of the F1 Generation study and from the test
weeks when the different cohorts of the F1 Generation were sacrificed.

The samples were labelled with the study number, species, type of sample, concentration, generation and cohort number, group number test week, test day, location (top/middle/bottom), and date.

Final results on analytical verification of doses/concentrations were not available at the time of preparation of this robust study summary, but will be given in an update.

Duration of treatment / exposure:

14-day palatability study:
14 days

EOGRTS:
Parental generation (F0)= 18-19 weeks (10 weeks of pre-mating, 2 weeks mating, 3 weeks pregnancy, 3 weeks lactation) (test day 132-136)
F1 Cohort 1A= approx. 12-13 weeks (PND 86 to 96)
F1 cohort 1B= approx. 17-18 weeks (PND 120 to 136, at PND 4-8 of F2 generation)
F1 cohort 2A= approx. 12-13 weeks (PND 84 to 90)
F1 cohort 2B= approx. 3 weeks (PND 21 to 23)
F1 cohort 3= approx. 8-9 weeks (PND 53 to 61)
F2 Generation= PND 4 to 8
Surplus pubs= on PND 4 or PND 22

Frequency of treatment:

ad libitum, via diet

Details on study schedule:

- F1 parental animals not mated until PND 94.

Dose / conc.:

0 mg/kg bw/day (nominal)

Dose / conc.:

100 mg/kg bw/day (nominal)

Remarks:

Mean actual dose received (based on food consumption)
Parental generation: males 103.82 mg/kg bw/day; females 109.24 mg/kg bw/day
Cohort 1A: males 100.33 mg/kg bw/day; females 105.27 mg/kg bw/day
Cohort 1B: males 104.39 mg/kg bw/day; females 103.46 mg/kg bw/day
Cohort 2A: males 101.54 mg/kg bw/day; females 116.38 mg/kg bw/day
Cohort 3: males 88.2 mg/kg bw/day; females 97.2 mg/kg bw/day

Dose / conc.:

300 mg/kg bw/day (nominal)

Remarks:

Mean actual dose received (based on food consumption)
Parental generation: males 311.78 mg/kg bw/day; females 319.31 mg/kg bw/day
Cohort 1A: males 304.64 mg/kg bw/day; females 319.58 mg/kg bw/day
Cohort 1B: males 314.12 mg/kg bw/day; females 327.92 mg/kg bw/day
Cohort 2A: males 307.76 mg/kg bw/day; females 311.80 mg/kg bw/day
Cohort 3: males 270.0 mg/kg bw/day; females 286.9 mg/kg bw/day

Dose / conc.:

1 000 mg/kg bw/day (nominal)

Remarks:

Mean actual dose received (based on food consumption)
Parental generation: males 1050.04 mg/kg bw/day; females 1084.90 mg/kg bw/day
Cohort 1A: males 1066.35 mg/kg bw/day; females 1011.11 mg/kg bw/day
Cohort 1B: males 1045.59 mg/kg bw/day; females 1089.04 mg/kg bw/day
Cohort 2A: males 1028.08 mg/kg bw/day; females 993.33 mg/kg bw/day
Cohort 3: males 956.4 mg/kg bw/day; females 1042.8 mg/kg bw/day

No. of animals per sex per dose:

14-day palatability study:
5 animals/sex/dose

EOGRTS:
Parental generation (F0): 24 males/24 females
Cohort 1A: 20 males/20 females
Cohort 1B: 20 males/20 females
Cohort 2A: 10 males/10 females
Cohort 2B: 10 males/10 females
Cohort 3: 10 males/10 females

Control animals:

yes, plain diet

Details on study design:

14-palatablity study:
- Rationale for animal assignment: Following the initial health check approximately upon delivery, the animals were weighed and allocated based on the body weight by means of a computerised randomisation program to the test groups; only healthy animals were used. Animals of the extremes of the weight distribution and/or any animal showing signs of illness during the period between allocation and the start of treatment were excluded and replaced by spare animals.

EORGTS:
- Dose selection rationale: The dose levels were selected by the Sponsor based on available toxicological data and the results of a palatability study conducted prior to the main study.
- Rationale for animal assignment:
F0- generation: The animals were allocated to the test groups at the last day of the pre-dosing period on test day 14 by using a Provantis10-generated randomization based on the body weights of the animals. Only those female animals were used for randomization that passed the estrous cycle test. The body weight range did not exceed 20 % of the mean weight for each sex at the time of selection.
F1- generation: When the pups from those dams that were the first to have descendants, had reached post-natal day 21, the pups that were used for the establishment of the adult F1 Generation were randomly selected from all litters that were available for their respective group of the F0 Generation. The randomization procedure was performed using Provantis4. The selected pups were separated from the surplus pups after they have reached post-natal day 21. When all selected pups have reached post-natal day 21, the F1 Generation started with test day 1.
- Fasting period before blood sampling for clinical biochemistry: overnight

Positive control:

none

Parental animals: Observations and examinations:

14-day palatabillity study:
Clinical signs and mortality were recorded as described below. Body weight of each rat was recorded at the time of group allocation, on the first treatment day and daily thereafter throughout the experimental period. The quantity of food left by each animal was recorded on a daily basis throughout the experimental period. The food intake per animal (g/animal/day) was calculated using the total amount of food given to and left by each animal in each group on completion of a treatment day. The drinking water consumption was monitored daily by visual appraisal throughout the study. Individual test item intake was calculated on a daily and weekly basis throughout the experimental period based on concentration in the diet, individual food intake and body weight.


EOGRTS:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: at least once daily
Mortality was recorded twice daily. Behavioural changes, signs of difficult or prolonged parturition, and all signs of toxicity were recorded. Any signs of illness or reaction to treatment were recorded for each individual animal. Cage side observations included skin/fur, eyes, mucous membranes, respiratory and circulatory systems, somatomotor activity and behaviour patterns. The onset, intensity and duration of any signs observed were recorded. In addition, animals were checked regularly throughout the working day from 7:00 a.m. to 3:45 p.m. On Saturdays and Sundays, animals were checked regularly from 7:00 a.m. to 11:00 a.m. with a final check performed at approximately 3:30 p.m.

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: weekly
This more detailed examination started for the F0 main study animals on test day 14 (one day before the start of treatment) to allow for within-subject comparisons. Thereafter the examination was performed weekly until termination.
These observations were made outside the home cage in a standard arena and at the same time of the day, each time preferably by observers unaware of the treatment. Signs observed included changes in skin, fur, eyes, mucous membranes, occurrence of secretions and excretions and autonomic activity (e.g. lacrimation, piloerection, pupil size, and unusual respiratory pattern), as well as changes in gait, posture and response to handling as well as the presence of clonic or tonic movements, stereotypy (e.g. excessive grooming, repetitive circling), difficult or prolonged parturition and bizarre behaviour (e.g. self-mutilation, walking backwards). Because of the postulated adverse effects (in the public domain) on the gastrointestinal tract, the clinical observations also included the status or condition of the faeces (i.e., whether dry pellets or loose diarrhoea) and any evidence of faecal blood.

BODY WEIGHT: Yes
- Time schedule for examinations: The animals were weighed on the first day of dosing (test day 15), weekly thereafter and at sacrifice. Additionally, the female animals were weighed on gestation days 0, 7, 14 and 21 as well as on lactation days 1, 4, 7, 14 and 21.

FOOD CONSUMPTION AND COMPOUND INTAKE: Yes
Food consumption was measured in weekly intervals for the F0 main study animals.
The actual test item-intake for each test week is given in mg per kg body weight and day. The actual test-item intake was calculated based on the relative food consumption per day and the nominal test item concentration that was used in the respective test week.

WATER CONSUMPTION AND COMPOUND INTAKE: Yes
- Time schedule for examinations: Water consumption was measured in weekly intervals for the F0 main study animals.

HAEMATOLOGY: Yes
Whole blood samples were taken from the retrobulbar venous plexus under isoflurane anaesthesia from 10 animals/sex/group fasted overnight at necropsy (TW 20) (in EDTA anticoagulant for haematological investigations and in Citrate anticoagulant for coagulation tests). The following parameters were investigated:
Differential blood count (relative- Neutrophilic, eosinophilic and basophilic granulocytes, lymphocytes and monocytes), Differential blood count (absolute), Erythrocytes (RBC), Leucocytes (WBC), Haematocrit value (PCV or HCT), Haemoglobin content (HGB), Platelets (PCT or PLT), Reticulocytes (Reti), Mean corpuscular volume (MCV), Mean corpuscular haemoglobin (MCH), Mean corpuscular haemoglobin concentration (MCHC), Prothrombin time (PT), Activated partial thromboplastin time (aPTT).

CLINICAL BIOCHEMISTRY: Yes
Whole blood samples were taken from the retrobulbar venous plexus under isoflurane anaesthesia from 10 animals/sex/group fasted overnight at necropsy (TW 20) (in Li-Heparin anticoagulant for clinical chemistry tests). The following parameters were investigated:
Sodium, potassium, calcium, chloride, albumin, total bilirubin, total cholesterol, glucose, total protein, blood urea (BUN), creatinine, alanine aminotransferase (ALAT/GPT), alkaline phosphatase (aP), aspartate aminotransferase (ASAT/GOT), bile acids, lactate dehydrogenase (LDH), low-density lipoprotein (LDL), high-denssity lipoprotein (HDL), sodium/potassium ratio, globulin, albumin/globulin ratio, BUN/creatinine ratio.

THYROID HORMONES: Yes
Blood samples were taken from the scheduled animals under isoflurane anaesthesia. The blood was processed for serum. T4, T3 and TSH were analysed in serum samples of 10 animals/sex/dose (same animals as selected for laboratory examinations).

DETERMINATION OF SEXUAL HORMONES: Yes
Blood samples were taken from the scheduled animals under isoflurane anaesthesia at sacrifice for the determination of sexual hormones. A sufficient amount of blood was withdrawn to obtain at least 530 μL serum. Estradiol, testosterone and estrone were determined in 10 animals/sex/dose using commerical ELISA kits. Animals were fasted overnight.

Ti ANALYSIS IN BLOOD AND URINE: Yes
Lithium-heparin stabilised whole blood and urine samples for Ti level analysis were collected from non-fasted main study animals of the F0 Generation. For these examinations, the remaining 10 animals from each group were used that were not used for blood sampling for laboratory examinations and hormone level analysis. For urine sampling the adult animals were placed in URIMAX funnel cages for 5 h. Blood and urine samples from the adult animals were collected at sacrifice.

URINALYSIS: Yes
Urine samples were collected from animals fasted overnight. The urine was collected for 16 hours in a URIMAX funnel cage. The collection of urine was terminated immediately before the start of blood withdrawal for the haematological and clinical chemical examinations as well as the hormone level determinations. The following parameters were measured: Volume, pH, specific gravity. The following tests were also performed using qualitative indicators (Combur®Test, Roche Diagnostics GmbH, 68305 Mannheim, Germany) of analyte concentration: Protein, Glucose, Bilirubin, Urobilinogen, Ketones, Haemoglobin, Nitrite.
Microscopic examination of urine samples was carried out by centrifuging samples and spreading the resulting deposit on a microscope slide. The deposit was examined for the presence of the following parameters: Epithelial cells, Leucocytes, Erythrocytes, Organisms, further constituents, crystalluria. The colour and the turbidity of the urine were examined visually.

CIRCADIAN CYCLE OD HORMONE LEVELS (in satellite animals):
Hormone analysis of Beta-estradiol, Testosterone, T3, T4 and TSH was performed twice, pre-dose and at the end of the pre-mating period (after 10 weeks) using the F0 Satellite animals of groups 1 and 4 (30 animals per sex and group; 120 in total) throughout a 24-hour cycle. Blood sampling was performed every 2 hours during the 24-hour period of the cycle. The examination was performed at pre-dose (starting at 10.00 a.m. on test day 14 and ending on test day 15 at 8.00 p.m.) and after 10 weeks of treatment (starting at 10 a.m. on test day 84 and ending on test day 85 at 8.00 p.m.) before start of mating of the F0 satellite animals. In order to reduce the total amount of blood collected from each animal within the 24-hour period, each group (control group and high dose group) was split into six subgroups of 5 male and 5 female animals. Blood withdrawal from the 5 animals of each subgroup was performed twice during the cycle period (24 hours) with an interval of 12 hours between each time point of blood withdrawal. At each time point approx. 1.2 mL blood was collected from the tail-vein, sufficient enough to obtain at least 410 μL serum from each animal. The serum was aliquoted for the determination of each hormone. The aliquots were stored at - 20°C ± 10 % until analysis.

Oestrous cyclicity (parental animals):

Estrous cycle monitoring of F0 parental animals:
- 14 days pre-exposure period to select 96 main study animals and 80 satellite animals with regular estrous cycles (4 to 5 days per cycle) (main and satellite animals)
- During 10 weeks of pre-mating until evidence of mating (main study animals only)
- On the day of sacrifice.

Estrous cycle monitoring of Cohort 1B parental animals:
- Starting from the time of pairing until mating evidence was detected
- On the day of sacrifice

Estrous cycle monitoring of Cohort 1A animals:
- Start after onset of vaginal patency until first appearance of cornified cells
- Two weeks starting around PND 75
- On the day of sacrifice

Sperm parameters (parental animals):

Parameters examined in all males of the parental generation (F0) and cohort 1A:
The right epididymis and the right testis were used for sperm count and the examination of sperm viability and sperm morphology. The examinations were performed according to the methods described by L. Chahoud and R. Franz (1993) as well as by S. Plassmann and H. Urwyler (2001).

Detailed histopathological examination with special emphasis on the qualitative stages of spermatogenesis and histopathology of interstitial testicular structure was performed on one testicle and one epididymis of the F0 and F1 Cohort 1A males of groups 1 and 4 following H-E and PAS staining.

Organ weights of both epididymides and testicles of all parental (F0) and cohort 1A animals were recorded.

Litter observations:

STANDARDISATION OF LITTERS
- Performed on day 4 postpartum: yes (for F0 generation)
- If yes, maximum of 10 pups/litter (5/sex/litter as nearly as possible);
After counting on PND 4, the litters were adjusted to 10 pups per litter (5 pups per sex and litter) by eliminating (culling) surplus pups using a randomization scheme generated by Provantis®. Selective elimination of pups e.g. based upon body weight is not appropriate and was not performed. In case of unequal gender distribution, a partial litter size adjustment was performed (e.g. 6 male and 4 female pups).
Culled pups were used for blood collection for the determination of thyroid hormone levels on PND 4 and PND 22.

PARAMETERS EXAMINED
The following parameters were examined in F1 and F2 offspring (besides the offspring indices): presence of nipples/areolae in male pups (F1 only), balano-preputial separation or vaginal opening (and body weight at that time point) (F1 only), anogenital distance (AGD), pup weight on the day of AGD, external and internal examination, body weight gain, thyroid hormones (T4 on PND 4 and T3,T4 and TSH on PND 22 of F1 pups only), pup organ weights (absolute) on PND 22 (from surplus F1 pups)

GROSS EXAMINATION OF DEAD PUPS:
Dead pups and F1 pups sacrificed on PND 4 (culled F1 Pups) were carefully examined externally for gross abnormalities. The external reproductive genitals were examined for signs of altered development. Those F1 Pups that were not selected for the F1 cohorts were sacrificed on PND 22 and dissected. They were examined macroscopically for external and internal gross abnormalities. From up to 10 pups per sex and group organs were weighed and preserved in 7% buffered formalin.

OBSERVATIONS AND EXAMINATIONS
COHORT 1A, 1B, 2A, 2B and 3

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: at least once daily (all F1 cohorts and satellite animals)
Mortality was recorded twice daily. Behavioural changes, signs of difficult or prolonged parturition, and all signs of toxicity were recorded. Any signs of illness or reaction to treatment were recorded for each individual animal. Cage side observations included skin/fur, eyes, mucous membranes, respiratory and circulatory systems, somatomotor activity and behaviour patterns. The onset, intensity and duration of any signs observed were recorded. In addition, animals were checked regularly throughout the working day from 7:00 a.m. to 3:45 p.m. On Saturdays and Sundays, animals were checked regularly from 7:00 a.m. to 11:00 a.m. with a final check performed at approximately 3:30 p.m.

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: weekly (Cohort 1B only)
This more detailed examination started after weaning until termination.
These observations were made outside the home cage in a standard arena and at the same time of the day, each time preferably by observers unaware of the treatment. Signs observed included changes in skin, fur, eyes, mucous membranes, occurrence of secretions and excretions and autonomic activity (e.g. lacrimation, piloerection, pupil size, and unusual respiratory pattern), as well as changes in gait, posture and response to handling as well as the presence of clonic or tonic movements, stereotypy (e.g. excessive grooming, repetitive circling), difficult or prolonged parturition and bizarre behaviour (e.g. self-mutilation, walking backwards). Because of the postulated adverse effects (in the public domain) on the gastrointestinal tract, the clinical observations also included the status or condition of the faeces (i.e., whether dry pellets or loose diarrhoea) and any evidence of faecal blood.

BODY WEIGHT: Yes
- Time schedule for examinations: Live pups were weighed during the lactation period on their post-natal days 1, 4, 7, 14 and 21. After weaning they were weighed weekly. The female animals of Cohort 1B were additionally weighed on their gestation days 0, 7, 14 and 21. The F2 Pups were weighed on their post-natal days 1 and 4 (sacrifice).

FOOD CONSUMPTION AND COMPOUND INTAKE: Yes
Food consumption was measured in weekly intervals for all animals of all cohorts starting shortly after weaning (except cohort 2B, since they were sacrificed after weaning).
The actual test item-intake for each test week is given in mg per kg body weight and day. The actual test-item intake was calculated based on the relative food consumption per day and the nominal test item concentration that was used in the respective test week.

WATER CONSUMPTION AND COMPOUND INTAKE: Yes
- Time schedule for examinations: Water consumption was measured in weekly intervals for all cohort 1B animals (starting shortly after weaning until sacrifice).

HAEMATOLOGY: Yes
Whole blood samples were taken from all Cohort 1A animals from the retrobulbar venous plexus under isoflurane anaesthesia from 10 animals/sex/group fasted overnight at necropsy (PND 87 to 96) (in EDTA anticoagulant for haematological investigations and in Citrate anticoagulant for coagulation tests). The following parameters were investigated:
Differential blood count (relative- Neutrophilic, eosinophilic and basophilic granulocytes, lymphocytes and monocytes), Differential blood count (absolute), Erythrocytes (RBC), Leucocytes (WBC), Haematocrit value (PCV or HCT), Haemoglobin content (HGB), Platelets (PCT or PLT), Reticulocytes (Reti), Mean corpuscular volume (MCV), Mean corpuscular haemoglobin (MCH), Mean corpuscular haemoglobin concentration (MCHC), Prothrombin time (PT), Activated partial thromboplastin time (aPTT).

CLINICAL BIOCHEMISTRY: Yes
Whole blood samples were taken from all cohort 1A animals from the retrobulbar venous plexus under isoflurane anaesthesia from 10 animals/sex/group fasted overnight at necropsy (PND 87 to 96) (in Li-Heparin anticoagulant for clinical chemistry tests). The following parameters were investigated:
Sodium, potassium, calcium, chloride, albumin, total bilirubin, total cholesterol, glucose, total protein, blood urea (BUN), creatinine, alanine aminotransferase (ALAT/GPT), alkaline phosphatase (aP), aspartate aminotransferase (ASAT/GOT), bile acids, lactate dehydrogenase (LDH), low-density lipoprotein (LDL), high-denssity lipoprotein (HDL), sodium/potassium ratio, globulin, albumin/globulin ratio, BUN/creatinine ratio.

THYROID HORMONES: Yes
Blood samples were taken from all Cohort 1A animals (fasted) and F1 surplus pups (on PND 4 (T4 only) or 22, both non-fasted) under isoflurane anaesthesia. The blood was processed for serum. T4, T3 and TSH were analysed in serum samples of 10 animals/sex/dose (same animals as selected for laboratory examinations).

DETERMINATION OF SEXUAL HORMONES: Yes
Blood samples were taken from the scheduled animals under isoflurane anaesthesia at sacrifice for the determination of sexual hormones. A sufficient amount of blood was withdrawn to obtain at least 530 μL serum. Estradiol, testosterone and estrone were determined in 10 animals/sex/dose of Cohort 1A, 1B, 2A and 2B using commerical ELISA kits. Animals were fasted overnight.

Ti ANALYSIS IN BLOOD AND URINE: Yes
Lithium-heparin stabilised whole blood and urine samples for Ti level analysis were collected from 10 non-fasted animals/sex/dose of Cohort 1A, 1B and F2 pups. For these examinations, the remaining 10 animals fromeach group were used that were not used for blood sampling for laboratory examinations and hormone level analysis. For urine sampling the adult animals were placed in URIMAX funnel cages for 5 h. Blood and urine samples from the adult animals were collected at sacrifice. For the F2 pups urine was directly collected from the urinary bladder before necropsy.

URINALYSIS: Yes
Urine samples were collected from Cohort 1A animals fasted overnight (on PND 87 to 96). The urine was collected for 16 hours in a URIMAX funnel cage. The collection of urine was terminated immediately before the start of blood withdrawal for the haematological and clinical chemical examinations as well as the hormone level determinations. The following parameters were measured: Volume, pH, specific gravity. The following tests were also performed using qualitative indicators (Combur®Test, Roche Diagnostics GmbH, 68305 Mannheim, Germany) of analyte concentration: Protein, Glucose, Bilirubin, Urobilinogen, Ketones, Haemoglobin, Nitrite.
Microscopic examination of urine samples was carried out by centrifuging samples and spreading the resulting deposit on a microscope slide. The deposit was examined for the presence of the following parameters: Epithelial cells, Leucocytes, Erythrocytes, Organisms, further constituents, crystalluria. The colour and the turbidity of the urine were examined visually.

ASSESSMENT OF DEVELOPMENTAL NEUROTOXICITY:
Neurological screening of Cohort 2A animals
Auditory startle response (on PND 22/23), observation screening and functional tests (males on PND 60 to 64 and females on PND 58 to 63). The observational screening included the following parameters: righting reflex, Body temperature, Salivation, Startle response, Respiration, Mouth breathing, Urination, Convulsions, Piloerection, Diarrhoea, Pupil size, Pupil response, Lacrimation, Impaired gait, Stereotypy, Toe pinch, Tail pinch, Wire manoeuvre, Hind-leg splay, Positional passivity, Tremors, Positive geotropism, Limb rotation, Auditory function, Posture, Palpebral closure, Vocalization, Arousal, Bizarre behaviour, Ease of removal/handling, Muscle tone, Approach response, Touch response
The functional tests included measurement of grip strength and locomotor activity

ASSESSMENT OF DEVELOPMENTAL IMMUNOTOXICITY:
The assessment of potential developmental immunotoxicity of cohort 3 animals included the determination of anti KLH-IgM antibodies in the blood serum using ELISA and teh examination of the splenic lympocyte subpopulations via flow cytometry (cohort 1A and 3).

KLH-assay:
All animals of cohort 3 and satellite animals nos. 941 to 960 (positive control animals) were used to determine the IgM values in serum after a single i.v. bolus injection of 1500 µg/animal KLH (keyhole limpet haemocyanin, in PBS) into the tail vein on PND 53 to 61. Positive control animals received an additional single oral administration on the same day of KLH injection of 40 mg/kg bw of CPA (Cycloposphamide) solved in 0.5% Methyl Cellulose. 5 days after injection, animals were sacrificed and sufficient blood was withdrawn from the animals to receive 2 aliquots of at least 150 μL serum each to determine the IgM values with a Rat Anti-KLH IgM ELISA kit.

Phenotypic analysis of spleen cells (cohort 1A and 3):
The analysis of the subpopulation of lymphocytes in the spleen was performed for all animals of Cohort 1A, Cohort 3 and the satellite animal nos. 941 to 960.
At necropsy, the spleen of the mentioned animals was split in 2 parts. The part of the spleen not preserved for histopathology (histopathology of the spleen was only performed for Cohort 1A) was minced using a mechanic dissociator to prepare single cell suspensions. The prepared spleen samples were used for the determination of the following lymphocyte subpopulations via flow cytometry: helper T cells (CD4+), suppresor/cytotoxic T cells (CD8+), T cells (CD3+), B cells (CD45RA+) and NK cells (CD161+)

Postmortem examinations (parental animals):

14-day palatability study:
On test day 15 (approximately 2 hours after removal of the diet), all animals were euthanized by carbon dioxide (CO2) inhalation, exsanguinated by cutting the aorta abdominalis, weighed, dissected and inspected macroscopically under the direction of a pathologist. All superficial tissues were examined visually and by palpation and the cranial roof was removed to allow observation of the brain, pituitary gland and cranial nerves. After ventral midline incision and skin reflection, all subcutaneous tissues were examined. The condition of the thoracic viscera was noted with due attention to the thymus, lymph nodes and heart. The abdominal viscera were examined before and after removal. The urinary bladder was examined externally and by palpation. The gastro-intestinal tract was
examined as a whole, and the stomach and caecum were incised and examined. The lungs were removed and all pleural surfaces were examined under suitable illumination. The liver and the kidneys were examined. Any abnormalities in the appearance and size of the gonads, adrenal glands, uterus, intra-abdominal lymph nodes, and accessory reproductive organs were recorded. The organs and carcasses were discarded.


EOGRTS:

SACRIFICE
- Male animals: All surviving animals on test day 135, 136 (after weaning of the F1 animals)
- Maternal animals: All surviving animals on test day 132 to 135 (after weaning of the F1 animals)
Satellite animals used for ACF determination: shortly after weaning of the F1 satellite animals
Satellite animals used for hormone cycle blood sampling: after the last blood sampling

GROSS NECROPSY
At the time of sacrifice or premature death during the study, the adult animals were examined macroscopically for any abnormalities or pathological changes. Special attention was paid to the organs of the reproductive system and GI tract, particularly whether for example the caecum was dilated. During necropsy the number of implantation sites was recorded in the female animals. Apparently non-pregnant uteri were placed in a 10% aqueous solution of ammonium sulfide for about 10 minutes to stain possible implantation sites in the endometrium according to SALEWSKI.
All superficial tissues were examined visually and by palpation and the cranial roof removed to allow observation of the brain, pituitary gland and cranial nerves. After ventral midline incision and skin reflection all subcutaneous tissues were examined. The condition to the thoracic viscera was noted with due attention to the thymus, lymph nodes and heart. The abdominal viscera were examined before and after removal; the urinary bladder was examined externally and by palpation. The gastro-intestinal tract was examined as a whole and the stomach and the caecum were incised and examined. The lungs were removed and all pleural surfaces were examined under suitable illumination. The liver and the kidneys were examined. Any abnormalities in the appearance and size of the gonads, adrenals, uterus, intra-abdominal lymph nodes and accessory reproductive organs were recorded.

HISTOPATHOLOGY / ORGAN WEIGHTS
Organ weights:
The weights of the organs were recorded from 20 males and female animals of the F0 Generation. With the exception of the thyroid weight (determined after fixation), all organs wereweighed before fixation. Paired organs were weighed individually and identified as left or right.
Organs weights recorded from brain, epidodymis, heart, kidney, liver, lymph node (mesenteric and cervical), ovary, spleen, testicle, thyroid, thymus, prostate, semincal vesicles with coagulating glands, pituitary, uterus including cervix and oviducts.

Histopathology:
The histopathology was performed on the preserved organs of all F0 animals of group 1 and 4 (control and high dose). Since no relevant findings were obserd in group 1 and 4, no further histopathological examination of the low and mid dose animals was required.
The organs listed below were examined histopathologically after preparation of paraffin sections and haematoxylin-eosin staining. Parathyroids could not always identified macroscopically. They were examined microscopically if in the plane of section. In addition, frozen sections of the heart, liver and one kidney were prepared, stained with Oil Red O, and examined.
Eye with optic nerve, epididymis, adrenal gland, bone, bone marrow (os femoris), brain (cerebrun, cerebellum, pons), gross lesions, heart, intestine (small and large), kidney and ureter, liver, lungs, lymph node (cervical and mesenteric), mammary gland, muscle (skeletal), testicle, nerve (sciatic), oesophagus, ovary, pituitary, prostate, semincal vesicles with coagulating glands, spinal cord, spleen, stomach, thyroid, thymus, trachea, urinary bladder, uterus, vagina

Bone marrow:
During dissection fresh bone marrow was obtained from the os femoris (3 airdried smears / animal) form 10 male and 10 female animals from all groups of the F0 Generation. The same animals were used as those that were selected for the laboratory examinations and the hormone level determinations. The myeloid:erythroid ratio was determined from animals of groups 1 and 4 by cell differentiation (counting of 200 nuclei-containing cells).

Evaluation of aberrant crypt foci (ACF):
10 male and 10 female F0 Satellite animals per group (80 animals in total) were used for evaluation of aberrant crypt foci (ACF). As gestation and lactation are companied by profound physiological changes, the F0 animals for ACF determination were paired and sacrificed after weaning (male and female parental F0 animals). This is necessary to allow correlation of results from ACF determination in the satellite animals with results of regular histopathologic examination of the gastrointestinal tract of parental F0 main study animals that were sacrificed after weaning. Following a randomization scheme, the animals were euthanized by carbon dioxide (CO2) inhalation and exsanguinated by cutting the aorta abdominalis. The colon was taken, opened longitudinally with scissors, and gently rinsed with 0.9% NaCl solution to remove the colon contents. The colon was cut into pieces of suitable size to fit into histological cassettes; the pieces were placed on numbered pieces of filter paper of the same size. The most cranial piece was put onto filter paper number 1, all following pieces on filter papers with ascending numbers going towards the
rectum. Filter papers of an individual animal were placed in one cassette in numerical order with the paper number 1 (the most cranial portion) located at the bottom. The colon portions had to be as flat as possible to facilitate later processing and evaluation. To make sure that the colon portions were pressed flat and do not bulge or roll in the closed cassette, rolls of filter paper were added to the cassette if necessary. The closed tissue cassettes were labelled with study number and animal number and immersed in 5 % buffered formalin. Prior to transfer to the Investigating Scientist the samples were coded to ensure blinded analysis. As the examination of aberrant crypt foci is not a standard method, this part of the
study is not in compliance with GLP.

Postmortem examinations (offspring):

SACRIFICE
- Cohort 1A on PND 86 to 96
- Cohort 1B on PND 120 to 136 (after sacrifice of the F2 pups)
- Cohort 2A on PND 84 to 90
- Cohort 2B on PND 21 to 23
- Cohort 3 on PND 53 to 61
- F2 generation on PND 4 to 8
- Surplus pups (F1 generation, culled): PND 4
- Surplus pups (not selected for the cohorts) on PND 22

GROSS NECROPSY

Surplus F1 pups and the F2 pups:
External inspection for gross abnormalities: Dead pups and F1 pups sacrificed on PND 4 (culled F1 Pups) were carefully examined externally for gross abnormalities. The external reproductive genitals were examined for signs of altered development.
External and internal inspection for gross abnormalities:
F1 Pups: Those F1 Pups that were not selected for the F1 cohorts were sacrificed on PND 22 and dissected. They were examined macroscopically for external and internal gross abnormalities. From up to 10 pups per sex and group organs were weighed and preserved in 7% buffered formalin.
F2 Pups: The F2 pups that were terminally sacrificed between postnatal days 4 and 8 were dissected and macroscopically examined for external and internal gross abnormalities.

Adult animals:
At the time of sacrifice or premature death during the study, the adult animals were examined macroscopically for any abnormalities or pathological changes. Special attention was paid to the organs of the reproductive system and GI tract, particularly whether for example the caecum was dilated. During necropsy the number of implantation sites was recorded in the female animals. Apparently non-pregnant uteri were placed in a 10% aqueous solution of ammonium sulfide for about 10 minutes to stain possible implantation sites in the endometrium according to SALEWSKI.
All superficial tissues were examined visually and by palpation and the cranial roof removed to allow observation of the brain, pituitary gland and cranial nerves. After ventral midline incision and skin reflection all subcutaneous tissues were examined. The condition to the thoracic viscera was noted with due attention to the thymus, lymph nodes and heart. The abdominal viscera were examined before and after removal; the urinary bladder was examined externally and by palpation. The gastro-intestinal tract was examined as a whole and the stomach and the caecum were incised and examined. The lungs were removed and all pleural surfaces were examined under suitable illumination. The liver and the kidneys were examined. Any abnormalities in the appearance and size of the gonads, adrenals, uterus, intra-abdominal lymph nodes and accessory reproductive organs were recorded.

HISTOPATHOLOGY / ORGAN WEIGTHS

Organ weights:
Surplus F1 pups:
For the F1 Pups that were sacrificed on PND 22 preservation and / or weighing of organs from up to 10 pups per sex and group was scheduled. Organs preserved and weighed were brain, gross abnormalities, heart, intestine (small and large, large were preserved only), kidney, liver, spleen, lungs, mammary gland, stomach, thymus.

Adult animals:
The weights of the organs listed below were recorded from 20 males and female animals from Cohort 1A (the listed lymph nodes were only weighed from 10 males and females of Cohort 1A). From all male and female animals of Cohort 2A and 2B only the brain weight was determined. With the exception of the thyroid weight (determined after fixation), all organs were weighed before fixation. Paired organs were weighed individually and identified as left or right.
Organs weights recorded from brain, epidodymis, heart, kidney, liver, lymph node (mesenteric and cervical), ovary, spleen, testicle, thyroid, thymus, prostate, semincal vesicles with coagulating glands, pituitary, uterus including cervix and oviducts. From Cohort 2A and 2B only the brain weight was determined.

Histopathology of cohort 1A animals (group 1 and 4):
The histopathology was performed on the preserved organs of all Cohort 1A animals of group 1 and 4 (control and high dose). Since no relevant findings were obserd in group 1 and 4, no further histopathological examination of the low and mid dose animals was required.
The organs listed below were examined histopathologically after preparation of paraffin sections and haematoxylin-eosin staining. Parathyroids could not always identified macroscopically. They were examined microscopically if in the plane of section. In addition, frozen sections of the heart, liver and one kidney were prepared, stained with Oil Red O, and examined.
Eye with optic nerve, epididymis, adrenal gland, bone, bone marrow (os femoris), brain (cerebrun, cerebellum, pons), gross lesions, heart, intestine (small and large), kidney and ureter, liver, lungs, lymph node (cervical and mesenteric), mammary gland, muscle (skeletal), testicle, nerve (sciatic), oesophagus, ovary, pituitary, prostate, semincal vesicles with coagulating glands, spinal cord, spleen, stomach, thyroid, thymus, trachea, urinary bladder, uterus, vagina.
Detailed histopathological examination with quantitative evaluation of primordial and small growing follicles as well as corpora lutea were performed on one ovary of the F1 Cohort 1A females of groups 1 and 4.

Bone marrow:
During dissection fresh bone marrow was obtained from the os femoris (3 airdried smears / animal) form 10 male and 10 female animals from all groups of Cohort 1A. The same animals were used as those that were selected for the laboratory examinations and the hormone level determinations. The myeloid:erythroid ratio was determined from animals of groups 1 and 4 by cell differentiation (counting of 200 nuclei-containing cells).

Neurohistopathology of Cohort 2A and 2B animals:
Haematoxylin-eosin and lyxul fast blue cresyl violet staining on the following organs of cohort 2A animals (group 1 and 4): Brain (olfactory bulbs, cerebral cortex, hippocampus, basal ganglia, thalamus, hypothalamus, mid-brain (thecum, tegmentum, and cerebral peduncles), brain-stem and cerebellum).
Haematoxylin-eosin staining on the following organs of cohort 2A animals (group 1 and 4): Eye with optic nerve and retina (left and right), Muscle (skeletal), Nerve (sciatic), Spinal cord (3 sections)
Haematoxylin-eosin staining on the following organs of cohort 2B animals (group 1 and 4): Brain (olfactory bulbs, cerebral cortex, hippocampus, basal ganglia, thalamus, hypothalamus, mid-brain (thecum, tegmentum, and cerebral peduncles), brain-stem and cerebellum).

Histopathological evaluation included:
- alterations in the gross size or shape of the olfactory bulbs, cerebrum or cerebellum;
- alterations in the relative size of various brain regions, including decreases increases in the size of regions resulting from the loss or persistence of normally transient populations of cells or axonal projections (e.g., externa germinal layer of cerebellum, corpus callosum);
- alterations in proliferation, migration, and differentiation, as indicated by areas of excessive apoptosis or necrosis, clusters or dispersed populations of ectopic,disoriented or malformed neurons or alterations in the relative size of various layers or cortical structures;
- alterations in patterns of myelination, including an overall size reduction or altered staining of myelinated structures.
- evidence of hydrocephalus, in particular enlargement of the ventricles, stenosis of the cerebral aqueduct and thinning of the cerebral hemispheres;
- morphometric (quantitative) evaluation: linear or areal measurement of the following major brain regions: cerebral cortex, mid-brain (thecum, tegmentum, and cerebral peduncles), brain-stem and cerebellum.

Statistics:

14-day palatability study:
Due to the limited number of characters in this field please refer to the field "any other information on materials and methods incl. tables".

EOGRTS:

DATA AQUISITION:
The following data were captured or calculated by the departmental computerized system (Provantis® integrated preclinical software, version 10.2.1, Instem LSS Ltd):
Clinical signs, body weight, body weight gain, food consumption, haematological and biochemical parameters.
Raw data not fully compatible with the computerized system (e.g. data from the ELISA experiments, neurological screening or pup data) were maintained on paper according to the appropriate SOPs.
Data maintained on paper (e.g. data from the ELISA experiments, the neurological screening or pup data) was entered in Provantis in a retrospective manner using the laboratory records according to the appropriate SOPs.

STATISTICS:
Parametrical data
The statistical evaluation of the parametrical values was done by Provantis (Provantis® integrated preclinical software, version 10.2.1, Instem LSS Ltd) using
the following settings: Homogeneity of variances and normality of distribution were tested using the BARTLETT’s and SHAPIRO-WILK’s test. In case of heterogeneity and/or nonnormality of distribution, stepwise transformation of the values into logarithmic or rank values was performed prior to ANOVA. If the ANOVA yielded a significant effect (p ≤ 0.05), intergroup comparisons with the control group were made by the DUNNETT’s test (p ≤ 0.01 and p ≤ 0.05) (see the decision tree on the following page).

Non-parametrical data
Bone marrow: statistical evaluation of the myeloid / erythroid ratio using the Chi2 test with StatXact 4.0.1 software.

Reproductive indices:

Reproducte indices for the F0 females and females of Cohort 1B:
Female fertility index (%)
Gestation index (%)
Pre-coital time
Gestation length

Offspring viability indices:

Implantation sites
- number per dam
- distribution in the uterine horns (implantation sites left or right)
- absolute number per group
- mean per group
Number of pups per group and per dam
- at birth (alive and dead)
- on post-natal day 4
- on post-natal days 7, 14 and 21 (only F0 females)
Number of male and female pups per group and per dam
- at birth (alive and dead)
- on post-natal day 4
- on post-natal days 7, 14 and 21 (only F0 females)
Number of stillbirths
- per group
- per dam
Number of pups with malformations
- per group
- per dam

Clinical signs:

effects observed, treatment-related

Description (incidence and severity):

At the intermediate and high dose level (300 or 1000 mg Titanium dioxide E171/kg b.w./day) test item-related changes were noted for the male and female animals in the form of pale faeces caused by the white colour of Titanium dioxide E171-E in the diet. This observation was considered to be of no toxicological relevance.

Dermal irritation (if dermal study):

not examined

Mortality:

no mortality observed

Body weight and weight changes:

no effects observed

Food consumption and compound intake (if feeding study):

no effects observed

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

no effects observed

Ophthalmological findings:

not examined

Haematological findings:

no effects observed

Clinical biochemistry findings:

no effects observed

Urinalysis findings:

no effects observed

Behaviour (functional findings):

not examined

Immunological findings:

not examined

Organ weight findings including organ / body weight ratios:

no effects observed

Histopathological findings: non-neoplastic:

no effects observed

Histopathological findings: neoplastic:

no effects observed

Other effects:

no effects observed

Reproductive function: oestrous cycle:

no effects observed

Reproductive function: sperm measures:

no effects observed

Reproductive performance:

no effects observed

14-day palatability study:

CLINICAL SIGNS:
None of the male and female rats treated with nominal dose levels of 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day via the diet for 14 consecutive days died during the course of the study. Changes in behaviour or external appearance did not occur. The faeces of all animals were of a normal consistency and did not reveal any further abnormalities.

BODY WEIGHT AND BODY WEIGHT GAIN:
No test item-related influence was noted on the body weight and the body weight gain of the male and female rats treated with nominal dose levels of 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day via the diet for 14 consecutive days compared to the control animals.

FOOD AND DRINKING WSTER CONSUMPTION:
No test item-related influence was noted on the food consumption of the male and female rats treated with nominal dose levels of 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day via the diet for 14 consecutive days compared to the control animals. The data did not reveal an influence of Titanium dioxide E171-E on the palatability of the diet: At the exposure level of 300 mg Titanium dioxide E171-E/kg b.w./day via the diet, the food consumption of the male and female animals ranged from 8% or 2% below to 11% or 24% above the values of the control group. There was no clear tendency and no statistically significant difference compared to the control group. In the high exposure level group of 1000 mg Titanium dioxide E171-E/kg b.w./day, the food consumption of the male and female animals ranged from 5% or 8% below to 11% or 16% above the value of the control group. There was no doseresponse- relationship compared to the low dose level group and no statistically significant difference compared to the control group. The visual appraisal of the drinking water consumption revealed no differences between the control and the test item-treated animals.

TEST ITEM INTAKE:
The test item-diet mixtures containing nominal dose levels of 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day and the calculation of the test item intake did not reveal any noteworthy effect on the palatability of the feed for the male and female rats after the 14-day treatment period. The mean intake of Titanium dioxide E171-E/kg b.w./day via the diet over 14 days was 305 mg/kg b.w./day (males) or 348 mg/kg b.w./day (females) at the low exposure level and 988 mg/kg b.w./day (males) or 1088 mg/kg b.w./day (females) at the high exposure level of the test item in the diet. Therefore, the calculated test item intake was 102% or 99% (of the nominal = theoretical values) for the male animals and 116% or 109% for the female animals after the 14-day treatment period. During the 14-day treatment period of the test item-diet mixture containing nominal dose levels of 300 mg Titanium dioxide E171-E/kg b.w./day the mean test item intake ranged from 253 to 358 mg/kg b.w./day for the male rats and from 258 to 462 mg/kg b.w./day for the female rats. Treatment with the higher nominal concentration of 1000 mg Titanium dioxide E171-E/kg b.w./day via the diet led to a test item intake in the range from 831 to 1120 mg/kg b.w./day for the males and 799 to 1322 mg/kg b.w./day for the females. Fluctuations noted on the test item intake during the 14-day treatment period, in particular at the nominal high dose level, are considered to be not test item-related as variations also occurred for the control animals (group 1) on the food consumption. Furthermore, the adjustment of test item concentration in the diet for the 2nd treatment week must be taken into account.

GROSS NECROPSY:
At necropsy on test day 15, no macroscopic changes were noted in the male and female rats treated with nominal dose levels of 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day via diet for 14 consecutive days.



EOGRTS:

For the assessment of general toxicity 24 males and females per group were examined. However, as 20 dams with live born pups were received for each litter (after exclusion of one low dose, two mid dose and two high dose non-pregnant females, discussed below), no necropsy was perfromed for the remaining dams (four control, three low dose, two mid dose and two high dose females). Hence, the number of implantation sites was not determined, and the animals were excluded from the evaluation of the reproductive parameter. However, pups from the excluded animals were used for the F1 generation.
For laboratory examinations (such as haematology and clinical biochemistry) only 10 animals/sex/dose were used. During necropsy and for the determination of organ weights and histopathology 20 animals/sex/dose were used.

GENERAL TOXICITY


CLINICAL SIGNS:
No test item-related changes in behaviour and the external appearance were noted for the male and female animals of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).
Non-test item-related changes in the external appearance were noted for one female animal of the low dose group (no. 93 with pale eyes on lactation days 1 and 2) and one female animal of the high dose group (no. 170 with piloerection during an elongated littering period on gestation days 23 to 25). As these observations were only noted for one animal each on 2 or 3 test days, they were considered to be spontaneous.

MORTALITY:
No premature death was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the male and female animals.

DETAILED CLINICAL OBSERVATIONS:
No observations were noted during the detailed clinical observations for the animals of the control group and the animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BODY WEIGHT AND BODY WEIGHT GAIN:
No test item-related changes in body weight and body weight gain were noted for the male and female rats between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

FOOD CONSUMPTION:
No test item-related difference in food consumption was noted between the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171- E/kg b.w./day).

WATER CONSUMPTION:
No test item-related difference in drinking water consumption was noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A slight but statistically significantly increased drinking water consumption was noted at the low dose level between test days 29 and 36 (8.3 % above the control value for the female animals, p ≤0.05). This slight and temporary increase was considered to be spontaneous.

TEST ITEM INTAKE:
The actual mean test item intake during the whole study was in the range of the nominal values for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). No significantly lower actual values in comparison to the nominal values were noted
for the male and female animals. However, for the female animals higher actual values than the nominal values were noted in test week 16 (last week of gestation period). As no signs of toxicity were noted in the whole study, the temporary higher actual dose levels were not relevant, moreover, as the mean values for the whole study period were in the range of the nominal values as mentioned above.

HAEMATOLOGY:
No test item-related differences for the examined haematological parameters were noted between the control group and the treatment groups (100, 300 or 1000 mg
Titanium dioxide E171-E/kg b.w./day). However, statistically significant changes in comparison to the control group were noted at the low dose level for the haemoglobin (HGB) concentration and the haematocrit (HCT) value for the male animals. However, as the differences were only marginal and no dose response-relationship was noted, the differences were considered to be spontaneous.

CLINICAL BIOCHEMISTRY:
No test item-related differences for the examined biochemical parameters were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). However, statistically significant changes in comparison to the control group were noted for the LDL cholesterol concentration, the alkaline phosphatase (aP) and the chloride concentration. However, as no dose response-relationship was noted and the differences were observed in one sex only, they were considered to be spontaneous.

URINALYSIS:
No test item-related differences for the examined urinalysis parameters were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

SEXUAL HORMONE LEVELS:
MALES:
No test item-related differences for the examined sexual hormone levels (estradiol, estrone and testosterone) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Increased estrone levels in comparison to the control group were noted at the intermediate and the high dose level, statistically significant only at the intermediate dose level. However, though the mean values were increased in comparison to the control group, the individual values from 10 of 10 animals of the intermediate dose group (range between 8.7 and 14.0 pg estrone/mL) and from 9 of 10 animals of the high dose group (range between 7.4 and 14.4 pg estrone/mL) were in or only marginally above the range of the control group (between 5.8 and 13.5 pg estrone/mL). One increased value was only noted for the high dosed animal no. 168 (36.4 pg estrone/mL). Hence, as this increased value was only noted for 1 animal of 20 animals of the high dose group, the observation was considered to be spontaneous. For the purpose of an assessment of the relevance of any minor alterations in hormone levels, a 24-hour circadian cycle was monitored in male and female animals at two different time points in satellite animals. As expected, these measurements document that during a 24 h day, all investigated hormones are subject to substantial variation.

FEMALES:
No test item-related differences for the examined sexual hormone levels (estradiol, estrone and testosterone) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

THYROID HORMONE LEVEL:
MALES:
No test item-related differences for the examined thyroid hormone levels (T3, T4 and TSH) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).
FEMALES:
No test item-related differences for the examined thyroid hormone levels were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A slight but statistically significantly increased T3 level in comparison to the control group was noted at the high dose level. However, all mean and all individual values of the control group and the treatment groups were inside the range of the mean values from the examined time points that were measured during the examination of the 24 hour cycle of T3 concentration. Hence, the slight increase was considered to be spontaneous.

CIRCADIAN CYCLES OF HORMONE LEVELS:
The results clearly document that (gender-specific) sexual and thyroid hormones are subject to substantial variation during the day. Overall, no test itemrelated effect on hormone levels were noted between the control group and the high dose group (1000 mg Titanium dioxide E171-E/kg b.w./day) for the examined time points of the daily circadian cycle after 10 weeks of dosing.

Ti ANALYSIS IN BLOOD AND URINE:
The analysis of titanium levels in blood and urine samples taken throughout the study is not yet finalised but will be provided in an update of this robust study summary.

BODY WEIGT AT AUTOPSY:
No test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the body weight at autopsy for the male and female animals. A slightly increased body weight at autopsy was noted for the female animals of the low dose group (9.1 % above the control value; statistically not significant). As no statistical significance and no dose-response relationship were noted, this observation was considered to be spontaneous.

GROS NECROPSY:
MALES:
No observations were noted for the male animals of the low and the intermediate dose group (100 or 300 mg Titanium dioxide E171-E/kg b.w./day) and no test itemrelated observation was noted for the male animals of the high dose group (1000 mg Titanium dioxide E171-E/kg b.w./day). The following observations were noted for the males of the control and the high dose group.
Group 1: Adrenal gland right and left: small (no. 16)
Group 4: Duodenum: Mucosa with white layer (no. 155)
The observation in animal no. 155 of the high dose group (mucosa with white layer) was confirmed by microscopy (fine granular deposit) and caused by the high concentration of the test-item in group 4. It was not associated with any pathological observation and considered to be of no toxicological relevance.

FEMALES:
No test item-related observations were noted for the female animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171/kg b.w./day).
The following isolated observations were considered to be spontaneous:
Group 1: Cervix: thickened, glazed induration, approx. 8 mm in diameter (no. 47)
Group 2: Pituitary: Haematoma, 2 mm diameter (no. 75)
Mammary gland: second left, solid, filled with light green mass, approx. 15x15x10 mm, 1.22 g (no. 92)
Group 3: Uterus: slightly thickened, glazed (no. 130)
Stomach: scattered haemorrhagic foci, approx. 1 - 5 mm in diameter (no. 136)
Group 4: Ovary, left: cystic with clear fluid, approx. 4 mm in diameter (no. 174)

BONE MARROW:
No test item-related differences for the myeloid/erythroid ratio of the bone marrow were noted between the control group and the high dose group (1000 mg Titanium dioxide E171-E/kg b.w./day).

ORGAN WEIGHT:
No test item-related differences for the examined absolute and relative organ weights were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

ACF SCORING:
The analysis of sections of the entire colons showed that the administration of Titanium Dioxide E171-E to rats over a time period of 122 days via the diet did not induce the development of any aberrant crypt foci (ACF) in this species.

HISTOPATHOLOGY:
No test item-related changes were observed for the examined male (n = 20) and female (n = 20) animals of the high dose group (1000 mg Titanium Dioxide E171-E/kg b.w./day) during the microscopic examination.
The histopathological examination performed on one testicle and one epididymis of the examined males (n = 20) of groups 1 and 4 (with special emphasis on the qualitative stages of spermatogenesis (proliferative, meiotic and spermiogenic phases) and histopathology of the interstitial testicular structure), did not reveal any test item-related effects.




REPRODUCTIVE FUNCTION/PERFORMANCE

ESTROUS CYCLE DURING EXPOSURE:
No test item-related differences were noted for the mean length and the mean number of estrous cycles per dam during the pre-mating period between the female animals of the control group and the female animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). None of the females showed a complete estrous cycle during the mating period from test day 85 until a positive mating sign was noted.

ESTROUS CYCLE AT NECROPSY:
A stage of diestrus was most commonly noted for the animals of the control group and the treatment groups. The proportion of the different stages at necropsy between the control group and the treatment groups were nearly equal.

SPERMIOGRAM:
Sperm number: No test item-related difference was noted between the rats of the control group and the rats treated with 100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day for the number of ultrasound-resistant sperm heads (sperm count) per gram testicular tissue.
Sperm motility: No test item-related differences were noted between the rats of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the percentage of motile spermatozoa in the epididymal cauda on the total number of motile and non-motile spermatozoa.
Sperm morphology: The examination of spermatozoa from the epididymal cauda revealed no increased number of spermatozoa with a malformation in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171/kg b.w./day) in comparison to the control group. In detail, the examination of 4000 spermatozoa (200 per animal) from each test group revealed 11 spermatozoa with a malformation in the control group, 6 in the low dose group and 5 in the high dose group. In all cases, the observed malformation was in the form of a banana-like sperm head.

FEMALE FERTILITY:
No test item-related influence on the fertility index of the female rats was noted for any of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). The observation of one non-pregnant female (no. 76) at the low dose group and 2 non-pregnant females each at the intermediate (nos. 125, 144) and the high dose group (nos. 173, 177) was considered to be spontaneous. A fertility index of 90 % is in the range of the historical control data.

GESTATION INDEX:
No test item-related influence on the gestation index was noted for the female rats of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg
b.w./day). One female of the intermediate dose group (no. 126) only delivered stillbirth (2 stillbirths), leading to a gestation index of 95% at the intermediate dose level. However, the singular observation of one female with only stillbirths was considered to be spontaneous.

PRE-COITAL TIME:
No test item-related differences were noted in the length of the pre-coital time between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

GESTATION LENGTH:
No test item-related differences were noted for the length of the gestation period between the rats of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171/kg b.w./day).

PRE-NATAL DEVELOPMENT F1-GENERATION:

BIRTH INDICES AND POST-IMPLANTATION LOSS:
No test item-related differences were noted for the mean number of implantation sites per dam, the mean number of pups born (alive and dead) per dam and the mean number of live born pups per dam between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Also the reproductive indices as the birth index, the live birth index and the percentage of post implantation loss revealed no test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

VIABILITY INDEX:
Pre- and post-cull period:
No test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the viability indices of the pre- and the post-cull period.

SEX RATIO:
No test item-related influence on the male to female ratio was noted for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BODY WEIGHT OF PUPS:
No test item-related influence on the body weight of pups was noted for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).
Two runts were noted at the high dose level (1000 mg Titanium dioxide E171-E/kg b.w./day). This is regarded to be spontaneous.

LITTER WEIGHT:
No test item-related influence on the litter weight was noted for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

NUMBER OF LIVE PUPS:
No test item-related or statistically significant differences were noted for the mean number of live pups per dam between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) during the lactation period.

THYROID HORMONES:
No test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171E/kg b.w./day) were noted for the T4 level on lactation day 4 and the T3, T4 and TSH levels on lactation day 22 for the male and female pups.

EXTERNAL AND INTERNAL EXAMINATION:
External examination (pups terminally sacrificed on lactation day 4 or 22 or found dead):
No gross abnormalities (e.g. malformations or variations) were noted during the macroscopic external examination of the control pups and the pups from the dams treated with 100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day after terminal sacrifice on lactation day 4 (culled pups), on lactation day 22 or for the pups that died (stillborn or found dead) during the lactation period.
Internal examination (pups terminally sacrificed on lactation day 22):
No gross abnormalities (e.g. malformations or variations) were noted during the macroscopic examination of the inner organs or tissues of the control pups and the pups from the dams treated with 100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day after terminal sacrifice on lactation day 22.

PUP ORGAN WEIGHTS (ABSOLUTE):
No test item-related differences for the examined absolute organ weights were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

Key result

Dose descriptor:

NOAEL

Effect level:

>= 1 000 mg/kg bw/day (nominal)

Based on:

test mat.

Sex:

male/female

Basis for effect level:

other: no test item-related effects were observed

Clinical signs:

effects observed, treatment-related

Description (incidence and severity):

COHORT 1B:
At the high dose level 1000 mg Titanium dioxide E171-E/kg b.w./day, test itemrelated changes were noted for all male and female animals in the form of pale faeces consecutively from test day 11 until necropsy. This observation was caused by the incorporated white colour of Titanium dioxide E171-E via the diet and was not considered to be of toxicological relevance.

Dermal irritation (if dermal study):

not examined

Mortality:

no mortality observed

Body weight and weight changes:

no effects observed

Food consumption and compound intake (if feeding study):

no effects observed

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

no effects observed

Ophthalmological findings:

not examined

Haematological findings:

not examined

Clinical biochemistry findings:

not examined

Urinalysis findings:

not examined

Behaviour (functional findings):

not examined

Immunological findings:

not examined

Organ weight findings including organ / body weight ratios:

no effects observed

Gross pathological findings:

no effects observed

Neuropathological findings:

not examined

Histopathological findings: non-neoplastic:

not examined

Histopathological findings: neoplastic:

not examined

Other effects:

no effects observed

Reproductive function: oestrous cycle:

no effects observed

Reproductive function: sperm measures:

not examined

Reproductive performance:

no effects observed

COHORT 1B

As recommended by the corresponding guideline OECD 443 20 animals/sex/dose of Cohort 1B were used for pairing and the establishment of the F2 generation. However, due to the elevated number of non-pregnant females in the mid (3) and in the high dose (3) group, a re-check was performed of the vaginal smears of the non-pregnant females. This additional check showed that no sperm could be found in the vaginal smears from five of six non-pregnant females (only for the non-pregnant female no. 625 the re-check confirmed the sperm detection during the first examination). Hence, the positive mating signal was incorrect for five females and it cannot be ruled out that the animals would have become pregnant during a longer mating period. Therefore, the erroneously considered inseminated non- pregnant females were excluded from the study (two mid dose and three high dose females). Further, no positive mating sign was noted for one low dose female (no. 591) after 14 days of mating. This female was laparotomized on test day 106 and excluded from the study. One mid dose female was not pregnant although a positive mating sign was noted (no. 625, as mentioned above). This non-pregnant female was considered in the calculation of reproductive performance indices but not for other paramters such as general toxicity.
Overall, the reduced number of pregnant females is not considered to be test item-related but a result of technical failure of the testing facility and thus does not invalate the study.


GENERAL TOXICITY

MORTALITY:
No premature death was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the male and female animals.

CLINICAL SIGNS:
No changes in behaviour, the external appearance and the consistency of the faeces were noted for the male and female animals of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

DETAILED CLINICAL OBSERVATION:
No observations were noted during the detailed clinical observations for the animals of the control group and the animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BODY WEIGHT AND BODY WEIGHT GAIN:
MALES: No test item-related changes in body weight and body weight gain were noted for the male rats between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Slight but statistically significantly (p ≤ 0.05) decreased body weights in comparison to the control group were noted on test days 1, 15, 22, 29 and 36 (at maximum 12.6 % below control on test day 1) for the intermediate dose group. As the difference between the intermediate dose group and the control group was only slight, nearly disappeared during the further course of the study and no decreased body weight was noted for the male animals of the high dose group, the difference between the animals of the intermediate dose group and the control animals at the beginning of the study was considered to be spontaneous.
FEMALES: No differences in body weight and body weight gain were noted between the female animals of the control group and the female animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

FOOD CONSUMPTION:
No test item-related difference in food consumption was noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Some minor changes were noted but not considered to be test item-related since no dose response was observed and the changes were only temporary/spontaneous.

WATER CONSUMPTION:
No test item-related difference in water consumption was noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Some minor changes were noted but not considered to be test item-related since no dose response was observed and the changes were only temporary/spontaneous.

TEST ITEM INTAKE:
The actual mean test item intake during the whole study was in the range of the nominal values for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

SEXUAL HORMONES:
MALES: No test item-related differences for the examined sexual hormone levels (estradiol, estrone and testosterone) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Decreased estradiol levels were noted at the intermediate and the high dose level in comparison to the control group, statistically significant at the intermediate dose level. However, though the estradiol concentrations of the intermediate and the high dose group of Cohort 1B were decreased in comparison to the control group of Cohort 1B, they were not unusually low, as the mean values of the intermediate group and the high dose group of Cohort 1B were still above the range of the mean values of the circadian cycle and the values that were noted for the F0 Generation and Cohort 1A and Cohort 2A. Hence, unusually increased estradiol concentrations were noted for the control group and the treatment groups of Cohort 1B when compared to the values of the circadian cycle and the values from the F0 Generation and from Cohort 1A and Cohort 2A. These unusually increased estradiol concentrations were more increased in the control group of Cohort 1B as in the intermediate and the high dose group of Cohort 1B, hence, the decrease that was noted between the control group and the intermediate and the high dose group can be considered as spontaneous and of no toxicological relevance.

FEMALES: No test item-related differences for the examined sexual hormone levels (estradiol, estrone and testosterone) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Statistically significantly increased testosterone levels were noted at the intermediate and the high dose level. However, all values were in the range of or slightly below or above the values of the circadian cycle. It is also worth mentioning that the testosterone concentrations of the Cohort 1B control animals were slightly lower than for the control groups of the F0 Generation and Cohort 1A and 2A. Hence, these variations of the female testosterone levels were considered to be spontaneous.

SEXUAL MATURATION:
MALES: No test item-related differences between the control group and the test item-treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the time point of balanopreputial gland cleavage and no differences were noted for the body weight at the time point of balanopreputial gland cleavage.
FEMALES: No test item-related differences between the control group and the test item-treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the time point of vaginal opening and no differences were noted for the body weight at the time point of vaginal opening.

BODY WEIGHT AT NECROPSY:
No test item-related differences were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

GROSS NECROPSY:
MALES: No observations were noted for the male animals of the low and the intermediate dose group (100 or 300 mg Titanium dioxide E171-E/kg b.w./day) and no test itemrelated observations were noted for the male animals of the high dose group (1000 mg Titanium dioxide E171-E/kg b.w./day). The following isolated observation was considered to be spontaneous:
Group 4: Testes and epididymides (bilateral): small (no. 609).

FEMALES: No test item-related observations were noted for the female animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). The following isolated observations were considered to be spontaneous:
Group 1: Kidney (bilateral: very soft consistency (no. 554)
Group 4 Uterus: dilatation (no. 669 NV: prenatal loss of 8 implants)

ORGAN WEIGHTS:
MALES: No test item-related differences for the examined absolute and relative organ weights were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Changes that were noted for the left testis and the liver of the male animals were considered to be spontaneous since no dose response was noted and and no test item-realted changes during histopathology were observed.

FEMALES: No test item-related differences for the examined absolute and relative organ weights were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).



REPRODUCTIVE PERFORMANCE/FUNCTION- COHORT 1B

ESTROUS CYCLE- MATING PERIOD
The obtained stages of the estrous cycle during the mating period did not show any sign of an impaired mating behaviour as a stage of estrus (E) or a stage of proestrus (P) that were not followed by a positive mating sign. On nearly all days of the mating period a stage of diestrus (D) was noted. A stage of proestrus (P) was noted on a few days and such a stage of proestrus (P) was in nearly all cases followed by a positive mating signal on the following day. Exceptions were only noted for animal no. 591 (group 2) and no. 640 (group 3). A stage of estrus (E), that was not followed by a positive mating signal (failed mating opportunity), was noted for 2 animals of the intermediate dose group (nos. 637 and 640) on one day each. These single observations can be considered as spontaneous.

ESTROUS CYCLE AT NECROPSY:
No test item-related changes in estrous cycle at necropsy were noted in the control group or in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

FEMALE FERTILIY:
No test item-related influence on the fertility index of the female rats was noted for any of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). All females of the control group, the low dose group and the high dose group that were mated with a male partner and showed a positive mating sign (sperm detection) became pregnant, leading to a fertility index of 100% for the control group and the low and the high dose group. Only one female with a positive mating sign did not become pregnant (no. 625 of the intermediate dose group). This single occurrence was considered to be spontaneous. The fertility index of all treatment groups was in the range of the historical control data.

GESTATION INDEX:
No test item-related influence on the gestation index was noted for the female rats of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). One pregnant female of the high dose group (no. 669) did not litter but showed a prenatal loss of all 8 implants, leading to a gestation index of 94%. However, this single occurrence was considered to be spontaneous.

PRE-COITAL TIME:
No test item-related differences were noted in the length of the pre-coital time between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). However, 2 females with an elongated pre-coital time of 14 test days were noted in the control group and 3 females with an elongated pre-coital time of 13 or 14 test days were noted in the intermediate dose group. The observation of 3 females with an elongated pre-coital time in the intermediate dose group was considered to be spontaneous, as 2 females with an elongated pre-coital time we noted in the control group and none in the high dose group.

GESTATION LENGTH:
No test item-related differences were noted for the length of the gestation period between the rats of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BIRTH INDICES AND POST-IMPLANTAION LOSS:
No test item-related differences were noted for the mean number of implantation sites per dam, the mean number of pups born (alive and dead) per dam and the mean number of live born pups per dam between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Also the reproductive indices as the birth index, the live birth index and the percentage of post implantation loss revealed no test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A slightly reduced birth index and consequently a slightly increased post implantation loss in comparison to the control group were noted at the high dose level for the group value (calculated from the whole numbers of implantation sites and pups per group) and the group mean value (mean value from the individual values of implantation sites and pups of each female per group). The group values for the birth index were 89.27% (high dose group) in comparison to 93.59 % for the control group.
For the post-implantation loss, the group value for the high dose group was 10.73 % in comparison to 6.73% for the control group. The reduced birth index and the increased post-implantation loss at the high dose level were due to the high dosed animal no. 669 with a total loss of all its 8 implantation sites, resulting in a birth index of 0.0% and a post-implantation loss of 100% for animal no. 669. However, as the post-implantation loss of all other remaining females of the high dose group was in the range of the animals of the control group, the slightly decreased birth index and the slightly increased post-implantation loss that were noted at the high dose level were considered to be spontaneous.

VIABILITY INDEX:
No test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the viability index (difference between the number of live born pups (pups alive on lactation day 0/1) and the number of live pups on lactation day 4).

SEX RATIO:
No test item-related influence on the male to female ratio was noted for any treatment group (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BODY WEIGHT OF PUPS (until PND 4):
No test item-related influence on the body weight of pups was noted for any treatment group (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) on lactation days 1 and 4. One runt each was noted in the control group and the low and the intermediate dose groups. No runt was noted at the high dose level.

LITTER WEIGHT:
No test item-related influence on the litter weight was noted for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

NUMBER OF LIVE LITTER:
No test item-related or statistically significant differences were noted for the mean number of live pups per dam between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) on lactation days 1 and 4.

EXTERNAL AND INTERNAL EXAMINATION OF THE PUPS:
External examination (pups terminally sacrificed between lactation days 4 to 8 or found dead):
No gross abnormalities (e.g. malformations or variations) were noted during the macroscopic external examination of the F2 pups from the control group and the F2 pups from the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) after terminal sacrifice between lactation days 4 to 8 or for the pups that died (stillborn or found dead) during the lactation period (PND 1 to 4).
Internal examination (pups terminally sacrificed between lactation days 4 to 8):
No gross abnormalities (e.g. malformations or variations) were noted during the macroscopic examination of the inner organs or tissues of the F2 pups from the control group and the F2 pups from the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) after terminal sacrifice between lactation days 4 to 8.

Key result

Dose descriptor:

NOAEL

Effect level:

>= 1 000 mg/kg bw/day (nominal)

Based on:

test mat.

Sex:

male/female

Basis for effect level:

other: no test item-related effects were observed

Clinical signs:

effects observed, treatment-related

Description (incidence and severity):

COHORT 1A:
At the high dose level (1000 mg Titanium dioxide E171-E/kg b.w./day) test itemrelated changes were noted for all male and female animals in the form of pale faeces caused by the white colour of Titanium dioxide E171 in the diet from test day 11 until sacrifice. This observation was not considered to be of toxicological relevance.

COHORT 2A:
At the high dose level (1000 mg Titanium dioxide E171-E/kg b.w./day) test itemrelated changes were noted for all male and female animals in the form of pale faeces caused by the white colour of Titanium dioxide E171-E in the diet from test day 11 until sacrifice on test days 64, 65 or 66. This observation was not considered to be of toxicological relevance.

COHORT 3 :
At the high dose level (1000 mg Titanium dioxide E171-E/kg b.w./day) test item related changes were noted for all animals in the form of pale faeces caused by the white colour of Titanium dioxide E171-E in the diet from test day 11 until sacrifice on test days 32, 33, 36 or 37 (postnatal days 53 - 61). This observation was not considered to be of toxicological relevance.

Dermal irritation (if dermal study):

not examined

Mortality / viability:

no mortality observed

Description (incidence and severity):

For Cohort 1A, 2A, 2B and 3

Body weight and weight changes:

no effects observed

Description (incidence and severity):

For Cohort 1A, 2A and 3.
Cohort 2B was sacrificed directly after weaning, but no changes were observed until weaning.

Food consumption and compound intake (if feeding study):

no effects observed

Description (incidence and severity):

For Cohort 1A, 2A and 3.
Cohort 2B was sacrificed directly after weaning, no food consumption was recorded.

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

not examined

Ophthalmological findings:

not examined

Haematological findings:

no effects observed

Description (incidence and severity):

For Cohort 1A (other Cohorts were not analysed)

Clinical biochemistry findings:

no effects observed

Description (incidence and severity):

For Cohort 1A (other Cohorts were not analysed)

Urinalysis findings:

no effects observed

Description (incidence and severity):

For Cohort 1A (other Cohorts were not analysed)

Sexual maturation:

no effects observed

Description (incidence and severity):

For Cohort 1A, 2A and 3.
Cohort 2B was sacrificed directly after weaning, no sexual maturation was recorded.

Anogenital distance (AGD):

no effects observed

Nipple retention in male pups:

no effects observed

Organ weight findings including organ / body weight ratios:

no effects observed

Description (incidence and severity):

For Cohort 1A
For Cohort 2A and 2B only brain weights were recorded.
Organ weight of Cohort 3 were not recorded.

Gross pathological findings:

no effects observed

Description (incidence and severity):

For Cohort 1A, 2A, 2B and 3

Histopathological findings:

no effects observed

Description (incidence and severity):

For Cohort 1A only.
Full histopatholocial examination was not conducted on Cohort 2A, 2B and 3 animals.
Neurohistopathology was conducted on Cohort 2A and 2B animals, no changes were observed.

Other effects:

no effects observed

Behaviour (functional findings):

no effects observed

Description (incidence and severity):

Cohort 2A only

Developmental immunotoxicity:

no effects observed

Description (incidence and severity):

Cohort 3 only

ANOGENITAL DISTANCE OF F1 PUPS
No test item-related difference in the absolute and the relative ano-genital distance (value normalized to cube root of pup body weight) of the male and the female pups was noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

NIPPLE RETENTION OF F1 PUPS:
No test item-related difference in the number of nipples was noted between the male pups of the control group and in the male pups of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).


COHORT 1A

For the assessment of general toxicity 20 animals/sex/dose were used. However, for laboratory examinations (such as haematology, clinical biochemistry and lymphocyte typing) only 10 animals/sex/dose were used. For necropsy, histopathology and the determination of organ weights all 20 animals/sex/dose were used.

MORTALITY:
No premature death was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the male and female animals.

CLINICAL SIGNS
No changes in behaviour, the external appearance and the consistency of the faces were noted for the male and female animals of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BODY WEIGHT AND BODY WEIGHT GAIN:
MALES: No test item-related changes in body weight and body weight gain were noted for the male rats between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). However, a slightly reduced body weight was noted for the male animals of the intermediate dose group during the whole study period from test day 1 (10.4% below the control) until test day 64 (5.6% below the control). The difference was statistically significant on test days 36 and 64 (6.1% or 5.6% below the control; p ≤ 0.05). However, as the difference between the intermediate and the control group slightly decreased during the course of the study and only marginal differences were noted between the high dose group and the control group, the differences in body weight that were noted between the intermediate and the control group were considered to be spontaneous.
FEMALES:
No differences in body weight and body weight gain were noted between the female animals of the control group and the female animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

FOOD CONSUMPTION:
MALES: No test item-related difference in food consumption was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A slightly increased food consumption was noted at the high dose level between test week 1 and test week 6 (p ≤0.05/0.01). At maximum 14.0% above the control in test week 1. This slight increase was considered to be spontaneous.
FEMALES: No test item-related difference in food consumption was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

TEST ITEM INTAKE:
The actual mean test item intake during the whole study was in the range of the nominal values for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

HAEMATOLOGY:
No test item-related differences for the examined haematological parameters were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). However, statistically significant changes in comparison to the control group were noted for the concentration of neutrophilic granulocytes of the male animals, which were considered to be spontaneous, as all values were still in the range of historical control data.

CLINICAL BIOCHEMISTRY:
No test item-related differences for the examined biochemical parameters were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). However, slightly increased sodium / potassium ratios were noted for the female animals at the intermediate and the high dose level, which were due to decreased potassium levels (statistically not significant). However, as the difference was only small, no dose-response relationship was noted and no changes were noted for the male animals, the increased sodium / potassium ratio at the intermediate and the high dose level was considered to be spontaneous. Furthermore, all values were in the range of historical control data, and no changes were seen in the sodium levels.

LYMPHOCYTE TYPING IN SPLEEN:
No test item-related influence was noted in the proportion of the examined lymphocyte subtypes to each other between the male and female animals of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

URINALYSIS:
No test item-related differences for the examined urinalysis parameters were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Increased values (statistically significant or not) were noted for the female animals of all treatment groups for the pH value and the urine volume. However, as no dose-response relationship was noted, no changes were noted for the male animals and all values were in the range of historical control data, the differences that were noted for the female animals were considered to be spontaneous.

SEXUAL HORMONES:
MALES: No test item-related differences for the examined sexual hormone levels (estradiol, estrone and testosterone) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Increased estrone levels in comparison to the control group were noted at the intermediate and the high dose level, statistically significant at the high dose level. Any increased estrone levels were considered to be spontaneous and within circadian variation.
FEMALES: No test item-related differences for the examined sexual hormone levels (estradiol, estrone and testosterone) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

THYROID HORMONE LEVELS:
MALES: No test item-related differences for the examined thyroid hormone levels (T3, T4 and TSH) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).
FEMALES: No test item-related differences for the examined thyroid hormone levels were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A statistically significantly decreased T3 level in comparison to the control group was noted at the low dose level. However, this slight decrease was considered to be spontaneous as no dose-response relationship was noted. Furthermore, all mean values and individual values were in the range of the circadian cycle.
Markedly increased values were noted for the TSH concentrations at the low, the intermediate and the high dose group, although they were not statistically significant. The lack of statistically significance was due to a high variability of the individual values. Only a few values were above the upper range of the individual values of the circadian cycle. However, as no statistical significance was noted, the few increased values were considered to be spontaneous.

SEXUAL MATURATION:
MALES: No test item-related differences between the control group and the test item-treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the time point of balanoprepuital gland cleavage and the body weight at the time point of balanoprepuital gland cleavage.
FEMALES: No test item-related differences between the control group and the test item-treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the time point of vaginal opening and the body weight at the time point of vaginal opening.
No test item-related influence was noted on the time point for the appearance of cornified cells in the vaginal smear and no test item-related influence was noted on the period between the day of vaginal opening and the day of the appearance of cornified cells in the vaginal smear for the females of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

ESTROUS CYCLE- 2-WEEK PERIOD:
No test item-related differences were noted for the mean length and the mean number of estrous cycles per female animal between the females of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

SPERMIOGRAM:
Sperm number: No test item-related difference was noted between the rats of the control group and the rats treated with 100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day for the number of ultrasound-resistant sperm heads (sperm count) per gram testicular tissue.
Sperm motility: No test item-related differences were noted between the rats of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the percentage of motile spermatozoa in the epididymal cauda on the total number of motile and non-motile spermatozoa.
Sperm morphology: The examination of spermatozoa from the epididymal cauda revealed no increased number of spermatozoa with a malformation in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) in comparison to the control group. In detail, the examination of 4000 spermatozoa (200 per animal) from each test group revealed no malformation.

BODY WEIGHT AT NECROPSY:
No test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the body weight at autopsy.

GROSS NECROPSY:
No observations were noted in the control group and no test item-related observations were noted in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).
The following observations were considered to be spontaneous:
Group 2 (males): Lungs: reddened (no. 410)
Group 2 (females): Stomach: scattered haemorrhagic foci, approx. 0.5 – 10 mm in diameter (no. 421)
Group 2 (females): Kidney, left: few light brownish foci, approx. 3.0 - 8.0 mm in diameter (no. 431)

ESTROUS CYCLE AT NECROPSY:
The examination at necropsy revealed an increased number of females in the high dose group that were in an estrous stage and a decreased number of females that were in a diestrous stage in comparison to the control group and the low and the intermediate dose group. However, this difference markedly decreased when comparing the results from the histopathological examination. In detail, during the histopathological examination 11 females with a diestrous stage were noted in the control group in comparison to 8 in the high dose group. Seven females of the control group were in an estrous or a proestrous stage in comparison to 11 in the high dose group. Hence, the difference that was noted between the control group and the high dose group during the examination at necropsy could be considered as irrelevant, as the more accurate histopathological examination revealed only small differences.

ORGAN WEIGHTS:
MALES: No test item-related differences for the examined absolute and relative organ weights were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day), with the following exceptions: Only slightly but statistically significantly reduced absolute organ weights were noted for the right kidney and the right testes at the intermediate and the high dose level. However, as the differences were only slight and statistically significantly reduced values were not noted for the absolute organ weights of the left kidney and the left testes, these changes were considered to be spontaneous and not of toxicological relevance.
FEMALES: No test item-related differences for the examined absolute and relative organ weights were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Decreased absolute and relative organ weights (statistically significant or not) in comparison to control were noted for the mesenteric lymph node of all treatment groups. However, the histopathological examination of the mesenteric lymph node of the high dose females revealed no test item-related changes. Furthermore, no dose response-relationship was noted (the percentage of organ weight decrease in comparison to control was equal in all treatment groups). Hence, the observed decrease in the absolute and relative organ weights was considered to be spontaneous.

BONE MARROW:
No test item-related differences for the myeloid/erythroid ratio of the bone marrow were noted between the control group and the high dose group (1000 mg Titanium dioxide E171-E/kg b.w./day).

HISTOPATHOLOGY:
Selected organs form all male and female animals of the control group and the high dose group (1000 mg Titanium Dioxide E171-E/kg b.w./day) (20 animals per sex and group) were examined microscopically. These examinations did not reveal any test item-related changes.
Histopathological examination performed on one testicle and one epididymis (with special emphasis on the qualitative stages of spermatogenesis (proliferative, meiotic and spermiogenic phases)and histopathology of the interstitial testicular structure) of all males of groups 1 and 4 (20 males per group), did not reveal any test itemrelated effects.
The examination of one ovary each from 20 females of the control group and 20 females of the high dose group revealed no test item-related differences in the number of primordial and small growing follicles and the number of corpora lutea.


COHORT 2A

All examinations were performed with 10 animals/sex/dose.

MORTALITY:
No premature death was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the male and female animals.

CLINICAL SIGNS:
No changes in behaviour, the external appearance and the consistency of the faeces were noted for the male and female animals of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BODY WEIGHT AND BODY WEIGHT GAIN:
MALES: No test item-related changes in body weight and body weight gain were noted for the male rats between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).
FEMALES: No differences in body weight and body weight gain were noted between the female animals of the control group and the female animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A slightly decreased body weight (statistically not significant) was noted for the female animals of the low dose group (8.0% below the control on test day 57). However, as no differences were noted between the control group and the intermediate and the high dose group, the decreased values that were noted at the low dose level were considered to be spontaneous.

FOOD CONSUMPTION:
MALES: No test item-related difference in food consumption was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A statistically significantly increased food consumption was noted at the intermediate and the high dose level between test days 8 to 15 (23.3 % or 22.5 % above the control; p ≤ 0.01). As the observation was only temporary, it was considered to be spontaneous. An additional statistically significant difference in the form of a temporary decreased food consumption was noted at the low dose level between test days 43 to 50 (6.7 % below the control, p ≤ 0.05). This observation was also considered to bespontaneous, due to its small and temporary nature.
FEMALES: No test item-related difference in food consumption was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Statistically significantly increased values were noted between test days 8 and 15 at the low and the intermediate dose level (24.6 % or 14.8 % above the control; p ≤ 0.05). A statistically significantly decreased value was noted between test days 22 to 29 at the high dose level (27.5 % below the control; p ≤ 0.01). As both observations were only temporary, they were considered to be spontaneous.

TEST ITEM INTAKE:
The actual mean test item intake during the whole study was in the range of the nominal values for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

SEXUAL HORMONES:
No test item-related differences for the examined sexual hormone levels (estradiol, estrone and testosterone) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the male and female animals. Differences (statistically not significant) between the mean values of the test groups and the control group were noted for the estrone and testosterone concentrations of the male animals and the estradiol concentration of the female animals. However, all individual concentrations of estradiol of the female animals were within the circadian variation. In case of the testosterone concentration of the male animals, one or a few individual values per group were slightly below the circadian variation, whereas all other values were within the circadian variation (the circadian variation of estrone was not examined). As the differences in the estrone, the testosterone and the estradiol concentrations were not statistically significant and completely (estradiol concentration) or nearly completely (testosterone concentration) within the circadian variation, they were considered to be spontaneous.

SEXUAL MATURATION:
MALES: No test item-related differences between the control group and the test item-treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the time point of balanopreputial gland cleavage and the body weight at the time point of balanopreputial gland cleavage.
FEMALES: No test item-related differences between the control group and the test item-treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the time point of vaginal opening and the body weight at the time point of vaginal opening.

BODY WEIGHT AT AUTOPSY:
No test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the body weight at autopsy.

GROSS NECROPSY:
No observations were noted for the male and female animals of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BRAIN WEIGHTS:
No test item-related differences for the absolute and relative brain weights were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the male and female animals.

NEUROHISTOPATHOLOGY:
The neurohistopathological examination of different brain regions from all male and female animals of the high dose group (1000 mg Titanium dioxide E171-E/kg b.w./day) did not reveal any test item-related effects when compared to the male and female animals of the control group (n = 10 animals per sex and group). The results of the quantitative evaluation (comparison of the relative size of different brain regions) of groups 1 and 4 was still ongoing at the time of preparation of this robust study summary, but will be given in an update.

SCREENING FOR NEURODEVELOPMENTAL EFFECTS

AUDITORY STARTLE RESPONSE:
No test item-related difference was noted for the reaction of the animals in the auditory startle response test between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day), when exposed to a short noise stimulus on postnatal days 23, 24 or 25. The reaction was noted as yes / no answer.
At the beginning all animals react with a short startle to nearly every sound stimulus. During the course of the experiment, the animals adapt to the stimuli. At the end of the experiment the animals showed no reaction to the stimulus in more than half of the cases. This course of the experiment was noted for in the control group and in the treatment groups.

OBSERVATIONAL SCREENING:
No test item-related observations of abnormal behaviour, no adverse effects on motoric skills, changes in the external appearance and the appearance of the faces were noted for the male and female animals of all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). Furthermore, no test item-related differences were noted in body temperature or the hind-leg splay in comparison to the control group. A statistically significantly increased hind-leg splay was noted for the female animals at the low and the high dose level. As no dose-response relationship was seen and no changes were noted for the male animals, the increased values at the low and the high dose level were considered to be spontaneous.

FUNCTIONAL TESTS:
GRIP STRENGTH
No test item-related influence on the fore- and hindlimb grip strength was noted for the male and female animals of all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A statistically significantly increased forelimb grip strength was noted for the male and the female animals of the intermediate dose group. However, as the values of the high dose group were again in the range of the control group (no dose-response relationship) the increased values at the intermediate dose level were considered to be spontaneous.

SPONTANEOUS MOTILITY:
No test item-related influence on the spontaneous motility was noted for the male and female animals of all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).


COHORT 2B

All examinations were performed with 10 animals/sex/dose.

BODY WEIGHT:
No test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the body weight of the male and female animals of Cohort 2B on their lactation day 21. Slight and statistically not significantly decreased body weights were noted for the male animals of each treatment group (between 10.4% below the control in the low dose group and 7.5% below the control in the high dose group). For the female animals a decreased body weight was noted for the animals of the intermediate dose group (7.6% below the control, statistically not significant). However, regarding the other cohorts, slightly reduced body weights in comparison to the control group were noted for the male and female animals of the intermediate and the high dose group in most cases on test day 1 of the F1 Generation study. In all these cases, the difference between the treatment groups and the control group decreased or even completely disappeared during the course of the study. Therefore, the differences in body weight that were noted on test day 1 for the different cohorts were considered to be spontaneous. Thus, it can be assumed that also the differences in body weight that were noted for the animals of cohort 2B on lactation day 21 (which was only a few days before test day 1 of the other F1 cohorts) would have been disappeared during the course of the study.

SEXUAL HORMONES:
No test item-related differences for the examined sexual hormone levels (estradiol, estrone and testosterone) were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the male and female animals. A decreased testosterone concentration was noted for the animals of the intermediate dose group. However, as no dose-response relationship was noted, the observation was considered to be spontaneous and not test item-related. Furthermore, nearly all individual values were within the circadian variation, only a few individual values form group 3 and one value from group 2 were slightly below the lower range of the circadian cycle.

BODY WEIGHT AT AUTOPSY:
No test item-related differences between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the body weight at autopsy. Slight but statistically significantly decreased body weights at autopsy were noted for the male and the female animals of the treatment groups, statistically significant for the female animals of the intermediate dose group. However, the differences corresponded to the differences that were noted for the last live body weight from lactation day 21, which was considered to be spontaneous (see above). Hence, the differences in the body weight at autopsy could also be considered as spontaneous.

GROSS NECROPSY:
No observations were noted for the male and female animals of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BRAIN WEIGHTS:
No test item-related differences for the absolute and relative brain weights were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) for the male and female animals. Slightly increased relative brain weights were noted for the male and female animals, statistically significant for the female animals of the intermediate dose group. However, no changes were noted for the absolute brain weights. Hence, the observed differences in the relative brain weight were only due to the differences that were noted for the body weight at autopsy and no dose response could be observed. Therefore, the observed differences in the relative brain weight were of no toxicological relevance.

NEUROHISTOPATHOLOGY:
The neurohistopathological examination of different brain regions from all male and female animals of the high dose group (1000 mg Titanium dioxide E171-E/kg b.w./day) did not reveal any test item-related effects when compared to the male and female animals of the control group (n = 10 animals per sex and group).


COHORT 3

All examinations were performed with 10 animals/sex/dose.

MORTALITY:
No premature death was noted in the control group and in the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

CLINICAL SIGNS:
No changes in behaviour, the external appearance and the consistency of the faeces were noted for the male and female animals of the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

BODY WEIGHT AND BODY WEIGHT GAIN:
MALES: No test item-related changes in body weight and body weight gain were noted for the male rats between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) between test day 1 and test day 29 (test day 29 was the last day of body weight measurement for the complete groups). However, a slightly reduced body weight (statistically not significant) was noted for the male animals of the intermediate and the high dose group (at maximum 10.3 % or 11.3 % below the control on test day 1). During the further course of the study the difference between the intermediate and the control group slightly decreased (7.6 % below the control on test day 29). In contrast to the intermediate dose group, the difference between the high dose group and the control group even nearly completely disappeared during the course of the study (1.4 % below control on test day 29). As the difference at the high dose level nearly completely disappeared, the reduced body weights that were noted at the beginning of the study for the intermediate and the high dose level were considered to be spontaneous and not test item-related.
FEMALES: No differences in body weight and body weight gain were noted between the female animals of the control group and the female animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). As for the male animals, reduced body weights in comparison to the control were noted for the animals of the intermediate (10.3 % below the control) and the high dose group (16.3 % below the control; p ≤ 0.01) on test day 1. As the differences in comparison to the control group nearly completely disappeared during the course of the study (1.5 % or 0.3 % below the control on test day 29), the differences that were noted at start of the study, were considered to be spontaneous and not test item-related.

FOOD CONSUMPTION:
MALES: No test item-related differences in food consumption were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A slightly but statistically significantly (p ≤ 0.01) increased food consumption was noted at the high dose level for test week 3 (test days 15 to 22) and for test week 4 (test days 22 to 29) (at maximum 12.4 % above the control in test week 3). This slight increase was considered to be spontaneous.
FEMALES: No test item-related differences in food consumption were noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). A statistically significantly increased food consumption (17.1 % above the control; p ≤ 0.01) was noted at the high dose level on test day 1. This temporary observation was considered to be spontaneous.

TEST ITEM INTAKE:
The actual mean test item intake during the whole study was in the range of the nominal values for all treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).

SEXUAL MATURATION:
MALES: No test item-related differences between the control group and the test item-treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the time point of balanoprepuital gland cleavage and the body weight at the time point of balanoprepuital gland cleavage.
FEMALES: No test item-related differences between the control group and the test item-treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) were noted for the time point of vaginal opening and the body weight at the time point of vaginal opening.

GROS NECROPSY:
No test item-related observations were noted for the male and female animals of the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day). The following singular observations were considered to be spontaneous:
- Dilatation of the renal pelvis (male no. 865; control)
- Small spleen (male no. 904; high dose group)

ESTROUS CYCLE AT NECROPSY:
No test item-realted changes in estrous cycle at necropsy were noted of the control and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day).


SCREENING FOR DEVELOPMENTAL IMMUNOTOXIC EFFECTS

LYMPHOCYTE TYPING IN SPLEEN:
No differences in the relative size of the lymphocyte subpopulations were noted between the control group and the animals treated orally with 100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day, after immunization of the animals of the control group and the treatment groups with KLH. However, when compared to animals of cohort 1A, the KLH injected animals of cohort 3 revealed a shift towards a higher percentage of B cells in the lymphocyte population of the spleen. This shift in the lymphocyte subpopulation indicates the successful activation of the immune system by injection of KLH with increased B-cell proliferation to produce the antigen-specific antibodies. As this shift towards the B cell population was equally noted for all animals of cohort 3 (animals of the control group and the treatment groups) this can be taken as a reference that the immune response was not adversely affected by the test item.

T-CELL DEPENDENT KLH RESPONSE (KLH ASSAY):
No test item-related difference was noted for the anti-KLH IgM serum levels between the male and the female animals of the control group and the test-item treated groups (100, 300 or 1000 mg Titanium dioxide E171-E/kg b.w./day) after treatment of the animals from group 1 to 4 with KLH.

It was decided that the KLH immunisation was not successful and thus no final conclusion on possible effects of Titanium dioxide E171-E on the developing immune system can be provided. Thus, it was decided to repeat this part of the study. In early 2021 parental animals will be dosed for 10 weeks with 0, 100, 300 and 1000 mg/kg bw/day, then paired and the F1 generation will be used to investigate possible developmental immunotoxic effects. Final results will be provided with the next dossier update.

Key result

Dose descriptor:

NOAEL

Generation:

F1 (cohort 1A)

Effect level:

>= 1 000 mg/kg bw/day (nominal)

Based on:

test mat.

Sex:

male/female

Basis for effect level:

other: no test item-related effects were observed

Key result

Dose descriptor:

NOAEL

Generation:

F1 (cohort 2A)

Effect level:

>= 1 000 mg/kg bw/day (nominal)

Based on:

test mat.

Sex:

male/female

Basis for effect level:

other: no test item-related effects were observed

Key result

Dose descriptor:

NOAEL

Generation:

F1 (cohort 2B)

Effect level:

>= 1 000 mg/kg bw/day (nominal)

Based on:

test mat.

Sex:

male/female

Basis for effect level:

other: no test item-related effects were observed

Key result

Dose descriptor:

NOAEL

Generation:

F1 (cohort 3)

Basis for effect level:

other: No final conclusion of effects of TiO2 E171-E available. Will be provided with the next dossier update.

Remarks on result:

other: No final conclusion of effects of TiO2 E171-E available. Will be provided with the next dossier update.

Clinical signs:

not examined

Dermal irritation (if dermal study):

not examined

Mortality / viability:

not examined

Body weight and weight changes:

not examined

Food consumption and compound intake (if feeding study):

not examined

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

not examined

Ophthalmological findings:

not examined

Haematological findings:

not examined

Clinical biochemistry findings:

not examined

Urinalysis findings:

not examined

Sexual maturation:

not examined

Anogenital distance (AGD):

no effects observed

Nipple retention in male pups:

not examined

Organ weight findings including organ / body weight ratios:

not examined

Gross pathological findings:

no effects observed

Histopathological findings:

not examined

Other effects:

not examined

Behaviour (functional findings):

not examined

Developmental immunotoxicity:

not examined

ANOGENITAL DISTANCE:
No test item-related difference in the absolute and the relative ano-genital distance (value normalized to cube root of pup body weight) of the male and the female pups was noted between the control group and the treatment groups (100, 300 or 1000 mg Titanium dioxide E171/kg b.w./day). A slight but statistically significantly decreased relative ano-genital distance was noted at the low dose level for the male pups. However, the slight decrease was considered to be spontaneous.

For further results for the F2 generation please refer to the field "Details on results (P1)".

Key result

Dose descriptor:

NOAEL

Generation:

F2 (cohort 1B)

Effect level:

>= 1 000 mg/kg bw/day (nominal)

Based on:

test mat.

Sex:

male/female

Basis for effect level:

other: no test item-related effects were observed

Key result

Reproductive effects observed:

no

Conclusions:

F0 Generation
General toxicity: NOAEL above 1000 mg Titanium dioxide E171-E/kg b.w./day via the diet.
Reproductive toxicity: NOAEL above 1000 mg Titanium dioxide E171-E/kg b.w./day via the diet.

F1 Generation
Reproductive and developmental toxicity (Cohort 1A and 1B): NOAEL above 1000 mg Titanium dioxide E171-E/kg b.w./day via the diet.
Developmental neurotoxicity (Cohort 2A and 2B): NOAEL above 1000 mg Titanium dioxide E171-E/kg b.w./day via the diet.
Developmental immunotoxicity (Cohort 3): needs to be determined (early 2021)

Executive summary:

The extended one generation reproduction (EOGRT) study was commissioned by the Titanium Dioxide Manufactures Association (TDMA) in response to a request by the European Food Safety Authority (EFSA) as set forth in their Scientific Opinion of June 2016 (EFSA, 2016) on the re-evaluation of titanium dioxide (E171) as a food additive.

During the fully guideline compliant extended one generation reproductive toxicity study (OECD 443) 20 male and 20 female rats were dosed with 0, 100, 300 and 1000 mg/kg bw/day of food-grade Titanium Dioxide E171-E via the diet for 10 weeks before mating. After mating, all males were analysed for general toxicity. After weaning of F1 animals, all parental females were sacrificed and analysed for reproductive and gerneral toxicity and all F1 pups were allocated to Cohort 1A, 1B, 2A, 2B and 3 for the assessment of reproductive/developmental endpoints (cohort 1A and 1B), potential impact on the developing nervous system (Cohort 2A adult rats, Cohort 2B weaning rats) and potential impact on the developing immune system (Cohort 3). Cohort 1B animals were maintained on treatment until postnatal day 90 and bred to obtain the F2 generation. The F2 generation was analysed for developmental toxicity on postnatal day 7.

In addition to that, 10 male and 10 female F0 satellite animals per group were paired and sacrificed after weaning and used for evaluation of aberrant crypt foci. Further F0 satellite animals were used for hormone analysis of beta-estradiol, Testosterone, T3, T4 and TSH (group 1 and 4 only). This analysis was performed twice, pre-dose and at the end of the pre-mating period (after 10 weeks) throughout a 24-hour cycle. Blood sampling was performed every 2 hours during the 24-hour period of the cycle.

Further, blood samples were taken at sacrifice from F0, Cohort 1A, 1B, 2A and 2B animals for the determination of sexual hormones (estradiol, testosterone and estrone) and whole blood and urine samples taken at sacrifice from F0, Cohort 1A, 1B and F2 pups were used for titanium level analysis.

No test item-related influence was noted on the general toxicity and the reproductive performance of the parental animals of the F0 Generation as well as on the pre- and postnatal development of the F1 pups. No test item-related changes were noted during the histopathological examination (including a detailed examination of the testis and the epididymides) and the examination of the intestines for aberrant crypt foci (ACF). There were no treatment-related effects on hormone levels (estradiol, estrone and testosterone, plus T3, T4 and TSH) in any of the treatment groups of each cohort compared to controls. During their post-weaning development, the animals of the F1 Generation showed no signs of general toxicity. No test item-related influence was noted on the development of the reproductive system (levels of sexual hormones, time points of sexual maturations, number and length of estrous cycles, sperm parameter, detailed histopathological examination of testis and epididymides, number of primordial and growing follicles and number of corpora lutea in the ovaries). Also no test item-related influence was noted on the reproductive performance of the F1 males and females (fertility index, gestation index, pre-coital time and gestation length) and on the pre- and postnatal development of the F2 pups until sacrifice on lactation day 4 (number of resorptions, stillborns, live born pups and the viability index after birth until lactation day 4-7). No test item-related influence was noted on the neurological function of the young adult male and female animals of cohort 2A. The neurohistopathological examination of the brains from the high dosed adult animals of cohort 2A and from the high dosed recently weaned animals of cohort 2B did not reveal any test item-related effects when compared to their control group. The examination of the lymphpcyte subpopulations in the spleen and the anti KLH IgM serum levels was not successful, thus complete cohort 3 will be repeated in early 2021. Final results on cohort 3 will be provided with the next dossier update. The analysis of titanium in whole blood and urine is also still ongoing and will be reported with the next dossier update.

Effect on fertility: via oral route

Endpoint conclusion:

no adverse effect observed

Study duration:

chronic

Species:

rat

Effect on fertility: via inhalation route

Endpoint conclusion:

no adverse effect observed

Study duration:

chronic

Species:

rat

Effect on fertility: via dermal route

Endpoint conclusion:

no study available

Additional information

Toxicity to reproduction

 

An extended one generation reproduction (EOGRT) study according to OECD 443 (2018) was commissioned by the Titanium Dioxide Manufactures Association (TDMA) in response to a request by the European Food Safety Authority (EFSA) as set forth in their Scientific Opinion of 28 June 2016 ((EFSA Journal 2016;14(9):4545)) on the re-evaluation of titanium dioxide (E171) as a food additive. This study was requested in order to establish a health-based guidance value (ADI) for the food additive titanium dioxide (EFSA Journal 2016;14(9):4545). The study is further described below.

 

During the fully guideline compliant extended one generation reproductive toxicity study (OECD 443) 20 male and 20 female rats were dosed with 0, 100, 300 and 1000 mg/kg bw/day of food-grade Titanium Dioxide E171-E via the diet for 10 weeks before mating. After mating, all males were analysed for general toxicity. After weaning of F1 animals, all parental females were sacrificed and analysed for reproductive and general toxicity and all F1 pups were allocated to Cohort 1A, 1B, 2A, 2B and 3 for the assessment of reproductive/developmental endpoints (cohort 1A and 1B), potential impact on the developing nervous system (Cohort 2A adult rats, Cohort 2B weaning rats) and potential impact on the developing immune system (Cohort 3). Cohort 1B animals were maintained on treatment until postnatal day 90 and bred to obtain the F2 generation. The F2 generation was analysed for developmental toxicity on postnatal day 7.

In addition to that, 10 male and 10 female F0 satellite animals per group were paired and sacrificed after weaning and used for evaluation of aberrant crypt foci. Further F0 satellite animals were used for hormone analysis of beta-oestradiol, Testosterone, T3, T4 and TSH (group 1 and 4 only). This analysis was performed twice, pre-dose and at the end of the pre-mating period (after 10 weeks) throughout a 24-hour cycle. Blood sampling was performed every 2 hours during the 24-hour period of the cycle.

Further, blood samples were taken at sacrifice from F0, Cohort 1A, 1B, 2A and 2B animals for the determination of sexual hormones (oestradiol, testosterone and estrone) and whole blood and urine samples taken at sacrifice from F0, Cohort 1A, 1B and F2 pups were used for titanium level analysis.

No test item-related influence was noted on the general toxicity and the reproductive performance of the parental animals of the F0 Generation as well as on the pre- and postnatal development of the F1 pups. No test item-related changes were noted during the histopathological examination (including a detailed examination of the testis and the epididymides) and the examination of the intestines for aberrant crypt foci (ACF). There were no treatment-related effects on hormone levels (oestradiol, estrone and testosterone, plus T3, T4 and TSH) in any of the treatment groups of each cohort compared to controls. During their post-weaning development, the animals of the F1 Generation showed no signs of general toxicity. No test item-related influence was noted on the development of the reproductive system (levels of sexual hormones, time points of sexual maturations, number and length of oestrous cycles, sperm parameter, detailed histopathological examination of testis and epididymides, number of primordial and growing follicles and number of corpora lutea in the ovaries). Also no test item-related influence was noted on the reproductive performance of the F1 males and females (fertility index, gestation index, pre-coital time and gestation length) and on the pre- and postnatal development of the F2 pups until sacrifice on lactation day 4 (number of resorptions, stillborns, live born pups and the viability index after birth until lactation day 4-7). No test item-related influence was noted on the neurological function of the young adult male and female animals of cohort 2A. The neurohistopathological examination of the brains from the high dosed adult animals of cohort 2A and from the high dosed recently weaned animals of cohort 2B did not reveal any test item-related effects when compared to their control group. The examination of the lymphocyte subpopulations in the spleen and the anti KLH IgM serum levels was not successful, thus complete cohort 3 will be repeated in early 2021. Final results on cohort 3 will be provided with the next dossier update. The analysis of titanium in whole blood and urine is also still ongoing and will be reported with the next dossier update.

 

In conclusion, the following No-observed-effect-levels (NOAELs) were determined based on the results:

 

F0 Generation:

General toxicity: NOAEL above 1000 mg Titanium dioxide E171-E/kg bw/day via the diet.

Reproductive toxicity: NOAEL above 1000 mg Titanium dioxide E171-E/kg bw/day via the diet.

 

F1 Generation

Reproductive and developmental toxicity (Cohort 1A and 1B): NOAEL above 1000 mg titanium dioxide E171-E/kg bw/day via the diet.

Developmental neurotoxicity (Cohort 2A and 2B): NOAEL above 1000 mg Titanium dioxide E171-E/kg bw/day via the diet.

Developmental immunotoxicity (Cohort 3): needs to be determined (early 2021)

 

Furthermore, during a comprehensive literature search for information on reproduction toxicity of TiO2, eight references were identified representing studies on toxicity to reproduction, conducted either in mice or rats receiving ultrafine or pigment-grade titanium dioxide via oral (gavage or diet), inhalation (whole body), subcutaneous or intravenous administration. After a thorough reliability screening, all of these references were considered of very limited relevance for hazard assessment purposes. The criteria for quality, reliability and adequacy of experimental data under REACH and for hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X) are not fulfilled. These references are discussed below, highlighting their findings and the reasons for their exclusion from hazard assessment:

 NIER (2009) describes a study which evaluated the reproductive and developmental toxicity of titanium dioxide according to OECD 421. Groups of 10 male and 10 female Sprague-Dawley rats were treated with 1,000 mg TiO2/kg bw/day in 1 % methylcellulose via gavage. According to the authors, no treatment-related changes were observed in the parental generation regarding general toxicity and reproductive performance. Also, the authors stated that no treatment-related changes were observed in the F1 generation. The NOAEL of titanium dioxide was considered to be above ,1000 mg/kg bw/day. The reference is however only available as a study summary without raw data and was therefore rated as not assignable (RL=4). Whereas the original study was not obtainable, it was nevertheless assessed and peer-reviewed within the OECD SIDS initial assessment report (OECD CoCAM4, April 2013) and approved by the OECD procedure on Mutual Acceptance of Data (MAD). Thus, the study is considered useful as supporting information.

 

Engler-Chiurazzi et al. (2016) investigated the effects of titanium dioxide nanoparticles on the behavioural and cognitive function of male Sprague-Dawley rats (20 weeks old) whose maternal parents (n = 4) were exposed to the substance via inhalation (whole body) for 4 days/week, 5 hours/day during gestation day 7 up to and including gestation day 20. The concentration level was 10.4 mg TiO2/m³ air (analytical). A vehicle control group was run concurrently. Eleven male rats were taken from four control or four exposed litters to undergo behavioural assessment. According to the authors, prenatal nano-TiO2exposure induced significant working or short-term memory impairments and changes in initial motivation in male pups. In addition, the authors mentioned that gestational nano-TiO2exposure significantly altered visible platform performance (group comparisons on the final trial not significant). The authors did not find any effects on locomotor, balance, affective, anxiety-like, or depressive-like behaviour in male pups following gestational nano-TiO2 exposure via maternal inhalation. Reference memory learning, retention, and perseveration in any phase of the Morris water maze test were not markedly altered in pups. Furthermore, no significant effects for either Working memory incorrect or Reference memory errors were observed. Lastly, the authors did not observe any marked impact on maternal weight, implantation site number or pup number/litter after test item exposure. This reference has several reporting and experimental deficiencies which do not allow a definitive conclusion on the exposure-effect correlation:

-      the number of pregnant rats was too low in this study (4 rats/group). The guideline foresees a total of 20 litters for each dose level in order to have an adequate number of offspring for the evaluation of neurotoxicity. This reduction of pregnant dams, which in turn leads to a low number of offspring available for testing, constitutes a significant reduction of the statistical power to levels below acceptability.

 

-      the number of offspring used for behavioural testing were too low, and it was not clearly stated how many offspring were taken from the respective groups. As a minimum number of animals in each dose group for pre-weaning and post-weaning examinations the guideline recommends 10/sex (1/litter) or 20/sex (1/sex/litter). However, only a total of eleven males was reported for testing behavioural and cognitive function, both for treatment and control groups. This low number of male pups again constitutes a substantial reduction in statistical power.

 

-      Only males were used for behavioural and cognitive testing which does not allow a conclusion for effects in females. It is widely accepted that female animals are usually more susceptible to behavioural and cognitive alterations. In addition, the guideline foresees testing of both genders.

 

-      The guideline foresees that at least three dose level and a concurrent control should be used. However, in this study only one high concentration level (10.4 mg TiO2/m³) was tested. Firstly, the usage of one concentration level precludes the possibility to demonstrate any dose-response and a determination of a No-Observed-Adverse Effect level (NOAEL) or a benchmark dose. Secondly, the concentration of 10 mg/m³ used in this study clearly exceeded the MTD, resulting in extreme stress and impaired breathing to be expected which may well influence foetal development indirectly without being test-item related.

 

-      The guideline foresees, as a minimum, that the test substance or control are given to the mated females from gestation day 6 (implantation) throughout lactation day 21. The pre- and postnatal neurological development phase of the offspring would then be covered by the exposure of the test item by using this dosing regime. In contrast, the test item exposure in this study was limited to the period of gestation (day 6 to day 20), which precludes the pups being exposed during the postnatal neurological development. In addition, the exposure did not occur continuously (4 d/week).

 

-      Observations of the dams with regard to clinical signs, mortality, detailed clinical observations or food consumption were missing. The observations of the offspring on clinical signs, mortality, body weight, and detailed clinical observations are missing. As stated by the guideline, body weight is a good indication of physical development of the pups. By not observing the health conditions of the offspring, it cannot be ruled out that alternative explanations for the results exist.

 

-      The guideline recommends that young adults be tested at postnatal day 60 – 70. In this study, behavioural testing of the male offspring was conducted at 20 weeks of age. Since testing was conducted at 20 weeks of age, it cannot be ruled out that the effects observed during behavioural testing might have been caused by an other factor due to the long lag tome between gestational exposure and testing.

 

-      Finally, the authors did not report any neuropathological examination of the male pups. The guideline states that the purposes of the examinations are to “(i) to identify regions within the nervous system exhibiting evidence of neuropathological alternations, (ii) to identify types of neuropathological alternations resulting from exposure to the test substance; (iii) to determine the range of severity of the neuropathological alterations”. A neuropathological examination of the pups would support and confirm the findings reported to have occurred during the behavioural testing. Due to the lack of this examination in this study, the findings of the authors cannot be further confirmed.

 

Due to the shortcomings presented above, this publication was disregarded for hazard and risk assessment.

 

 

Kubo-Irie et al. (2016) exposed groups of six pregnant ICR mice by subcutaneous injection to alumina-coated titanium dioxide nanoparticles (35 nm) in saline containing 0.05 % Tween-80. Concentrations used were 1, 10, 100, and 1000 µg/mL. A vehicle control group was run concurrently. The mice were treated on gestation day 5, 8, 11, 14, and 17 and they gave birth on gestation day 19. The effects of prenatal exposure to the test item on spermatogenesis in 12-week old mice was investigated. According to the authors, the test item was detected in the seminiferous epithelium of the male offspring and low cellular adhesion and degenerated Sertoli cells were observed in the seminiferous epithelium. These observations were made in all treatment groups. The detrimental function of the Sertoli cells resulted in the formation of abnormal spermatozoa. Lastly, they mentioned that the incident of abnormal sperm were a tendency to increase in line with the concentrations. The study represents a mechanistic investigation with very limited value for hazard assessment purposes. The following reporting and experimental deficiencies were noted in the publication:

 

- the test item was administered via subcutaneous injection. This unphysiological route of exposure is irrelevant for health hazard assessment of industrial chemicals.

 

- 20 pregnant females per treatment group are usually required for the investigation of the effect of the test item, but in this study only six females were used per treatment group. This significantly reduces the statistical power and precludes any meaningful evaluation of the effect of the test item.

 

- the female mice were exposed only on gestational day 5, 8, 11, 14, and 17. The guideline however foresees that the female mice are exposed daily prior to mating, during the mating period, pregnancy and up to the weaning of offspring.

 

- clinical signs, mortality, food consumption, body weight and gross pathology were not recorded for the female mice. Without this kind of information, it is unknown whether the test item had any adverse effect on the female animals, which in turn could have an adverse outcome on the developing offspring.

 

- full examination of offspring are missing (spermatogenesis examined only), it can therefore not be excluded that the findings were secondary to other effects in the offspring.

 

Based on the shortcomings given above, this publication was disregarded for hazard and risk assessment purposes.

 

Morgan et al. (2015) investigated the reproductive toxicity of titanium dioxide nanoparticles (Aeroxide P25; anatase; particle size: 10 nm) in male rats. Forty adult male albino rats were divided into two equal groups. One group was used as control receiving no treatment, whereas the second group received 100 mg/kg bw/day of the test item in distilled water via gavage. Half of the animals per group were killed after 4 weeks of exposure and the remaining animals were killed after 8 weeks of exposure. Frequency of exposure was daily. According to the authors, the results showed that titanium dioxide nanoparticles affected the male reproductive system as evidenced by marked reduction in body weight and relative sex organ weights (testis, epididymis, seminal vesicle and prostate gland). Furthermore, they stated that the sperm motility, concentration and viability percentage were also significantly reduced after exposure to titanium dioxide nanoparticles with increased incidences of sperm morphological abnormalities (deformed head, detached head, curved tail and coiled tail). In addition, serum testosterone levels were also said to be significantly decreased compared with control. Lastly, the investigators stated that marked histopathological alterations in testis, epididymis, seminal vesicle and prostate gland were observed. This reference however exhibits major reporting and experimental deficiencies, which do not allow firm conclusions on any reprotoxic effects, as follows:

 

- the rat strain used in the study was not characterised further, which does not allow any comparison with strain-specific historical control data, thus a meaningful evaluation of findings is not possible.

- although the bodyweight of 180-200 g at the start of the study appears plausible for a young rat (aged 6 -7 weeks), the bodyweight after 4 and 8 weeks of exposure appears grossly implausible: after 4 and 8 weeks the average body weight is given as 271 g and 272 g respectively. For an adult rat after 8 weeks post study initiation (age 14 - 15 weeks), one would expect an average body weight of either 470 - 480 g (Sprague Dawley) or 390 - 410 g (Wistar), for example.

- the body weight gain of 1 g between week 4 and week 8 after study initiation in the control group animals appears grossly implausible: for a young rat one would expect approx. 20 g of weight gain.

- the implausible body weight and the unusual bodyweight gain during the conduct of the study indicate inappropriate housing conditions and a failure of the lab to handle the animals properly, so that effects on relative organ weights cannot be excluded because of this.

- the SD of the testosterone levels appear surprisingly low, since these hormone levels are known to undergo a periodic alteration in the circadian cycle, resulting in a large variation over the day and also between animals; in this study, 

- the variation of the testosterone levels between exposed and control animals are in fact within the normal range of physiological fluctuation.

Based on the shortcomings given above, this publication was disregarded for hazard and risk assessment purposes.

 

Gao et al. (2012) investigated the ovarian injury and gene-expressed characteristics in female CD-1 (ICR) mice (30 mice/group) induced by intragastric administration of titanium dioxide nanoparticles (anatase; self-synthesised) in 0.5 % (w/v) hydroxypropylmethylcellulose at 10 mg/kg bw daily for 90 consecutive days. A vehicle control group was run concurrently. Following the 90-day administration, ten treated females were mated with ten males in order to provide information on fertility. According to the authors, the findings indicated that titanium dioxide nanoparticles can accumulate in the ovary and result in ovarian damage, cause an imbalance of mineral element distribution and sex hormones, decrease fertility or pregnancy rate and oxidative stress in mice. Furthermore, they state that microarray analysis showed that in ovaries from mice treated with TiO2 NPs compared to controls, 223 genes of known function were up-regulated, while 65 ovarian genes were down-regulated. The study represent a mechanistic investigation and is considered not to be of any value for hazard assessment purposes, because of the following deficiencies in reporting and experimental design:

 

- the data of this publication show standard deviations of exactly 5 % for all measured data. Standard deviations of exact 5 % throughout all experiments covering multiple endpoints in an in vivo system are implausible and factually impossible and therefore put into question whether this study represents authentic, reliable research data.

 

- the authors suggest that the test substance caused several adverse effects following oral administration of (assumed) nano-sized titanium dioxide in CD-1 mice. The test item was however a self-synthesised material by the authors, but with a poor description of the generated test substance. Since the test item is poorly described in the publication and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2particles at all were used in these experiments, and (ii) whether the effects can really be attributed to titanium dioxide.

 

- the authors describe a decrease of fertility in the female mice. Normally, 20 females per treatment group are used to investigate this effect of the test item on fertility, but in this study only 10 females were used. This significantly reduces the statistical power and reduces the meaningful evaluation of the effect of the test item.

 

- the guideline foresees that at least three dose levels and a concurrent control should be used. In this study, only one dose level (10 mg/kg bw) was tested. The usage of one dose level precludes the possibility to demonstrate any dose-related response and the determination of a No-Observed-Adverse Effect level (NOAEL).

 

Based on the shortcomings given above, this publication was disregarded for hazard and risk assessment purposes.

 

Hougaard et al. (2010, 2011; please refer to Section 7.9.1. Neurotoxicity) exposed time-mated C57BL/6BomTac female mice (22 - 23 mice/group) by inhalation 1 h/day to 42.4 mg/m³ air (analytical data) aerosolised titanium dioxide nanoparticles (UV-titan L 181; rutile; major particle size: ~100 nm (geometric mean number diameter: 97 nm)) on gestation days 8 - 18. A vehicle control group was run concurrently. The developmental neurotoxicity in offspring was investigated as well as maternal inflammatory response. In a follow-up study (Kyjovska et al., 2013), the influence of the maternal airway exposure to titanium dioxide nanoparticles on male reproductive function in the two following generations (F1 and F2) were evaluated. In order to evaluate the reproductive function, F1 offspring were cross-mated with naive CBA/J mice. According to Hougaard et al. (2010, 2011), inhalation of nano-sized coated titanium dioxide induced long-term lung inflammation in time-mated adult mice, and their gestationally exposed offspring displayed neurobehavioral alterations. Furthermore, Kyjovska et al. (2013) stated that maternal particulate exposure did not affect daily sperm production statistically significantly in the F1 generation, although titanium dioxide tended to reduce sperm counts. They also stated that time-to-first deliver the F2 litter increased with decreasing sperm production and there was no effect on sperm production in the F2 generation originating after TiO2 exposure. Furthermore, the authors report statistically significant differences in sperm production between mouse strains. This investigation has several reporting and experimental deficiencies, which do not allow firm conclusions on hazard potential to be drawn, for the following reasons:

 

-      the number of litters was too low in this study (13 - 14/dose). The guideline recommends 20 litters for each dose level in order to have an adequate number of offspring for the evaluation of neurotoxicity. This reduction of litters causes a significant reduction of the statistical power.

 

-      the guideline foresees that at least three dose levels and a concurrent control should be used. In this study, only one high concentration level (42.4 mg/m³ (analytical)) was tested. Firstly, the usage of one concentration level precludes the possibility to demonstrate any dose-related response and the determination of a No-Observed-Adverse Effect level (NOAEL). Secondly, the concentration of 42.4 mg/m³ used in this study clearly exceeded the MTD, resulting in extreme stress with impairment of normal breathing behaviour to be assumed, which mas well influence foetal development indirectly.

 

-      the guideline also foresees, as a minimum, that the test substance or control are given to the mated females from gestation day 6 (implantation) throughout lactation day 21. The pre- and postnatal neurological development phase of the offspring would then be covered by the exposure of the test item by this dosing regime. In this study, however, the test item exposure occurred only during gestation (days 8 to 18), which precludes that the pups were exposed during the postnatal neurological development.

 

-      behavioural investigations of the animals were conducted during the light period. Since rats are nocturnal animals, testing the animals during a period where they are not normally active might result in findings which one might not obtain otherwise.

 

-      observations of the dams with regard to food consumption are missing, just as data on clinical signs alsonot being reported. In addition, clinical signs, detailed clinical observations, learning and memory testing (adolescences), and body weight (weanlings) of the offspring are not documented. By not observing the health conditions of the offspring, it cannot be ruled out that alternative explanations for the results exist.

 

-      the authors did not report neuropathological examinations of the offspring. The guideline states that the purposes of the examinations are to “(i) to identify regions within the nervous system exhibiting evidence of neuropathological alternations, (ii) to identify types of neuropathological alternations resulting from exposure to the test substance; (iii) to determine the range of severity of the neuropathological alterations”. A neuropathological examinations of the pups would have supported and confirmed the findings observed during behavioural testing. Due to the lack of this examination in this study, the findings of the authors cannot be verified.

 

-      The guideline foresees that young adults be tested for motor and sensory function at postnatal day 60 – 70. In this study, testing of the offspring was conducted at 4 months of age. Since testing was conducted at 4 months of age, it cannot be ruled out that the effects observed during behavioural testing might have been caused by another factor during this long period between gestational exposure and testing.

 

-      Motor activity testing was too short and missing for the pre-weaning period. Firstly, since motor activity testing was not conducted during the pre-weaning period, there is no possibility to check whether a change in motor activity occurred during the developmental period of the offspring. Secondly, by measuring the motor activity over a shorter period of time than recommended, different results may have been obtained as compared to if the activity had been measured over longer period of time.

  

Based on the shortcomings given above, this publication was disregarded for the hazard and risk assessment.

 

 

Song et al. (2017) investigated the effects of titanium dioxide nanoparticles (5 - 10 nm) in PBS (pH 7.4) with 0.5 % Tween 80 on spermatogenesis in ICR mice (15 males/group). Groups of mice were treated with dosages of 10, 50, and 100 mg/kg bw/day for 28 days. A vehicle control group was run concurrently. The results showed that anatase titanium dioxide nanoparticles could lead to sperm malformation. However, the testicle and epididymis indexes were not significantly different compared with those of the control group. Furthermore, the nanoparticles reduced the germ cell number and led to spherospermia, interstitial glands vacuole, malalignment, and vacuolization of spermatogenic cells in mice testes. Lastly, no effects on clinical signs and body weight were observed.

 

This reference had several reporting and experimental deficiencies, which do not allow an independent review about the exposure-effect correlation:

 

- the duration of treatment with the test item was too short in this study. According to the guideline, males should be dosed with the test item for at least once complete spermatogenic cycle in order to observe any adverse effects on spermatogenesis caused by the test material. The complete spermatogenic cycle of mice lasts approx. 56 days, but the animals in the current study were only dosed for 28 days. Therefore, the exposure to the early stages of developing spermatozoa will not be covered and the effects on functional fertility can not be investigated.

 

- the dose levels were too low. The guideline foresees as highest dose level a dose level inducing toxicity but no mortality. In the current study no clinical signs or body weight effects were observed at the highest dose. Furthermore, no justification was given for the dosing regime.

 

-  males were not mated following test item administration. Therefore, it not possible to investigate the effect of the test item on the ability of the sperm to produce healthy and alive offspring, which makes it impossible to obtained further information about the effect on fertility by the test material.

 

- information on food consumption and gross pathology was missing. Furthermore, the testes were only investigated during histopathology. Gross pathology and histopathology of the accessory sex organs would provided further information on fertility effects.

 

Based on the shortcomings given above, this publication was disregarded for the hazard and risk assessment.

Rodriguez-Escamilla et al. (2019) investigated the effects of E171 consumption in a solid matrix (test substance in pellets) and liquid suspension on testis structure, inflammation infiltrate and blood testis barrier disruption. Groups of 4 male BALB/c mice were given 0.1%, 0.5% and 1% via diet administration (pellets) for a duration of 7 weeks. A control group was run concurrently. Furthermore, a group of 4 male mice were treated with 5 mg/kg bw in a liquid suspension via gavage for a duration of 10 weeks. Also, a control group was run concurrently. The results showed that none of the administration routes had influence on body weight. Mice fed with pellets containing 1% E171 showed moderate anxiety and facial grooming. No other differences in behaviour were observed. Food consumption was not influenced in animals that received E171 in drinking water via gavage. However, mice fed with pellets containing 0.1% E171 showed a decreased food consumption throughout the study. Germinal epithelium detachment was observed in all dose groups treated via the diet but not in animals treated via gavage. Germ cells sloughing was observed in all dose groups, independently of the administration method. Parenchymal cell counting was observed in 1% group of animals treated via diet and MHC-II expression was increased in mid and high dose groups treated via diet and also in animals treated via gavage. N-cadherin and collagen type I expression was decreased in all dose groups (diet and gavage). Hydroxyproline was decreased in mid and high dose groups of animals treated via diet but not in animals treated via gavage. This reference had several reporting and experimental deficiencies, which do not allow an independent review about the exposure-effect correlation:

 

-        the test item data on particle characterisation is poorly described, which makes it impossible to verify the identity of the test item. Furthermore, food grade titanium dioxide was obtained from SENSIENT. However, no further information was provided and on the respective website no information can be obtained.

 

-       toxicity on reproduction in female rats was not evaluated, only males were evaluated

 

-        only selected parameters such as germinal epithelium detachment, germ cells sloughing, and expression analysis of various proteins were investigated, which are parameters not required for building an expert judgement and further assessment of the test substance.

 

-       the number of animals per group (n=4) is not sufficient to perform an appropriate statistical analysis.

 

-       the authors tried to compare the effects observed after dietary and intragastric E171 administration. However, the animals in the dietary administration study were 3 weeks younger than those used for the gavage study and the animals in the dietary administration study were treated for 7 weeks and in the gavage study for 10 weeks. These differences as such are sufficient to conclude that these both experiments cannot be compared.

Due to the shortcomings presented above, this publication was disregarded for hazard and risk assessment.

 

 

Summary entry – Toxicity to reproduction

Another sixteen references were also identified during the comprehensive literature search, representing in vivo experiment with investigation on reproduction toxicity. These experiments were conducted with mice or rats receiving ultrafine or pigment-grade titanium dioxide via oral (gavage or unspecified), inhalation (whole body), intravenous, or intratracheal administration. The study designs are not in accordance with accepted guidelines and are therefore of limited relevance for chemicals hazard assessment. The references usually lack significance due to, e.g., poor test item characterisation, low number of animals used, missing dose response relationship, unjustified dosing regime, or unphysiological route. It is therefore concluded that all references do not fulfil the criteria for quality, reliability and adequacy of experimental data for the fulfilment of data requirements under REACH and hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X). The studies given below were included in the IUCLID for information purposes only:

 

Zhao, X. et al. (2013): The publication is not relevant for human health risk assessment since the test item is self synthesized and data on particle characterisation is poorly described. The experimental design is insufficiently documented and not in accordance with any relevant guideline. The test item was administered 90 days before mating and no treatment occurred during the mating period. Additionally, females were not treated during pregnancy and nursing period. Although 100 animals per group were stated, only 5 females per group were used for the evaluation of TiO2-induced toxicity to reproductive system. Thus, the number of animals per group is too low for an appropriate statistical analysis (about 20 pregnant females are necessary). Besides of these main restrictions the following parameters were not stated: housing conditions during mating and for pregnant females, mating procedure, handling of pregnant animals (additional stress from outside factors), dosing volume, clinical observation, mortality of P1 and pups, food consumption record, litter size and adjustment, full gross pathology of P generation, full histopathology (only ovaries), physical or behavioural abnormalities, individual raw data.

 

Miura, N. et al. (2017): The publication is not relevant for human health risk assessment since the test item data on particle characterisation is poorly described. The experimental design is insufficiently documented and not in accordance with any relevant guideline. Additionally, the test item was administered by an unphysiological route and the toxicity on reproduction in female mice was not evaluated (only male testicular function). The dosing regime (weekly for 4 weeks) is not justified and not in accordance with a guideline for testing reproductive toxicity. The number of animals per group is not clearly stated and therefore the results are difficult to interpret. Additionally, the investigators suspended the acidic P25 in an assumed alkaline DSP solution. No investigations regarding pH effects were performed. Besides of these main restriction the following parameters were not stated: body weight record, food and water consumption, mortality, clinical observation, full histopathology (only male reproductive organs) and full gross pathology. 

 

Gao, G. et al. (2013): The publication is not relevant for human health risk assessment. The test item was self-synthesis by the authors, but they gave a poor description of the acquired test substance. Since the test item is poorly described in the publication and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in these experiments, and (ii) whether the effects can really be attributed to titanium dioxide. Furthermore, the data of this publication shows standard deviations of exactly 5 % for all measured data. Standard deviations of exact 5 % throughout all experiments covering multiple endpoints in an in vivo system is impossible and puts into question whether this study represent reliable research data. Also, the experimental design is not in accordance with any relevant guideline. The toxicity on reproduction in female mice was not evaluated (males only) and the dosing regime (daily for 90 days) is not justified. Besides of these main restriction the following parameters were not stated: body weight (at the end of study only), food consumption, full histopathology (male reproductive organs only) and full gross pathology. Lastly, individual data was missing.

 

Hong, F. et al. (2015a): The publication is not relevant for human health risk assessment. The test item was self-synthesis by the authors, but they gave a poor description of the acquired test substance. Since the test item is poorly described in the publication and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in these experiments, and (ii) whether the effects can really be attributed to titanium dioxide. Also, the experimental design is not in accordance with any relevant guideline. The toxicity on reproduction in female mice was not evaluated (males only) and the dosing regime (daily for 6 months) is not justified. Furthermore, the number of males per dose was too low. Besides of these main restriction the following parameters were not stated: body weight (at the end of study only), full histopathology and full gross pathology. Lastly, individual data was missing.

 

Hong, F. et al. (2015b): The publication is not relevant for human health risk assessment. The test item was self-synthesis by the authors, but they gave a poor description of the acquired test substance. Since the test item is poorly described in the publication and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in these experiments, and (ii) whether the effects can really be attributed to titanium dioxide. Also, the experimental design is not in accordance with any relevant guideline. The toxicity on reproduction in female mice was not evaluated (males only). Furthermore, the following parameters were not stated: clinical signs, mortality, full gross pathology, and full histopathology. Food consumption was recorded, but not weekly as foreseen by the guideline. Furthermore, acclimatisation period and housing conditions were not described. Lastly, individual data was missing.

 

Hong, F. et al. (2016): The publication is not relevant for human health risk assessment. The test item was self-synthesis by the authors, but they gave a poor description of the acquired test substance. Since the test item is poorly described in the publication and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in these experiments, and (ii) whether the effects can really be attributed to titanium dioxide. Also, the experimental design is not in accordance with any relevant guideline. The toxicity on reproduction in female mice was not evaluated (males only) and the dosing regime (daily for 8 or 9 months) is not justified. A low number of animals was used for mating in order to investigate the fertility of the males and mating procedure was not described in detail. No investigation of the litters was conducted. Furthermore, the following parameters were not stated: clinical signs, mortality, food consumption, body weight, full gross pathology and full histopathology (testes only). Lastly, individual data was missing.

 

Zhou, Y. et al. (2017): The publication is not relevant for human health risk assessment. The test item was self-synthesis by the authors, but they gave a poor description of the acquired test substance. Since the test item is poorly described in the publication and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in these experiments, and (ii) whether the effects can really be attributed to titanium dioxide. Also, the experimental design is not in accordance with any relevant guideline. The number of pregnant females per dose was too low. Furthermore, the following parameters were not investigated/reported in dams: clinical signs, mortality, detailed clinical observation, body weights, and food consumption. In addition, the following parameters were not investigated/reported in the offspring: clinical signs, mortality, clinical observations, body weights, functional/behavioural endpoints, sex determination, incomplete neuropathology (hippocampus only).

 

Hong, F. & Wang, L. (2018): The publication is not relevant for human health risk assessment. The test item was self-synthesis and a poor description of the acquired test substance was given. Since the test item is poorly described in the publication and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in these experiments, and (ii) whether the effects can really be attributed to titanium dioxide. Also, the experimental design is not in accordance with any relevant guideline. The strain of the mice was not stated and the number of pregnant females per dose group was too low (n = 10/dose; even the authors stated at the beginning that 50 females/dose were treated, only 10 females per dose were mated and it is not clearly stated what happened with the remaining animals). Furthermore, the females were only dosed prior to mating and it was unclear how long the mating period lasted. No investigation of litter was conducted. In addition, the following parameters were not investigated/reported in dams: clinical signs, mortality, food consumption, duration of gestation and gross pathology. Body weight was only measured at the beginning and at the end of the study. Lastly, individual data was missing.

 

Sharafutdinova L.A. (2018): The publication is not relevant for human health risk assessment since the test item data on particle characterisation is poorly described. The experimental design is insufficiently documented and not in accordance with any relevant guideline. Additionally, toxicity on reproduction in female rats was not evaluated (only immunohistochemical and morphometric characteristics of the spermatogenic epithelium). The dosing scheme of just 14 days is not justified and the number of animals per group and time point (n=10) is not sufficient to perform an appropriate statistical analysis. Further, only one dose group was tested which does not allow a dose-response related analysis. Only selected parameters were investigated and the investigated endpoints are not required for building an expert judgement and further assessment of the test substance under REACH (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X).

 

Miura, N. et al. (2019a): The publication is not relevant for human health risk assessment since the test item data on particle characterisation is poorly described. The experimental design is insufficiently documented and not in accordance with any relevant guideline. Additionally, toxicity on reproduction in female mice was not evaluated (only sperm motility and an in vitro fertilisation assay with mature oocytes). The dosing scheme of a single administration is not justified and the number of animals per group (n=1) is not sufficient to perform an appropriate statistical analysis (it is also worthy of note that the author justified the low number of animals with technical difficulties). Further, only two dose groups were tested (and unclear if 10 or 20 mg/kg) which does not allow a dose-response related analysis. The used oocytes for the in vitro fertilisation assay were obtained from 3-4 weeks old female mice. These females were clearly too young to obtain oocytes for an in vitro fertilisation test. Only selected parameters were investigated and the investigated endpoints are not required for building an expert judgement and further assessment of the test substance under REACH (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X).

 

Miura, N. et al. (2019b): The publication is not relevant for human health risk assessment since the test item data on particle characterisation is poorly described. The experimental design is insufficiently documented and not in accordance with any relevant guideline. Additionally, toxicity on reproduction in female mice was not evaluated (only sperm motility, number and ATP content). The dosing scheme of a single oral administration and 3 days p.a. observation or alternatively single administration once a week for 4 weeks is not justified. The number of animals per group (n=5) is not sufficient to perform an appropriate statistical analysis. Further on, the non-physiological route of administration via i.v. injection is not guideline conform and not suitable for human hazard assessment. Overall, only selected parameters were investigated and the investigated endpoints are not required for building an expert judgement and further assessment of the test substance under REACH (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X).

 

Karimipour, M. et al. (2018): The publication is not relevant for human health risk assessment since the test item data on particle characterisation is poorly described. The experimental design is insufficiently documented (e.g. mating procedure, whether male mice were also treated before mating) and not in accordance with any relevant guideline. Additionally, toxicity on reproduction in male mice was not evaluated. Although the number of animals per group is high (n=27), the number of animals per evaluated endpoint (n=7-10) is not sufficient to perform an appropriate statistical analysis. Further, only one dose group was tested which does not allow a dose-response related analysis. The signs of systemic toxicity such as body weight or food consumption were not evaluated throughout the study. Overall, only selected parameters were investigated, individual data is missing and the investigated endpoints are not required for building an expert judgement and further assessment of the test substance under REACH (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X).

 

Zhou, Y. et al. (2018): Since the self-synthesized test substance is poorly described and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in this experiment, and (ii) whether the effects can really be attributed to titanium dioxide and (iii) the methodical setup is not in accordance with any guideline. The number of animals per group in each experiment (n=5) is too low for a statistical evaluation and the experimental design is insufficiently reported. Further, the potential toxicity of TiO2 was only investigated in female mice and signs of systemic toxicity were not investigated. Apart from that, this publication is part of the publication series of the University of Soochow. The publications of this working group were subject to an investigation of the University of Soochow, resulting in the retraction of several publication. The shortcomings identified were incorrect statistics, experimental errors and missing original data. Thus, due to these major study restrictions this study is disregarded.

 

Elnagar, A.M. B. et al. (2018): The publication is not relevant for human health risk assessment since the test item data on particle characterisation is poorly described. The experimental design is insufficiently documented (e.g. informations about source and characterisation of the vehicle, the test animals and environmental conditions) and not in accordance with any relevant guideline. Additionally, toxicity on reproduction in male mice was not evaluated. Although the number of animals per group is too low (n=10), the number of animals per evaluated endpoint (n=10) is not sufficient to perform an appropriate statistical analysis. Further, only one dose group was tested which does not allow a dose-response related analysis. The strain of the used males was not stated and signs of systemic toxicity such as body weight or food consumption were not evaluated throughout the study. Overall, only selected parameters were investigated and the investigated endpoints are not required for building an expert judgement and further assessment of the test substance under REACH (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X).

 

Stapleton, P.A. et al. (2018): The publication is not relevant for human health risk assessment since the test item is poorly described (e.g. purity, stability and particle characterisation). The experimental design is insufficiently documented and not in accordance with any relevant guideline. Firstly, the number of animals per endpoint is not sufficient to perform an appropriate statistical analysis. Further, only one dose was tested, which does not allow a dose-response related analysis. In addition, the test item was only given once to the animals for a duration of 5 hours. Normally, the test item should be given for a longer time span (seven-day per week basis). No justification was given for single administration. Also, it cannot be ruled out that the rats in the whole-body chamber also ingest the test substance orally during the exposure by cleaning their bodies by themselves. Additionally, toxicity on the foetuses was not evaluated. Females were not investigated for clinical sings, mortality as well as food consumption and a complete gross pathology is missing. Overall, only selected parameters (body weight, oestrus cycle, leukocyte trafficking, vascular function) were investigated and the investigated endpoints are not required for building an expert judgement and further assessment of the test substance under REACH (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X). Lastly, historical control and individual data are missing.

 

Hong, F. et al. (2020): The experimental design is insufficiently described. It is not explicitly described that the TiO2 nanoparticles were administered in the same solution as what the control group received. Only male mice were exposed and investigated. The males were not mated after test item administration therefore, it is not possible to investigate the observed effects in the testis of the test substance on the ability of the sperm to produce healthy and alive offspring, which makes it impossible to obtained further information about the effect on fertility by the test material. Some parameters were investigated with a low number of animals. Information on clinical signs or body weight effects, daily food consumption and gross pathology are missing. The sperm parameters (motility, morphology and sperm counts) were not examined. Furthermore, the testis were only investigated during histopathology. Especially a gross pathology and histopathology of the associated sex organs would provide further information on fertility effects. Historical and individual data are missing. Overall, only selected parameters were investigated and the investigated endpoints are not required for building an expert judgement and further assessment of the test substance under REACH (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X).

 

 

Conclusion

An extended one generation reproduction (EOGRT) study according to OECD guideline 443 (2018) was commissioned by the Titanium Dioxide Manufactures Association (TDMA) in response to a request by the European Food Safety Authority (EFSA) as set forth in their Scientific Opinion of 28 June 2016 ((EFSA Journal 2016;14(9):4545)) on the re-evaluation of titanium dioxide (E171) as a food additive. Based on the results of this guideline conform study no toxicity to reproduction could be determined. All derived No-observed-effect-levels (NOAELs) are given below:

F0 Generation:

General toxicity: NOAEL above 1000 mg Titanium dioxide E171-E/kg bw/day via the diet.

Reproductive toxicity: NOAEL above 1000 mg Titanium dioxide E171-E/kg bw/day via the diet.

 

F1 Generation

Reproductive and developmental toxicity (Cohort 1A and 1B): NOAEL above 1000 mg titanium dioxide E171-E/kg bw/day via the diet.

Developmental neurotoxicity (Cohort 2A and 2B): NOAEL above 1000 mg Titanium dioxide E171-E/kg bw/day via the diet.

Developmental immunotoxicity (Cohort 3): needs to be determined (early 2021)

 

In addition, one reproductive toxicity screening assay is available, but only as a brief study summary without raw data and was therefore rated as not assignable (RL=4). This study showed no impairment of male or female fertility. Due to its use within the OECD HPV programme, the study is used as supporting information.

 

Further twenty-two references report mechanistic studies with a focus on organ specific investigations, such as ovarian injury, investigation of male reproductive system or female reproductive system only, fertility of adults, behavioural/neurotoxicity testing of offspring, or fertility in offspring of exposed dams. However, the study designs of these references are not in accordance with any accepted guideline and are therefore of limited relevance for chemical hazard assessment. The references also lack significance for example due to poor test item characterisation, non-physiological routes of administration, low number of animals available for testing, offspring too old during behavioural testing, one concentration tested only, short exposure duration, wrong dosing regime, implausible body weight documentation, and/or incomplete observation data. It is therefore concluded that all of these references do not fulfil the criteria for quality, reliability and adequacy of experimental data under REACH for hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X). The studies were therefore included for information purposes only.

 

Effects on developmental toxicity

Description of key information

Based on the information from the available six pre-natal developmental toxicity studies via gavage in rats (according to OECD 414, GLP) it is concluded that maternal toxicity up to the highest dose level of 1000 mg/kg bw/d, and reproduction data of dams were not affected by treatment. There was no embryo-toxicity and no effect on the development of foetuses. No external visceral or skeletal malformations were observed and the incidence of variations was not different between treated and control groups. No incidences of skeletal retardations were observed.

Link to relevant study records

Reference

Reference 1

Endpoint:

developmental toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2013-07-15 to 2013-08-29

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 414 (Prenatal Developmental Toxicity Study)

Version / remarks:

2001-01-22

Deviations:

no

GLP compliance:

yes (incl. QA statement)

Remarks:

signed 2013-01-22

Limit test:

no

Species:

rat

Strain:

Wistar

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River, 97633 Sulzfeld, Germany
- Age at the start of the treatment period: 11-12 weeks old
- Weight at the initiation of pairing: males: 285-323 g (mean: 304.5 g, ± 20% = 243.62 – 365.42 g); females: 184-225 g (mean: 201.14 g, ± 20% = 161.71 – 242.57 g)
- Housing: kept individually in IVC cages (except during the mating period when two females were paired with one male), type III H, polysulphone cages on Altromin saw fibre bedding (lot Nr. 240113)
- Diet (ad libitum): Altromin 1324 maintenance diet for rats and mice (lot Nr. 1426)
- Water (ad libitum): tap water, sulphuric acid acidified to a pH of approximately 2.8
- Acclimation period: at least 5 days

ENVIRONMENTAL CONDITIONS
- Temperature: 22 ± 3°C
- Humidity: 55 ± 10%
- Air changes: 10 x / hour
- Photoperiod (hrs dark / hrs light): 12/12

Route of administration:

oral: gavage

Vehicle:

other: aqua ad iniectabilia (water for injection)

Details on exposure:

PREPARATION OF DOSING SOLUTIONS:
The test item formulation was prepared freshly on each administration day immediately prior to dosing.
The test item was weighed into a tared plastic vial on a suitable precision balance and the vehicle was added to give the appropriate final concentration of the test item. The formulation was vortexed for 2-3 minutes.
Homogeneity of the test item in the vehicle was maintained by vortexing the prepared suspensions thoroughly before every dose administration.
Application volume for all groups was 5 mL/kg body weight.
For each animal the individual dosing volume was calculated on the basis of the body weight most recently measured (measured weekly).

VEHICLE
- Batch no.: 26210S1-2

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

The assessment of homogeneity, stability as well as a determination of the nominal concentration of the test item in the vehicle was performed at various intervals.
Samples for analysis of the nominal concentration of the dose formulations of the test item in the vehicle were taken in the first and last week of the study for all doses (8 samples in total).
Samples for the analysis of homogeneity were taken from the top, middle and bottom of the high dose, medium dose and low dose preparation. Samples were taken in the first and last week of the study (18 samples in total).
Samples for the stability analysis were taken from low, medium and high dose groups in study week 1 at 0 hr and 6 hrs (6 samples in total).
All formulation samples were stored at -20° C.

The analytical method for the determination of titanium in dose formulation samples was successfully validated according to SANCO/3029/99 rev. 4 (11/07/00).
The determinations were performed by ICP-OES using two independent emission wavelengths highly specific for titanium, one wavelength for quantification and one for confirmation.
The requirements of SANCO/3029/99 rev. 4 (11/07/00) regarding linearity, precision (repeatability), accuracy (recovery) and specificity were fulfilled.
The mean recovery values for titanium at all fortification levels obtained by ICP-OES comply with the standard acceptance criteria of SANCO/3029/99, which demands that the mean recovery at each fortification level should be in the range of 70% - 110%.
Test substance concentration, stability and homogeneity in dosing formulations were obtained by ICP-OES.

Results:
The analytical results obtained for the individual dose groups were consistent with the analysis of the % of nominal of the test item for the concentration, stability and homogeneity analyses, with the exception of homogeneity sample numbers 18 (300 mg/kg bw/day, week 1) and 31 (100 mg/kg bw/day, last week) for which the recoveries were higher. The mean recovery noted for sample number 20a (300 mg/kg bw/day, week 1) was 43.1%. These variations were considered to be caused by a sampling error.
In this case the samples need to be vortexed for approximately 2 minutes before sampling. The dose formulation sample preparation was made every day freshly before the dose administration. On the day of analytical sample collection the formulation samples were prepared in excess and analytical samples were collected from the same stock as that used for the dose administration. The formulation samples were assumed to have not been subjected to proper homogenization procedure before the collection of sample numbers 18, 20 and 31. This was evident from the % recoveries of the nominal concentration and the stability analysis of sample numbers 3, 11, 12 of 300 mg/kg bw/day group and 4, 8, 13 and 14 of 100 mg/kg bw/day group (collected from the same formulation preparation as that of the sample numbers 18, 20 (300 mg/kg bw/day group) and 31 (100 mg/kg bw/day group), respectively, which had a consistent and an acceptable recovery. As there were no toxicity findings seen at any of the tested dose levels including the HD group, this concentration increase and decrease in isolated analytical samples of the 100 mg/kg bw/day and/or 300 mg/kg bw/day groups had no impact on the validity and integrity of the study.

Details on mating procedure:

- Impregnation procedure: cohoused
- If cohoused/M/F ratio per cage: females were paired with males as per the ratio of 1:2 (male to female).
- Proof of pregnancy: sperm in vaginal smear referred to as day 0 of gestation

Duration of treatment / exposure:

Gestation day 5 through gestation day 19

Frequency of treatment:

once daily, 7 days per week

Duration of test:

20 days

Dose / conc.:

100 mg/kg bw/day (actual dose received)

Dose / conc.:

300 mg/kg bw/day (actual dose received)

Dose / conc.:

1 000 mg/kg bw/day (actual dose received)

No. of animals per sex per dose:

24-25 mated female rats

Control animals:

yes, concurrent vehicle

Details on study design:

- Dose selection rationale: since little or no toxicity was anticipated for the test substance, the highest dose level was set at 1000 mg/kg bw/d corresponding to a limit dose for this study. Thereafter, a descending sequence of dose levels was selected with a view to demonstrate any dose-related response and a NOAEL.

Maternal examinations:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: general clinical observations: once a day; morbidity and mortality: twice daily, except during holidays and weekends where the observation was made once daily.
- Cage side observations checked: spontaneous activity, lethargy, recumbent position, convulsions, tremors, apnoea, asphyxia, vocalisation, diarrhoea, changes in the skin and fur, eyes and mucous membranes (salivation, discharge), piloerection and pupil size, changes in gait, posture, response to handling as well as the presence of clonic or tonic movements, stereotypes, difficult or prolonged parturition or bizarre behaviour.

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: prior to the start of the mating, a detailed clinical observation outside the home cage was made.

BODY WEIGHT: Yes
- Time schedule for examinations: once before the assignment to the experimental groups, on the first day of administration and weekly during the treatment. The sperm positive females were weighed during gestation days 0, 5, 8, 11, 14, 17 and 20.

FOOD CONSUMPTION AND COMPOUND INTAKE: Yes
- Food consumption for each animal determined: Yes, food consumption of pregnant females was measured on gestation days 5, 8, 11, 14, 17 and 20.
- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: No

WATER CONSUMPTION AND COMPOUND INTAKE: No

POST-MORTEM EXAMINATIONS: Yes
- Sacrifice on gestation day #20
At the time of termination or death during the study, each dam (presumed pregnant female) was examined macroscopically for any structural abnormalities or pathological changes which may have influenced the pregnancy.
- Organs examined: immediately after the termination, the uteri were removed and the pregnancy status of the dams was confirmed. Uteri that appeared non-gravid were further examined by staining with 10 % ammonium sulphide solution to confirm the non-pregnant status.

Ovaries and uterine content:

The ovaries and uterine content was examined after termination: Yes
Examinations included:
- Gravid uterus weight: Yes, with cervix
- Number of corpora lutea: Yes
- Number of implantations: Yes
- Number of early resorptions: Yes
- Number of late resorptions: Yes
- Other: the uterine contents were examined for embryonic or foetal deaths as well as the number of viable foetuses. The position and number of foetuses in each uterine horn were also recorded.

Fetal examinations:

- External examinations: Yes, all per litter
- Soft tissue examinations: Yes, half per litter
- Skeletal examinations: Yes, half per litter
- Head examinations: Yes, craniofacial examination of the heads of the foetuses used for the soft tissue examination

All foetuses were weighed and sexed based on the anogenital distance.

Statistics:

A statistical assessment of the results of the body weight and food consumption was performed by comparing values of dosed with control animals using a one-way ANOVA and a post-hoc Dunnett Test. Foetal evaluation parameters like external, visceral, craniofacial and skeletal parameters were analysed using a Chi-square test. The statistics were performed with GraphPad Prism V.6.01 software (p<0.05 is considered as statistically significant).
For abnormality “pelvic girdle ilium bone offset” the statistical analyses was performed by combining all unilateral and bilateral findings.
For abnormality “cervical vertebral centra- unossified” the statistical analyses was performed by combining all the unossified cervical vertebral centra (vertebral centra 1 to 7).

Indices:

no data

Historical control data:

Historical control data is provided for the following:
- mean uterine data
- mean litter weight (g) data
- foetal external examination
- foetal visceral examination
- foetal craniofacial examination
- foetal skeletal examination

Clinical signs:

no effects observed

Dermal irritation (if dermal study):

not examined

Mortality:

mortality observed, non-treatment-related

Body weight and weight changes:

no effects observed

Food consumption and compound intake (if feeding study):

no effects observed

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

not examined

Ophthalmological findings:

not examined

Haematological findings:

not examined

Clinical biochemistry findings:

not examined

Urinalysis findings:

not examined

Behaviour (functional findings):

not examined

Immunological findings:

not examined

Organ weight findings including organ / body weight ratios:

not examined

Gross pathological findings:

no effects observed

Neuropathological findings:

no effects observed

Histopathological findings: non-neoplastic:

not examined

Histopathological findings: neoplastic:

not examined

Other effects:

not examined

Details on results:

- mortality: none of the animals died due to treatment in the study. However, one female animal in the 100 mg/kg bw/day group showed signs of paraplegia during the initial days of dose administration. This animal was euthanized on gestation day 10. This finding was considered to be incidental and not related to treatment.

- clinical observations: no clinical signs of toxicological relevance noted in any of the animals of the treated groups in comparison to the control.

- body weight development: no treatment related effect noted for body weight and body weight change in the treated groups in comparison to the controls. No treatment related changes noted for terminal and adjusted maternal body weight in the treated groups in comparison to the controls. The statistical analysis of the body weight data showed no statistical significance.

- food consumption: no treatment related effect noted for food consumption in treated groups in comparison to the controls. The statistical analysis of the food consumption data showed no statistically significant changes between the treated and the control group.

- pathology: there were no macroscopic findings considered to be related to the treatment in any of the animals of the control and or test item treated groups at necropsy. However, there was a fluid distended uterus observed in one female of the 1000 mg/kg bw/day group. This animal was non pregnant and the fluid distension could be the normal physiological change of uterus during normal oestrus cycle. Also considering the finding was reported in a single female of the 1000 mg/kg bw/day group this was not considered to be related to the treatment.

Please also refer to the field "Attached background material" below.

Number of abortions:

no effects observed

Pre- and post-implantation loss:

no effects observed

Total litter losses by resorption:

not examined

Early or late resorptions:

no effects observed

Dead fetuses:

no effects observed

Changes in pregnancy duration:

no effects observed

Changes in number of pregnant:

no effects observed

Other effects:

not examined

Details on maternal toxic effects:

- none of the females showed signs of abortion or premature delivery prior to the scheduled terminal sacrifice.

- no treatment related changes noted for the prenatal parameters including the number of corpora lutea, number of implantation sites, early and late resorptions, or pre- and post-implantation loss.
However, there was an increase in the number of early resorptions in the 300 mg/kg bw/day group, which also accounted for the increased total resorption in the 300 mg/kg bw/day group. There was no statistically significant or dose-related response noted for the increase in early resorptions and these were not considered treatment related.


- no effects on the pregnancy rate of the animals. The rates in the control and treated groups were as follows: control group: 80%; 100 mg/kg bw/day group: LD 92%; 300 mg/kg bw/day group: 83.33% and 1000 mg/kg bw/day: 83.33%.

Please also refer to the field "Attached background material" below.

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Fetal body weight changes:

not examined

Reduction in number of live offspring:

not examined

Changes in sex ratio:

no effects observed

Changes in litter size and weights:

no effects observed

Changes in postnatal survival:

not examined

External malformations:

no effects observed

Skeletal malformations:

no effects observed

Visceral malformations:

no effects observed

Other effects:

not examined

Details on embryotoxic / teratogenic effects:

Embryotoxic / teratogenic effects:no effects

Details on embryotoxic / teratogenic effects:
- no treatment related changes noted for the prenatal parameters including the live foetuses, number of dead foetuses, number of male and female foetuses, or sex ratio.
In addition, there was a slight decrease in the number of female foetuses in the 1000 mg/kg bw/day group, but considering the slightly higher number of male foetuses and total live foetuses in the 1000 mg/kg bw/day group being comparable to the control it was assumed that decreased numbers of female foetuses were compensated by a slightly higher number of male foetuses. In addition the mean values for the numbers of male and female foetuses were within the historical control data range. Hence, the finding was not considered to be associated with treatment. Because foetal sex is determined shortly after conception and well before the onset of dosing on gestation day 5, such changes in sex ratio are not considered to be indicative of a test substance-related effect.
The statistical analysis of data showed no statistically significant changes between the treated and the corresponding control group.
- no effect of treatment noted for the litter data including mean litter weight, total litter weight and male and female litter weight. However, there was a slight but and not statistically significant decrease in female litter weight noted in the 1000 mg/kg bw/day group (-12.96%). Given that all litter weight data reported were within the historical control data range, the finding was not considered to be associated with the treatment.

- no external abnormalities considered to be of toxicological relevance noted in any of the treated groups. The statistical analysis showed no significant changes. However, there were a few abnormalities noted in a few isolated foetuses of the control, 300 mg/kg bw/day and 1000 mg/kg bw/day groups. The abnormalities were gastroschisis in the control, micrognathia in the 300 mg/kg bw/day and 1000 mg/kg bw/day groups and small upper jaw in the 300 mg/kg bw/day group. The single foetus with micrognathia and small upper jaw in the 300 mg/kg bw/day group was used for visceral examination and could not be verified by skeletal examination.
The foetuses with micrognathia in the 1000 mg/kg bw/day group were checked during skeletal examination and only 1 of 2 foetus was confirmed with micrognathia and the other looked normal. These abnormalities were observed in a single foetuses of a single isolated female animal from either control or treated groups and therefore were considered to be spontaneous in their origin and unrelated to the treatment.

- skeletal examination revealed a range of abnormalities in the control and treated groups that were either within historical control ranges recorded for this laboratory; were significantly lower than the corresponding control values; or were seen only in the 300 mg/kg bw/day or 100 mg/kg bw/day dose groups and were not dose dependent.
There was statistically significant increase in the foetal incidences for unossification of vertebral cervical centrum in the 1000 mg/kg bw/day groups. However, the percent litter incidence in the 300 mg/kg bw/day and 1000 mg/kg bw/day groups was lower than in the concurrent control group. This variation, from the developmental perspective was of minimal significance and is normal in the foetuses of this strain of rats with C-section on gestation day 20. Therefore, this abnormality was not considered to be an adverse effect related to the treatment.

- internal examinations of the foetal viscera revealed a range of visceral abnormalities in all groups including the control. There were no abnormalities of toxicological relevance.

- craniofacial examination revealed a range of abnormalities in all groups including controls. These abnormalities either did not differ significantly from control values or did not show dose-related responses and were thus considered to have no relevance to treatment.

Please also refer to the field "Attached background material" below.

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Developmental effects observed:

not specified

Conclusions:

In this prenatal developmental toxicity study, the repeated dose oral administration of TiO2 pg-2 to pregnant female Wistar rats at doses of 100, 300 and 1000 mg/kg bw/ day from gestation day 5 through gestation day 19 produced no adverse toxicological effects in the females or foetuses or significant developmental effects at any administered dose.
Based on the findings from this study, the NOAEL (No-Observed-Adverse-Effect-Level) of TiO2 pg-2 in the Wistar rat for both maternal toxicity and developmental toxicity is considered to be 1000 mg/kg bw/ day.

Reference 2

Endpoint:

developmental toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2013-07-10 to 2014-04-24

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 414 (Prenatal Developmental Toxicity Study)

Version / remarks:

2001-01-22

Deviations:

no

GLP compliance:

yes

Limit test:

no

Species:

rat

Strain:

other: Crl:CD(SD)

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River Laboratories International, Inc., Raleigh, North Carolina
- Age at study initiation: approximately 67 days
- Weight at gestation day 0:
0 mg/kg/day: 200 g - 225 g
100 mg/kg/day: 200 g - 225 g
300 mg/kg/day: 200 g - 225 g
1000 mg/kg/day: 200 g - 225 g
- Housing: animals were housed in pairs (when possible) in solid-bottom caging with Bed-o'Cobs® bedding and nestlets as enrichment.
- Diet (ad libitum): PMI® Nutrition International, LLC Certified Rodent LabDiet® 5002
- Water (ad libitum): tap water
- Quarantine period: yes, but length of period was not stated

ENVIRONMENTAL CONDITIONS
- Temperature: 20-26ºC
- Relative humidity: 30-70%
- Photoperiod: approximate 12 hour light/dark cycle

Route of administration:

oral: gavage

Vehicle:

other: sterile water for injection

Details on exposure:

PREPARATION OF DOSING SOLUTIONS:
The formulations of the test substance in the vehicle were prepared at least weekly. The stability of the dosing formulations at concentrations bracketing the concentrations administered to the animals was demonstrated for up to 8 days.
The test substance was suspended in sterile water for injection. The dosing formulations were not adjusted for purity; a purity of value of 100% was assumed for the purposes of dose formulation calculations. Dosing formulations were stored at room temperature until used.
The volume administered (5 mL/kg) was based on the most recent body weight.

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

Samples of pigment grade titanium dioxide (TiO2 pg-1) in water at concentrations of a control, 20, 60 and 200 mg/mL were analyzed for homogeneity and concentration verification. Samples of each formulation were taken 2 times: near the beginning and end of the study. Dose samples at the same concentrations prepared near the end of the study were analyzed for concentration verification. Samples were digested in a chemical microwave using nitric and hydrofluoric acid and then analyzed by Inductively Coupled Plasma-Atomic Emission Spectroscopy (ICP-AES).
The analysis results for samples prepared at the beginning of the study show that the test substance was homogeneously mixed as represented by top, middle and bottom samples from each dose preparation. The mean concentrations were 19.6, 64.6, and 198 mg/mL of the dose preparations.
The measured values were 98.0%, 108%, and 99.0% of nominal, all within the acceptable range of 80 to 120% of the respective targeted suspension concentrations of 20, 60, and 200 mg/mL samples.
The analysis results for samples prepared at the end of the study show that the test substance was at the targeted concentrations for each dose preparation. The concentrations were 19.9, 59.9 and 203 mg/mL for the dose preparations. The measured values were 99.7%, 99.9% and 102% of nominal, all within the acceptable range of 80-120% of the respective targeted suspension concentrations of 20, 60, and 200 mg/mL samples.
The test substance was not detected in the control samples.

Details on mating procedure:

- Impregnation procedure: timed mated

Duration of treatment / exposure:

Gestation days 6 - 20

Frequency of treatment:

once daily

Duration of test:

20 days

Dose / conc.:

100 mg/kg bw/day (actual dose received)

Dose / conc.:

300 mg/kg bw/day (actual dose received)

Dose / conc.:

1 000 mg/kg bw/day (actual dose received)

No. of animals per sex per dose:

22 timed mated females

Control animals:

yes, concurrent vehicle

Maternal examinations:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule:
Quarantine and pretest: mortality/moribundity at least once daily; clinical observation on gestation day 4
Testing period: mortality/moribundity twice daily (morning and evening); clinical observations twice daily on gestation days 6 - 20 (during weighing and at least 2 hours post-dosing) and once on gestation day 21.

DETAILED CLINICAL No

BODY WEIGHT: Yes
- Time schedule for examinations:
Quarantine and pretest: gestation day 4
Testing period: daily on gestation days 6 - 21

FOOD CONSUMPTION AND COMPOUND INTAKE (if feeding study): Yes
- Food consumption for each animal determined and mean daily diet consumption calculated as g food/animal/day: Yes
Quarantine and pretest: gestation day 4

Testing period: gestation days 6, 8, 10, 12, 14, 16, 18, 20, and 21
Food consumption was measured as food consumed per cage, which was then divided by the number of animals in each cage for individual food consumption values. Excess food spillage was recorded.

- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: No

WATER CONSUMPTION AND COMPOUND INTAKE: No

POST-MORTEM EXAMINATIONS: Yes
- Sacrifice on gestation day 21
- Organs examined: a gross external and a visceral examination were performed immediately after euthanasia.

Ovaries and uterine content:

The ovaries and uterine content was examined after termination: Yes
Examinations included:
- Gravid uterus weight: Yes
- Number of corpora lutea: Yes
- Number of implantations: Yes
- Number of early resorptions: Yes
- Number of late resorptions: Yes
- Other: uteri with no visible implantation sites were placed in a 10% aqueous solution of ammonium sulfide to detect very early resorptions.

Fetal examinations:

- External examinations: Yes, all foetuses classified as live were examined for alterations. External sex was recorded for each live foetus.
- Soft tissue examinations: Yes, approximately half of the live foetuses
- Skeletal examinations: Yes, all live foetuses; The skeletal bodies of all the foetuses and the skulls of half the foetuses (foetuses that were not designated for head examination) were examined for alterations.
- Head examinations: Yes, approximately half of the foetuses
- Body weight of each live foetus was recorded.

Statistics:

Please refer to the field "Any other information on materials and methods incl. tables" below.

Indices:

no data

Historical control data:

no data

Clinical signs:

no effects observed

Dermal irritation (if dermal study):

not examined

Mortality:

no mortality observed

Body weight and weight changes:

no effects observed

Food consumption and compound intake (if feeding study):

no effects observed

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

not examined

Ophthalmological findings:

not examined

Haematological findings:

not examined

Clinical biochemistry findings:

not examined

Urinalysis findings:

not examined

Behaviour (functional findings):

not examined

Immunological findings:

not examined

Organ weight findings including organ / body weight ratios:

not examined

Gross pathological findings:

no effects observed

Neuropathological findings:

no effects observed

Histopathological findings: non-neoplastic:

not examined

Histopathological findings: neoplastic:

not examined

Other effects:

not examined

Details on results:

- no test substance-related mortality at any level tested; all animals on study survived until scheduled euthanasia.
- no test substance-related clinical observations at any level tested; the observations that were recorded were unremarkable and occurred infrequently.
- no test substance-related effects on maternal body weight parameters at any level tested; data for maternal body weights and weight changes were comparable across all dose levels tested. Mean final body weights (absolute or adjusted) GD 21 were within 2% of the control group mean at every dose level tested.
- no test substance-related effects on maternal food consumption at any level tested. Mean maternal food consumption from GD 6-21 was within 3% of the control group mean for all dose levels tested.
- 100 mg/kg/day: mean food consumption from gestation days 6-8 was slightly increased relative to the control mean. At 300 mg/kg/day, mean food consumption from days 14-16 and 16-18 of gestation were slightly lower than the respective control means. These instances were statistically significant; however, they are considered spurious and unrelated to the test substance. The changes are minimal in magnitude, variably increased or decreased, not dose dependent, and had no impact on cumulative food consumption values (gestation days 6-21) for these groups.
- no test substance-related maternal gross postmortem observations at any level tested. All animals appeared normal at the group at necropsy.

Number of abortions:

not examined

Pre- and post-implantation loss:

no effects observed

Total litter losses by resorption:

not examined

Early or late resorptions:

no effects observed

Dead fetuses:

no effects observed

Changes in pregnancy duration:

not specified

Changes in number of pregnant:

not specified

Other effects:

not examined

Details on maternal toxic effects:

- no test substance-related effects on any reproductive outcome endpoint.
- litter means for numbers of implantation sites, early and late resorptions were all comparable to control group values for every dose level tested.

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Fetal body weight changes:

no effects observed

Reduction in number of live offspring:

not examined

Changes in sex ratio:

no effects observed

Changes in litter size and weights:

not examined

Changes in postnatal survival:

not examined

External malformations:

no effects observed

Skeletal malformations:

no effects observed

Visceral malformations:

no effects observed

Other effects:

not examined

Details on embryotoxic / teratogenic effects:

- Litter means for numbers of live and dead foetuses as well as for foetal weight and sex ratio were all comparable to control group values for every dose level tested.
-100 mg/kg/day dose level: mean foetal weight was significantly increased relative to the control group mean. This increase was considered to be spurious and unrelated to the test substance because it was minimal in magnitude and not dose dependent. Further, an increase in mean foetal weight is not typically considered to reflect an adverse outcome as a decrease in this endpoint might.
- no test substance-related foetal malformations or variations observed at any dose level tested. The foetal alterations occurred with low frequency across all groups tested and occur with similar frequency in the test facility historical control database.

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Developmental effects observed:

not specified

Conclusions:

Under the conditions of this study, there was no evidence of either maternal or developmental toxicity at doses up to 1000 mg/kg/day. Therefore, the no-observed-adverse-effect level (NOAEL) for maternal and developmental toxicity is at the limit dose of 1000 mg/kg/day.

Reference 3

Endpoint:

developmental toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2013-05-06 to 2013-07-24

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 414 (Prenatal Developmental Toxicity Study)

Version / remarks:

2001-01-22

Deviations:

no

GLP compliance:

yes (incl. QA statement)

Remarks:

signed 2013-01-22

Limit test:

no

Species:

rat

Strain:

Wistar

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River, 97633 Sulzfeld, Germany
- Age at the start of the treatment period: 11-12 weeks old
- Weight at the initiation of pairing: males: 306 – 348 g (mean: 323.02 g, ± 20% = 258.42 – 387.62 g); females: 188 – 222 g (mean: 205.44 g, ± 20% = 164.35 – 246.52 g)
- Housing: kept individually in IVC cages (except during the mating period when two females were paired with one male), type III H, polysulphone cages on Altromin saw fibre bedding (lot Nr. 240113)
- Diet (ad libitum): Altromin 1324 maintenance diet for rats and mice (lot Nr. 0902)
- Water (ad libitum): tap water, sulphuric acid acidified to a pH of approximately 2.8
- Acclimation period: at least 5 days

ENVIRONMENTAL CONDITIONS
- Temperature: 22 ± 3°C
- Humidity: 55 ± 10%
- Air changes: 10 x / hour
- Photoperiod (hrs dark / hrs light): 12/12

Route of administration:

oral: gavage

Vehicle:

other: aqua ad iniectabilia (water for injection)

Details on exposure:

PREPARATION OF DOSING SOLUTIONS:
The test item was weighed into a tarred plastic vial on a precision balance.
The test item was suspended in the vehicle.
The vehicle was employed based on the test item’s characteristics and testing guideline. The test item and control formulations were prepared freshly on each administration day immediately prior to dosing.
Homogeneity of the test item in the vehicle was maintained by vortexing the prepared suspension thoroughly before every dose administration.
The application volume for all groups was 5 mL/kg bw/day.
For each animal the individual dosing volume was calculated on the basis of the body weight most recently measured on various gestation days.

VEHICLE
- Batch no.: 2621051-2

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

The assessment of homogeneity, stability as well as a determination of the nominal concentration of the test item in the vehicle was performed at various intervals.
Samples for analysis of nominal concentration of the dose formulations of the test item in the vehicle were taken in the first and last week of the study for all doses (8 samples in total).
Samples for homogeneity were taken from the top, middle and bottom of the high dose, medium dose and low dose preparation. Samples were taken in the first and last week of the study (18 samples in total).
Samples for stability analysis were taken from the low dose, medium dose and high dose group in study week 1 at 0 hr and 6 hrs (6 samples in total).
All formulation samples were stored at -20° C.

Results:
Determination of the nominal substance concentrations (in week 1 and last week of the study), stability analysis (0 and 6 hours in week 1) and homogeneity (samples from the top, middle and bottom of container in week 1 and last week of the study) were performed and the results revealed that mean recovery values for TiO2 at all fortification levels obtained by ICP-OES were between 81.1 to 97.1 which complied with the standard acceptance criteria of SANCO/3029/99, which demands that the mean recovery at each fortification level should be in the range of 70% - 110%.
The relative standard deviations per fortification level ranged from 0.52% to 3.16% (overall: 2.38%) for TiO2 analysis in dose formulations. These values meet the guideline requirements of SANCO/3029/99 rev. 4 (RSD ≤ 20%).

Details on mating procedure:

- Impregnation procedure: cohoused
- If cohoused/M/F ratio per cage: females were paired with males as per the ratio of 1:2 (male to female).
- Proof of pregnancy: sperm in vaginal smear referred to as day 0 of gestation

Duration of treatment / exposure:

Gestation day 5 through gestation day 19

Frequency of treatment:

once daily, 7 days per week

Duration of test:

20 days

Dose / conc.:

100 mg/kg bw/day (actual dose received)

Dose / conc.:

300 mg/kg bw/day (actual dose received)

Dose / conc.:

1 000 mg/kg bw/day (actual dose received)

No. of animals per sex per dose:

24-25 mated female rats

Control animals:

yes, concurrent vehicle

Details on study design:

- Dose selection rationale: since little or no toxicity was anticipated for the test substance, the highest dose level was set at 1000 mg/kg bw/day corresponding to a limit dose for this study. Thereafter, a descending sequence of dose levels was selected with a view to demonstrate any dose-related response and a NOAEL.

Maternal examinations:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: general clinical observations: at least once a day (approximately within half an hour after application); morbidity and mortality: at least once daily.
- Cage side observations checked: spontaneous activity, lethargy, recumbent position, convulsions, tremors, apnoea, asphyxia, vocalisation, diarrhoea, changes in the skin and fur, eyes and mucous membranes (salivation, discharge), piloerection and pupil size, changes in gait, posture, response to handling as well as the presence of clonic or tonic movements, stereotypes, difficult or prolonged parturition or bizarre behaviour.

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: prior to the start of the mating, a detailed clinical observation outside the home cage was made.

BODY WEIGHT: Yes
- Time schedule for examinations: once before the assignment to the experimental groups, on the first day of administration and weekly during the treatment. The sperm positive females were weighed during gestation days 0, 5, 8, 11, 14, 17 and 20.

FOOD CONSUMPTION AND COMPOUND INTAKE: Yes
- Food consumption for each animal determined: Yes, food consumption of pregnant females was measured on gestation days 0, 5, 8, 11, 14, 17 and 20.
- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: No

WATER CONSUMPTION AND COMPOUND INTAKE: No

POST-MORTEM EXAMINATIONS: Yes
- Sacrifice on gestation day #20
At the time of termination, each dam (presumed pregnant female) was examined macroscopically for any structural abnormalities or pathological changes which may have influenced the pregnancy.
- Organs examined: immediately after the termination, the uteri were removed and the pregnancy status of the dams was confirmed. Uteri that appeared non-gravid were further examined by staining with 10 % ammonium sulphide solution to confirm the non-pregnant status.

Ovaries and uterine content:

The ovaries and uterine content was examined after termination: Yes
Examinations included:
- Gravid uterus weight: Yes, with cervix
- Number of corpora lutea: Yes
- Number of implantations: Yes
- Number of early resorptions: Yes
- Number of late resorptions: Yes
- Other: the uterine contents were examined for embryonic or foetal deaths as well as the number of viable foetuses. The position and number of foetuses in each uterine horn were also recorded.

Fetal examinations:

- External examinations: Yes, all per litter
- Soft tissue examinations: Yes, half per litter
- Skeletal examinations: Yes, half per litter
- Head examinations: Yes, craniofacial examination of the heads of the foetuses used for the soft tissue examination
All foetuses were weighed and sexed based on the anogenital distance.

Statistics:

A statistical assessment of the results of the body weight and food consumption was performed by comparing values of dosed with control animals using a one-way ANOVA and a post-hoc Dunnett Test. Foetal evaluation parameters like external, visceral, craniofacial and skeletal parameters were analysed using a Chi-square test. The statistics were performed with GraphPad Prism V.6.01 software (p<0.05 is considered as a statistically significant).

Indices:

no data

Historical control data:

Historical control data is provided for the following:
- mean uterine data
- mean litter weight (g) data
- foetal external examination
- foetal visceral examination
- foetal craniofacial examination
- foetal skeletal examination

Clinical signs:

no effects observed

Dermal irritation (if dermal study):

not examined

Mortality:

no mortality observed

Body weight and weight changes:

no effects observed

Food consumption and compound intake (if feeding study):

no effects observed

Food efficiency:

not examined

Water consumption and compound intake (if drinking water study):

not examined

Ophthalmological findings:

not examined

Haematological findings:

not examined

Clinical biochemistry findings:

not examined

Urinalysis findings:

not examined

Behaviour (functional findings):

not examined

Immunological findings:

not examined

Organ weight findings including organ / body weight ratios:

not examined

Gross pathological findings:

no effects observed

Neuropathological findings:

no effects observed

Histopathological findings: non-neoplastic:

not examined

Histopathological findings: neoplastic:

not examined

Other effects:

not examined

Details on results:

- none of the females showed signs of abortion or premature delivery prior to the scheduled terminal sacrifice.

- mortality: all sperm positive females survived until the scheduled necropsy.

- clinical signs: no test item-related clinical signs were observed in the study.

- body weight development: statistical analysis of body weight and body weight gain data during the gestation period revealed no statistically significant effect on body weight development in treatment groups when compared with the controls.

- food consumption: no statistically significant effect of TiO2 pg-3 on food consumption was observed in any of the treatment groups when compared with controls and in correlation to the body weight and body weight gain, the food consumption was comparable in all groups.

- pathology: no macroscopic findings were observed in any female of the control and treatment groups at necropsy.

Please also refer to the field "Attached background material" below.

Number of abortions:

no effects observed

Pre- and post-implantation loss:

no effects observed

Early or late resorptions:

no effects observed

Dead fetuses:

no effects observed

Changes in pregnancy duration:

no effects observed

Changes in number of pregnant:

no effects observed

Other effects:

not examined

Details on maternal toxic effects:

- statistical analysis of prenatal data revealed no significant effect on prenatal parameters like gravid uterus weight, adjusted maternal weight, number of corpora lutea, number of implantation sites, resorptions (early and late), percent preimplantation loss, or post implantation loss in the treated groups when compared with the controls.

- successful mating resulted in 20/23 pregnancies in the control, 21/23 pregnancies in 100 mg/kg bw/day, 22/23 in 300 mg/kg bw/day and 21/22 in 1000 mg/kg bw/day groups. Nine females were not mated and therefore excluded from the study without any further observations.

- pregnancy rates (Nr. of pregnancies achieved /Nr. of females mated or sperm positive x 100) in treatment groups 100 mg/kg bw/day, 300 mg/kg bw/day and 1000 mg/kg bw/day were 91.30 %, 95.65% and 95.45%, respectively, compared to 86.96 % in control group. The differences in pregnancy rates between the 100 mg/kg bw/day and 300 mg/kg bw/day groups and the control were considered to be due to biological variation and of no toxicological relevance.

Please also refer to the field "Attached background material" below.

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Fetal body weight changes:

not examined

Reduction in number of live offspring:

not examined

Changes in sex ratio:

no effects observed

Changes in litter size and weights:

no effects observed

Changes in postnatal survival:

not examined

External malformations:

no effects observed

Skeletal malformations:

no effects observed

Visceral malformations:

no effects observed

Other effects:

not examined

Details on embryotoxic / teratogenic effects:

- statistical analysis of prenatal data revealed no significant effect on prenatal parameters like live foetuses, dead foetuses, sex ratio, group mean number of male and female foetuses in the treated groups when compared with the controls.

- no statistically significant or treatment-related effects were observed in any treatment group as compared with the control on group mean litter weight, total litter weight, male litter weight, female litter weight, group mean number of live foetuses, or number of males and number of females.

- no test item related gross external abnormalities were seen in either the control or treatment groups. However, some few and unremarkable abnormalities were observed, generally in single foetuses, and were considered to be incidental in nature and unrelated to treatment.

- skeletal examination revealed a range of abnormalities which were generally of a type or which occurred at an incidence comparable to or lower in treated groups when compared to the control group or were seen only at the 300 mg/kg bw/day or 100 mg/kg bw/day treated groups and were not dose dependent.

- statistically significant increases were observed in the incidences of unossified 4th forelimb phalanx (L) and unossified 5th hindlimb phalanx (R) in the 1000 mg/kg bw/day group, wavy ribs in the 300 mg/kg bw/day and 1000 mg/kg bw/day groups, incomplete ossification on the interparietal and incomplete ossification of the right parietal. In all cases, the incidence rate of these findings were within recent historical control ranges reported in the laboratory. Increases in incidences of wavy ribs and incomplete ossification are considered to be variations and not associated with long term consequences on survival, general growth and development.

- internal examinations of foetal viscera revealed a range of visceral abnormalities in all groups including the control. However, there was a statistically significant increase in incidence of hemorrhagic bladder content in the 1000 mg/kg bw/day group when compared with the controls. The bladder finding is most likely attributable to a dissection artefact during visceral examinations and it is not a developmental variation and therefore not considered to be adverse.
The remaining visceral abnormalities observed in the treated groups were at frequencies generally comparable to or in some cases slightly higher or lower in frequency compared to the controls.

- craniofacial examination revealed a range of visceral abnormalities in all groups including controls. Statistical analysis of the data revealed no significant effect in any of the findings when compared with controls.

Please also refer to the field "Attached background material" below.

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Developmental effects observed:

not specified

Conclusions:

In this prenatal developmental toxicity study, the repeated dose oral administration of TiO2 pg-3 to pregnant female Wistar rats at doses of 100, 300 and 1000 mg/kg bw/day on gestation days 5 to 19 produced no significant toxicological effects in the females or foetuses or significant developmental effects at any administered dose.
Based on the findings from this study, the NOAEL for TiO2 pg-3 for both maternal and developmental toxicity in the Wistar rat is considered to be 1000 mg/kg bw/day.

Reference 4

Endpoint:

developmental toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2013-07-15 to 2013-08-22

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

other: GLP guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 414 (Prenatal Developmental Toxicity Study)

Version / remarks:

adopted 2001-01-22

Deviations:

no

GLP compliance:

yes (incl. QA statement)

Remarks:

signed 2013-01-22

Limit test:

no

Species:

rat

Strain:

Wistar

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River, 97633 Sulzfeld, Germany
- Age at the start of the treatment period: 11-12 weeks old
- Weight at the initiation of pairing: males: 287-330 g (mean: 306.7 g, ± 20% = 368.04 – 245.36 g); females: 174-223 g (mean: 203.99 g, ± 20% = 244.79 – 163.19 g)
- Housing: kept individually in IVC cages (except during the mating period when two females were paired with one male), type III H, polysulphone cages on Altromin saw fibre bedding (lot Nr. 240113)
- Diet (ad libitum): Altromin 1324 maintenance diet for rats and mice (lot Nr. 1531)
- Water (ad libitum): tap water, sulphuric acid acidified to a pH of approximately 2.8
- Acclimation period: at least 5 days

ENVIRONMENTAL CONDITIONS
- Temperature: 22 ± 3°C
- Humidity: 55 ± 10%
- Air changes: 10 x / hour
- Photoperiod (hrs dark / hrs light): 12/12

Route of administration:

oral: gavage

Vehicle:

other: aqua ad iniectabilia (water for injection)

Details on exposure:

PREPARATION OF DOSING SOLUTIONS:
The test item formulation was prepared freshly on each administration day immediately prior to dosing.
The test item was weighed into a tared plastic vial on a suitable precision balance and the vehicle was added to give the appropriate final concentration of the test item. The formulation was vortexed for 2-3 minutes.
Homogeneity of the test item in the vehicle was maintained by vortexing the prepared suspension thoroughly before every dose administration.
Application volume for all groups was 5 mL/kg body weight.
For each animal the individual dosing volume was calculated on the basis of the body weight most recently measured (measured weekly).

VEHICLE
- Batch no.: 26210S1-2

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

The assessment of homogeneity, stability as well as a determination of the nominal concentration of the test item in the vehicle was performed at various intervals.
Samples for the analysis of nominal concentration of the dose formulations of the test item in the vehicle were taken in the first and last week of the study for all doses (8 samples in total).
Samples for the analysis of homogeneity were taken from the top, middle and bottom of the high dose, medium dose and low dose preparation. Samples were taken in the first and last week of the study (18 samples in total).
Samples for the stability analysis were taken from the low dose, medium dose and high dose group in study week 1 at 0 hr and 6 hrs (6 samples in total).
All formulation samples were stored at -20° C.

The analytical method for the determination of titanium in dose formulation samples was successfully validated according to SANCO/3029/99 rev. 4 (11/07/00).
The determinations were performed by ICP-OES using two independent emission wavelengths highly specific for titanium, one wavelength for the quantification and one for the confirmation.
The requirements of SANCO/3029/99 rev. 4 (11/07/00) regarding linearity, precision (repeatability), accuracy (recovery) and specificity were fulfilled.
The mean recovery values for titanium at all fortification levels obtained by ICP-OES comply with the standard acceptance criteria of SANCO/3029/99, which demands that the mean recovery at each fortification level should be in the range of 70% - 110%.
Test substance concentration, stability and homogeneity in dosing formulations were obtained by ICP-OES.

Results:
The analytical results obtained for the individual dose groups were consistent with the analysis of the % of nominal of the test item for the concentration, stability and homogeneity analyses, with the exception of the sample number 13 (LD, week 1) for which recoveries were 125.8 and 122.8 (‘a’ and ‘b’, respectively). This increase in the concentration was noticed in one sample of the LD group meant for stability analysis for 0 hours. This isolated result was considered to be due to sampling error.
In this case the formulations need to be vortexed for approximately 2 minutes before sampling. The dose formulation sample preparation was made every day freshly before the dose administration. On the day of analytical sample collection the formulation samples were prepared in excess and the analytical samples were collected from the same stock as that used for the dose administration. The formulation sample was assumed to have not been subjected to a proper homogenization procedure before the collection of sample number 13. This was evident from the sample numbers 4, 14, 21, 22 and 23 (LD, week 1) for which the recoveries were 90% to 103.3%. The sample numbers 4, 14, 21, 22 and 23 were collected from the same formulation preparation as that of sample number 13. As there were no toxicity findings seen at any of the tested dose levels, this concentration increase in an isolated analytical sample of the LD group had no impact on the validity and integrity of the study.

Details on mating procedure:

- Impregnation procedure: cohoused
- If cohoused/M/F ratio per cage: females were paired with males as per the ratio of 1:2 (male to female).
- Proof of pregnancy: sperm in vaginal smear referred to as day 0 of gestation

Duration of treatment / exposure:

Gestation day 5 through gestation day 19

Frequency of treatment:

once daily, 7 days per week

Duration of test:

20 days

Dose / conc.:

100 mg/kg bw/day (actual dose received)

Dose / conc.:

300 mg/kg bw/day (actual dose received)

Dose / conc.:

1 000 mg/kg bw/day (actual dose received)

No. of animals per sex per dose:

24-25 mated female rats

Control animals:

yes, concurrent vehicle

Details on study design:

- Dose selection rationale: since little or no toxicity was anticipated for the test substance, the highest dose level was set at 1000 mg/kg bw/day corresponding to a limit dose for this study. Thereafter, a descending sequence of dose levels was selected with a view to demonstrate any dose-related response and a NOAEL.

Maternal examinations:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: general clinical observations: once a day; morbidity and mortality: twice daily, except during holidays and weekends where the observation was made once daily.
- Cage side observations checked: spontaneous activity, lethargy, recumbent position, convulsions, tremors, apnoea, asphyxia, vocalisation, diarrhoea, changes in the skin and fur, eyes and mucous membranes (salivation, discharge), piloerection and pupil size, changes in gait, posture, response to handling as well as the presence of clonic or tonic movements, stereotypes, difficult or prolonged parturition or bizarre behaviour.

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: prior to the start of the mating, a detailed clinical observation outside the home cage was made.

BODY WEIGHT: Yes
- Time schedule for examinations: once before the assignment to the experimental groups, on the first day of administration and weekly during the treatment. The sperm positive females were weighed during gestation days 0, 5, 8, 11, 14, 17 and 20.

FOOD CONSUMPTION AND COMPOUND INTAKE: Yes
- Food consumption for each animal determined: Yes, food consumption of pregnant females was measured on gestation days 5, 8, 11, 14, 17 and 20.
- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: No

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

POST-MORTEM EXAMINATIONS: Yes
- Sacrifice on gestation day #20
At the time of termination, each dam (presumed pregnant female) was examined macroscopically for any structural abnormalities or pathological changes which may have influenced the pregnancy.
- Organs examined: immediately after the termination, the uteri were removed and the pregnancy status of the dams was confirmed. Uteri that appeared non-gravid were further examined by staining with 10 % ammonium sulphide solution to confirm the non-pregnant status.

Ovaries and uterine content:

The ovaries and uterine content was examined after termination: Yes
Examinations included:
- Gravid uterus weight: Yes, with cervix
- Number of corpora lutea: Yes
- Number of implantations: Yes
- Number of early resorptions: Yes
- Number of late resorptions: Yes
- Other: the uterine contents were examined for embryonic or foetal deaths as well as the number of viable foetuses. The position and number of foetuses in each uterine horn were also recorded.

Fetal examinations:

- External examinations: Yes, all per litter
- Soft tissue examinations: Yes, half per litter
- Skeletal examinations: Yes, half per litter
- Head examinations: Yes, craniofacial examination of the heads of the foetuses used for the soft tissue examination

All foetuses were weighed and sexed based on the anogenital distance.

Statistics:

A statistical assessment of the results of the body weight and food consumption was performed by comparing values of dosed with control animals using a one-way ANOVA and a post-hoc Dunnett Test. Foetal evaluation parameters like external, visceral, craniofacial and skeletal parameters were analysed using a Chi-square test. The statistics were performed with GraphPad Prism V.6.01 software (p<0.05 is considered as statistically significant).
For the abnormality “hindlimb talus- unossified” the statistical analyses was performed by combining all unilateral and bilateral findings.

Indices:

no data

Historical control data:

Historical control data is provided for the following:
- mean uterine data
- mean litter weight (g) data
- foetal external examination
- foetal visceral examination
- foetal craniofacial examination
- foetal skeletal examination

Details on maternal toxic effects:

Maternal toxic effects:no effects

Details on maternal toxic effects:
- none of the females showed signs of abortion or premature delivery prior to the scheduled terminal sacrifice.

- mortality: no deaths recorded during the course of the study in any of the groups.

- clinical observations: no clinical signs of toxicological relevance noted during the study.

- body weight development: no treatment related effect noted for body weight and body weight change in the treated groups in comparison to the controls. No treatment related changes noted for terminal and adjusted maternal body weight in the treated groups in comparison to the controls. The statistical analysis of the body weight data showed no statistical significance.

- food consumption: no treatment related effect noted for food consumption in the treated groups in comparison to the controls. The statistical analysis of the food consumption data showed no statistically significant changes between the treated and the control group.

- no treatment related changes noted for prenatal parameters including number of corpora lutea, number of implantation sites, early and late resorptions, or pre- and post-implantation loss.
However, there was a slight decrease in the uterus weight, number of implantation sites and number of live foetuses in the 100 mg/kg bw/day and 1000 mg/kg bw/day groups in comparison to the control. In the absence of statistical significance, dose related response and the values being within the historical control data range, the changes were considered to be incidental.
There was an increase in pre implantation loss in the 100 mg/kg bw/day and 1000 mg/kg bw/day groups. As the dose administration begins from gestation day 5, the changes noted for pre implantation loss is not associated with the treatment.

- there were no macroscopic findings noted in any of the animals of the control or the test item treated groups at necropsy.

Please also refer to the field "Attached background material" below.

Dose descriptor:

NOAEL

Effect level:

1 000 mg/kg bw/day (actual dose received)

Based on:

test mat.

Basis for effect level:

other: maternal toxicity

Details on embryotoxic / teratogenic effects:

Embryotoxic / teratogenic effects:no effects

Details on embryotoxic / teratogenic effects:
- no treatment related changes noted for prenatal parameters including live foetuses, number of dead foetuses, number of male and female foetuses, or sex ratio.

-there was a slight decrease in the number of male foetuses in the 1000 mg/kg bw/day group, which was due to the absence of male foetuses in 2/ 24 females in the 1000 mg/kg bw/day group. As this was observed in two isolated females and also was within the historical control data range, the finding was considered incidental and not associated with treatment. Because foetal sex is determined shortly after conception and well before the onset of dosing on gestation day 5, such changes in sex ratio are not considered to be indicative of a test substance-related effect.
The statistical analysis of data showed no statistically significant changes between the treated and the corresponding control group.

- no effects of treatment noted for the litter data including mean litter weight, total litter weight and male and female litter weight.
However, there was a slight decrease in the total and female litter weights noted in the 100 mg/kg bw/day and 1000 mg/kg bw/day groups. In the absence of the statistical significance and a dose related response and in addition, all mean values being within historical control ranges, the changes were not considered to be associated with treatment.
Furthermore, there was also a decrease (-13.75%) in the mean male litter weight noted in the 1000 mg/kg bw/day group in comparison to the control, which corresponded to the decreased number of male foetuses in the 1000 mg/kg bw/day group. But this finding was due to the absence of male foetuses in 2/ 24 females in the 1000 mg/kg bw/day group. As this was noted in two isolated females and also taking into account the data being within the historical control data range the finding was considered to be incidental and not associated with the treatment.

- no test item related gross external abnormalities were seen in either the control or treatment groups. There were no statistically significant changes noted. However, There were abnormalities like hematoma of the neck, flank and tail in the control and hematoma of the forelimb in the 1000 mg/kg bw/day group. These were observed in a single isolated foetus of a single female of the control and/ or 1000 mg/kg bw/day groups and were considered spontaneous in origin. Micrognathia was observed in a single foetus of a single female of the 100 mg/kg bw/day group and was considered to be spontaneous in origin.

- skeletal examination foetuses revealed a range of abnormalities in the control and treated groups that were either within historical control ranges recorded for this laboratory; were significantly lower than the corresponding control values; or were seen only in the 300 mg/kg bw/day or 100 mg/kg bw/day dose groups and were not dose dependent.
There was statistically significant increase in 14th full rib-bilateral in the 100 mg/kg bw/day and 1000 mg/kg bw/day groups. The percent foetal incidence of 14th full rib- bilateral was within the historical control data range and there was no dose related response observed. This abnormality was considered to have no relevance to treatment.

- internal examinations revealed a range of visceral abnormalities in all groups including the control. There were no abnormalities of toxicological relevance.

- craniofacial examination revealed a range of abnormalities in all groups including the controls. There were no abnormalities of toxicological relevance. The statistical analysis of data showed no significant changes.

Please also refer to the field "Attached background material" below.

Dose descriptor:

NOAEL

Effect level:

1 000 mg/kg bw/day (actual dose received)

Based on:

test mat.

Sex:

male/female

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Developmental effects observed:

not specified

Conclusions:

In this prenatal developmental toxicity study, the repeated dose oral administration of TiO2 uf-2 to pregnant female Wistar rats at doses of 100, 300 and 1000 mg/kg bw/ day from gestation day 5 through gestational day 19 produced no adverse toxicological effects in the females or foetuses or significant developmental effects at any administered dose.
Based on the findings from this study, the NOAEL (No-Observed-Adverse-Effect-Level) for TiO2 uf-2 for both maternal and developmental toxicity in the Wistar rat is considered to be 1000 mg/kg bw/day.

Reference 5

Endpoint:

developmental toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2013-05-05 to 2014-04-22

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

other: GLP guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 414 (Prenatal Developmental Toxicity Study)

Version / remarks:

adopted 2001-01-22

Deviations:

no

GLP compliance:

yes

Limit test:

no

Species:

rat

Strain:

other: Crl:CD(SD)

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River Laboratories International, Inc., Raleigh, North Carolina
- Age at study initiation: approximately 62-66 days
- Weight at gestation day 0:
0 mg/kg/day: 201 g - 225 g
100 mg/kg/day: 200 g - 225 g
300 mg/kg/day: 200 g - 225 g
1000 mg/kg/day: 200 g - 225 g
- Housing: animals were housed in pairs in solid-bottom caging with Bed-o'Cobs® bedding and nestlets as enrichment.
- Diet (ad libitum): PMI® Nutrition International, LLC Certified Rodent LabDiet® 5002
- Water (ad libitum): tap water
- Quarantine period: yes, but length of period was not stated

ENVIRONMENTAL CONDITIONS
- Temperature: 20-26ºC
- Relative humidity: 30-70%
- Photoperiod: approximate 12 hour light/dark cycle

Route of administration:

oral: gavage

Vehicle:

other: sterile water for injection

Details on exposure:

PREPARATION OF DOSING SOLUTIONS:
The formulations of the test substance in the vehicle were prepared at least weekly. The stability of the dosing formulations at concentrations bracketing the concentrations administered to the animals was demonstrated for up to 8 days.
The test substance was suspended in sterile water for injection. The dosing formulations were not adjusted for purity; a purity of value of 100% was assumed for the purposes of dose formulation calculations. Dosing formulations were stored at room temperature until used.
The volume administered (5 mL/kg) was based on the most recent body weight.

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

Samples of ultrafine titanium dioxide (TiO2 uf-1) at concentrations of 20, 60 and 200 mg/mL were analysed for homogeneity and concentration verification. Samples of each formulation were taken 2 times: near the beginning and end of the study. Dose samples at the same concentrations prepared near the end of the study were analysed for concentration verification. A 0 mg/mL control sample containing the vehicle only, sterile water, was included and analysed with each set of study samples.
Samples were prepared in a chemical microwave using nitric and hydrofluoric acid and then analysed by Inductively Coupled Plasma-Atomic Emission Spectroscopy (ICP-AES).
The analysis results show that the test substance was homogeneously mixed and at the targeted concentrations.
The test substance was not detected in the control samples.

Details on mating procedure:

- Impregnation procedure: timed mated

Duration of treatment / exposure:

Gestation days 6 - 20

Frequency of treatment:

once daily

Duration of test:

20 days

Dose / conc.:

100 mg/kg bw/day (actual dose received)

Dose / conc.:

300 mg/kg bw/day (actual dose received)

Dose / conc.:

1 000 mg/kg bw/day (actual dose received)

No. of animals per sex per dose:

22 timed mated females

Control animals:

yes, concurrent vehicle

Maternal examinations:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule:
Quarantine and pretest: mortality/moribundity at least once daily; clinical observation on gestation day 4
Testing period: mortality/moribundity twice daily (morning and evening); clinical observations twice daily on gestation days 6 - 20 (during weighing and at least 2 hours post-dosing) and once on gestation day 21.

DETAILED CLINICAL No

BODY WEIGHT: Yes
- Time schedule for examinations:
Quarantine and pretest: gestation day 4
Testing period: daily on gestation days 6 - 21

FOOD CONSUMPTION AND COMPOUND INTAKE (if feeding study): Yes
- Food consumption for each animal determined and mean daily diet consumption calculated as g food/animal/day: Yes
Quarantine and pretest: gestation day 4

Testing period: gestation days 6, 8, 10, 12, 14, 16, 18, 20, and 21
Food consumption was measured as food consumed per cage, which was then divided by the number of animals in each cage for individual food consumption values. Excess food spillage was recorded.

- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: No

WATER CONSUMPTION AND COMPOUND INTAKE: No

POST-MORTEM EXAMINATIONS: Yes
- Sacrifice on gestation day 21
- Organs examined: a gross external and a visceral examination were performed immediately after euthanasia.

Ovaries and uterine content:

The ovaries and uterine content was examined after termination: Yes
Examinations included:
- Gravid uterus weight: Yes
- Number of corpora lutea: Yes
- Number of implantations: Yes
- Number of early resorptions: Yes
- Number of late resorptions: Yes
- Other: uteri with no visible implantation sites were placed in a 10% aqueous solution of ammonium sulfide to detect very early resorptions.

Fetal examinations:

- External examinations: Yes, all foetuses classified as live were examined for alterations. External sex was recorded for each live foetus.
- Soft tissue examinations: Yes, approximately half of the live foetuses
- Skeletal examinations: Yes, all live foetuses; The skeletal bodies of all the foetuses and the skulls of half the foetuses (foetuses that were not designated for head examination) were examined for alterations.
- Head examinations: Yes, approximately half of the foetuses
- Body weight of each live foetus was recorded.

Statistics:

Please refer to the field "Any other information on materials and methods incl. tables" below.

Indices:

no data

Historical control data:

Historical control data for rat teratology studies was provided for the following parameters:
- corrected maternal weight gain
- mean food consumed
- mean foetal sex ratio

Details on maternal toxic effects:

Maternal toxic effects:no effects

Details on maternal toxic effects:
- no test substance-related mortality at any level tested; all animals on study survived until scheduled euthanasia.
- no test substance-related clinical observations at any level tested; the observations that were recorded were unremarkable and occurred infrequently.
- no adverse test substance-related effects on maternal body weight parameters at any level tested; data for maternal body weights and weight changes were comparable across all dose levels tested.
- 1000 mg/kg/day dose level: the mean body weight gain from days 20 to 21, and the cumulative weight gain (gestation days 6 to 21) calculated using the final body weight corrected for the weight of the products of conception, were significantly higher than for the control group (cumulative interval:1000 mg/kg/day group: 47.78 g (average); control group: 36.86 g (average)). This apparent increase is the result of an atypically low value for the concurrent control group. The test facility historical control mean is 56.5 g and ranges from 41.4 to 68.7 g.
- no adverse test substance-related effects on maternal food consumption at any level tested.
- during gestation days 18 to 20, the mean food consumption values for the 100 and 1000 mg/kg/day groups were significantly higher than for the control group (100 mg/kg/day: 26.6 g; 1000 mg/kg/day: 26.5 g; control group: 23.9 g). The test facility historical control mean for this interval is 28.5 g and ranges from 25.3 g to 31.1 g; therefore, these apparent increases are reflective of an atypically low concurrent control value.
- no adverse test substance-related maternal gross postmortem observations at any level tested.
- one female in the 300 mg/kg/day group was observed to have a large placenta; this is considered to be an incidental finding and was not dose-related.

Dose descriptor:

NOAEL

Effect level:

1 000 mg/kg bw/day (actual dose received)

Based on:

test mat.

Basis for effect level:

other: maternal toxicity

Details on embryotoxic / teratogenic effects:

Embryotoxic / teratogenic effects:no effects

Details on embryotoxic / teratogenic effects:
- no adverse test substance-related effects on endpoints quantitating intrauterine growth or embryo-foetal survival.
- 1000 mg/kg/day dose level: mean foetal sex ratio and the means for male and female foetuses per litter were statistically significantly different from the control group means.
male foetuses (mean): 7.2; control group: 5.7 historical control group: 5.2 to 7.4
female foetuses (mean): 4.8; control group: 6.7 historical control group: 5.8 to 8.3
mean foetal sex ratio: 60%; control group: 46%; historical control group: 43 to 53%.
These changes in sex ratio are not considered to be indicative of a test substance-related change because foetal sex is determined shortly after conception and well before the onset of dosing on gestation day 6. In the absence of any other effect on determinants of litter size such as post-implantation losses in the form of resorptions or dead foetuses, these statistical changes are spurious.
- no test substance-related foetal malformations or variations observed at any dose level tested. The foetal alterations occurred with low frequency across all groups tested and occur with similar frequency in the test facility historical control database.

Dose descriptor:

NOAEL

Effect level:

1 000 mg/kg bw/day (actual dose received)

Based on:

test mat.

Sex:

male/female

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Developmental effects observed:

not specified

Conclusions:

Under the conditions of this study, there was no evidence of either maternal or developmental toxicity at doses up to 1000 mg/kg/day. Therefore, the no-observed-effect level (NOAEL) for maternal and developmental toxicity was considered 1000 mg/kg/day.

Reference 6

Endpoint:

developmental toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2013-07-14 to 2014-04-24

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

other: GLP guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 414 (Prenatal Developmental Toxicity Study)

Version / remarks:

adopted 2001-01-22

Deviations:

no

GLP compliance:

yes

Limit test:

no

Species:

rat

Strain:

other: Crl:CD(SD)

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River Laboratories International, Inc., Raleigh, North Carolina
- Age at study initiation: approximately 62-66 days
- Weight at gestation day 0:
0 mg/kg/day: 200 g - 225 g
100 mg/kg/day: 201 g - 225 g
300 mg/kg/day: 201 g - 225 g
1000 mg/kg/day: 201 g - 221 g
- Housing: animals were housed in pairs in solid-bottom caging with Bed-o'Cobs® bedding and nestlets as enrichment.
- Diet (ad libitum): PMI® Nutrition International, LLC Certified Rodent LabDiet® 5002
- Water (ad libitum): tap water
- Quarantine period: yes, but length of period was not stated

ENVIRONMENTAL CONDITIONS
- Temperature: 20-26ºC
- Relative humidity: 30-70%
- Photoperiod: approximate 12 hour light/dark cycle

Route of administration:

oral: gavage

Vehicle:

other: sterile water for injection

Details on exposure:

PREPARATION OF DOSING SOLUTIONS:
The formulations of the test substance in the vehicle were prepared at least weekly. The stability of the dosing formulations at concentrations bracketing the concentrations administered to the animals was demonstrated for up to 8 days.
The test substance was suspended in sterile water for injection. The dosing formulations were not adjusted for purity; a purity of value of 100% was assumed for the purposes of dose formulation calculations. Dosing formulations were stored at room temperature until used.
The volume administered (5 mL/kg) was based on the most recent body weight.

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

Samples of ultrafine titanium dioxide (TiO2 uf-3) in water at concentrations of a control, 20, 60 and 200 mg/mL were analyzed for homogeneity and concentration verification. Samples of each formulation were taken 2 times: near the beginning and end of the study. Dose samples at the same concentrations prepared near the end of the study were analysed for concentration verification.
Samples were digested in a chemical microwave using nitric and hydrofluoric acid and then analyzed by Inductively Coupled Plasma-Atomic Emission Spectroscopy (ICP-AES).
The analysis results for samples prepared at the beginning of the stud show that the test substance was homogeneously mixed and at the targeted concentrations.
The analysis results for samples prepared at the end of the study show that the test substance was at the targeted concentration for the 20 mg/mL concentration and lower than expected at the 60 and 200 mg/mL concentration. The 60 mg/mL concentration found 54.3 and 48.2 mg/mL in the samples sent for analysis. The relative standard deviation is 8.34% indicating that the sample concentrations are not consistent. For the high concentration (200 mg/mL), the relative standard deviation was 5.20% also indicating that the sample concentrations are not consistent. The analyzed concentrations ranged from 83.0 to 89.4% with an average of 86.2% of the target. This finding indicates the samples were not completely homogeneous. The increased density of the test substance affects the mixing and sampling of the samples.
The test substance was not detected in the control samples.
The results from the second set of analyses reveal that the intermediate level formulation (60 mg/mL) concentration was slightly lower than intended and that the intermediate and high level formulations (60 and 200 mg/mL) were less homogeneous that expected. The instance of low concentration and the slight increase in heterogeneity do not adversely affect the study because the ability to interpret the animal endpoint data is not impaired. Even considering an occasional increase in variability of the formulation concentration at either the intermediate or high dose levels, the spacing of the dose levels was adequate for discerning any potential effects of dose if any dose-related outcomes had been present. There were no test substance-related effects at any dose level on either maternal or foetal endpoints.

Details on mating procedure:

- Impregnation procedure: timed mated

Duration of treatment / exposure:

Gestation days 6 - 20

Frequency of treatment:

once daily

Duration of test:

20 days

Dose / conc.:

100 mg/kg bw/day (actual dose received)

Dose / conc.:

300 mg/kg bw/day (actual dose received)

Dose / conc.:

1 000 mg/kg bw/day (actual dose received)

No. of animals per sex per dose:

22 timed mated females

Control animals:

yes, concurrent vehicle

Maternal examinations:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule:
Quarantine and pretest: mortality/moribundity at least once daily; clinical observation on gestation day 4
Testing period: mortality/moribundity twice daily (morning and evening); clinical observations twice daily on gestation days 6 - 20 (during weighing and at least 2 hours post-dosing) and once on gestation day 21.

DETAILED CLINICAL No

BODY WEIGHT: Yes
- Time schedule for examinations:
Quarantine and pretest: gestation day 4
Testing period: daily on gestation days 6 - 21

FOOD CONSUMPTION AND COMPOUND INTAKE (if feeding study): Yes
- Food consumption for each animal determined and mean daily diet consumption calculated as g food/animal/day: Yes
Quarantine and pretest: gestation day 4

Testing period: gestation days 6, 8, 10, 12, 14, 16, 18, 20, and 21
Food consumption was measured as food consumed per cage, which was then divided by the number of animals in each cage for individual food consumption values. Excess food spillage was recorded.

- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: No

WATER CONSUMPTION AND COMPOUND INTAKE: No

POST-MORTEM EXAMINATIONS: Yes
- Sacrifice on gestation day 21
- Organs examined: a gross external and a visceral examination were performed immediately after euthanasia.

Ovaries and uterine content:

The ovaries and uterine content was examined after termination: Yes
Examinations included:
- Gravid uterus weight: Yes
- Number of corpora lutea: Yes
- Number of implantations: Yes
- Number of early resorptions: Yes
- Number of late resorptions: Yes
- Other: uteri with no visible implantation sites were placed in a 10% aqueous solution of ammonium sulfide to detect very early resorptions.

Fetal examinations:

- External examinations: Yes, all foetuses classified as live were examined for alterations. External sex was recorded for each live foetus.
- Soft tissue examinations: Yes, approximately half of the live foetuses
- Skeletal examinations: Yes, all live foetuses; The skeletal bodies of all the foetuses and the skulls of half the foetuses (foetuses that were not designated for head examination) were examined for alterations.
- Head examinations: Yes, approximately half of the foetuses
- Body weight of each live foetus was recorded.

Statistics:

Please refer to the field "Any other information on materials and methods incl. tables" below.

Indices:

no data

Historical control data:

no data

Details on maternal toxic effects:

Maternal toxic effects:no effects

Details on maternal toxic effects:
- no test substance-related mortality at any level tested; all animals on study survived until scheduled euthanasia.
- no test substance-related clinical observations at any level tested; the observations that were recorded were unremarkable and occurred infrequently.
- no test substance-related effects on maternal body weight parameters at any level tested; data for maternal body weights and weight changes were comparable across all dose levels tested. Mean final body weights (absolute or adjusted) on GD 21 were within 1% of the control group mean at every dose level tested.
- no test substance-related effects on maternal food consumption at any level tested. Mean maternal food consumption from GD 6-21 was within 3% of the control group mean for all of the dose levels tested.
- no test substance-related maternal gross postmortem observations at any level tested. All animals appeared normal at the group at necropsy.
- no test substance-related effects on any reproductive outcome endpoint
- litter means for numbers of implantation sites, early and late resorptions were all comparable to control group values for every dose level tested.

Dose descriptor:

NOAEL

Effect level:

1 000 mg/kg bw/day (nominal)

Based on:

test mat.

Remarks on result:

not determinable due to absence of adverse toxic effects

Details on embryotoxic / teratogenic effects:

Embryotoxic / teratogenic effects:no effects

Details on embryotoxic / teratogenic effects:
- litter means for live and dead foetuses as well as for foetal weight and sex ratio were all comparable to control group values for every dose level tested.
- no test substance-related foetal malformations or variations observed at any dose level tested. The foetal alterations occurred with low frequency across all groups tested and occur with similar frequency in the test facility historical control database.

Dose descriptor:

NOAEL

Effect level:

1 000 mg/kg bw/day (nominal)

Based on:

test mat.

Sex:

male/female

Remarks on result:

not determinable due to absence of adverse toxic effects

Abnormalities:

not specified

Developmental effects observed:

not specified

Conclusions:

Under the conditions of this study, there was no evidence of either maternal or developmental toxicity at doses up to 1000 mg/kg/day. Therefore, the no-observed-effect level (NOAEL) for maternal and developmental toxicity was considered 1000 mg/kg/day.

Effect on developmental toxicity: via oral route

Endpoint conclusion:

no adverse effect observed

Dose descriptor:

NOAEL

1 000 mg/kg bw/day

Study duration:

subchronic

Species:

rat

Quality of whole database:

6 guideline- and GLP-conform studies available, the overall quality of the database is therefore high.

Effect on developmental toxicity: via inhalation route

Endpoint conclusion:

no study available

Effect on developmental toxicity: via dermal route

Endpoint conclusion:

no study available

Additional information

Developmental toxicity

Human data

Note: Human data on developmental toxicity are included in IUCLID dossier under the endpoint Section 7.10.3 Direct observations: clinical cases, poisoning incidents and other.

During a comprehensive literature search, human data on developmental toxicity could be identified. After a thorough reliability screening, this reference was considered of limited relevance for hazard assessment purposes. The criteria for quality, reliability and adequacy of experimental data under REACH for hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X) are not fulfilled. For information purposes, the reference is discussed briefly below, highlighting their findings and the reasons for their exclusion from health hazard assessment:

 

Guillard et al. (2020) aimed to assess titanium dioxide exposure in the human foetus by performing ICP-MS analysis of titanium contents in the human placenta collected at term from pregnancies and in the meconium. These matrices were also analysed with scanning transmission electron microscopy coupled to energy dispersive X-ray (STEM-EDX) analysis to ensure the presence of particles and their chemical nature. Second, by using the ex vivo human placenta perfusion model, it was assessed whether TiO2-NPs of dietary origin (E171 additive) may cross the placental barrier by combining ICP-MS, confocal microscopy, scanning electron microscopy (SEM) and STEM-EDX for compositional analysis and elemental mapping of the perfused samples in the foetal side and particle distribution into perfused placental tissues.

In the current study, the authors analysed human placentae collected at term from normal pregnancies and meconium from unpaired mothers/children using inductively coupled plasma mass spectrometry and scanning transmission electron microscopy coupled to energy-dispersive X-ray (EDX) spectroscopy for their titanium contents and for analysis of titanium dioxide particle deposition, respectively. Using an ex vivo placenta perfusion model, the authors also assessed the transplacental passage of food-grade TiO2 particles.

Using the ICP-MS analysis, the authors evidenced the presence of titanium in all placentae (basal level ranging from 0.01 to 0.48 mg/kg of tissue; n =22). Seven women displayed a total placental titanium content above 0.1 mg/kg of tissue, with 2 of them reaching 0.4–0.5 mg/kg of tissue. Furthermore, the ICP-MS analysis showed that in 50% of the meconium samples (0.02–1.50 mg/kg; n = 18) contained titanium. Also, STEM-EDX observation of the placental tissues confirmed the presence of titanium dioxide nanoparticles in addition to iron, tin, aluminium and silicon as mixed or isolated particle deposits. Titanium dioxide particles, as well as silicon, aluminium, iron and zinc particles were also recovered in the meconium.

In a placenta perfusion experiments (n = 7 placentae), confocal imaging and SEM-EDX analysis of foetal exudate showed a low transfer of food-grade titanium dioxide particles to the foetal side, which was barely quantifiable by ICP-MS. After the perfusion of placentae with food-grade titanium dioxide (E171), titanium was detected by ICP-MS in most foetal exudates (0.41 to 3.46 ng Ti/mL), but 92% were in the range of the blank level (0.33 to 1.92 ng Ti/mL). To assess whether a transfer occurred for titanium dioxide particles, the authors examined the exudate samples collected every 5 minutes on the foetal side by confocal microscopy to detect laser-diffracting (titanium dioxide-like) particles. The placental translocation of laser-diffracting titanium dioxide particles increased from 10 minutes of E171 perfusion, reached a plateau at 20–30 minutes, and then decreased until the end of the perfusion. Further, diameter measurements showed that 70 to 100% of the titanium dioxide particles recovered in the foetal exudate were nanosized.

Lastly, for placental tissue, which was used to investigate the distribution of titanium dioxide particles into the perfused cotyledon using (S)TEM-EDX, the results showed that enriched tissue areas with titanium + oxygen contained isolated round shaped particles or small aggregates of titanium dioxide. Titanium dioxide particles from the E171 suspension were recovered in the syncytiotrophoblast microvilli and had translocated in deeper areas of the placental chorionic mesenchyme surrounding foetal vessels. Size measurement of the 33 titanium dioxide (i.e., EDX-characterized) particles translocated into placental tissues showed 26 particles with a diameter below 250 nm, and 17 of them in the nanorange.

This reference had several reporting and experimental deficiencies:

Firstly, the test item data on particle size characterisation is poorly described and purity data is missing. In addition, the source of the test item is not clearly stated. It is only stated that the test substance was purchased from a website of a French commercial supplier of food colouring agents, but the name of the website was not stated in order to obtain any information.

 

The reference by Guillard et al. 2020 was reviewed by two expert toxicologists in the field of reproductive toxicity (John M. DeSesso and Amy Lavin Williams). The following critique was provided:

The experimental design of the study is insufficiently documented and shows some deficiencies. The authors stated that they used placentae obtained after birth and use these for testing. However, placental transfer of materials at term is not always predictive of transfer processes earlier in gestation (Carney et al., 2005; DeSesso et al., 2012) as a young placenta has a different absorptive capacity compared to older placentae and as the placenta differentiates and matures, discontinuities in the syncytiotrophoblast appear and are filled in by fibrin plaques (Burton and Fowden, 2015); these areas may allow access of materials to the foetal circulation.

In addition, the placentae used in the study were obtained either by vaginal delivery or caesarean delivery. The type of delivery is likely to have an impact on the placental perfusion properties and rates of perfusion. Firstly, placentae acquired from caesarean delivery can be damaged during removal (Mathiesen et al., 2010). This can cause the perfusion not to be successful (Mathiesen et al., 2010). Then, with vaginal delivery, the time spent awaiting the transfer of blood from the placental into the infant after delivery may contribute to some portions of the placenta becoming ischemic through the collapse of vessels and the 20-30% change in the thickness of villainous membranes (Kaufmann and Burton, 1994). On the other hand, with cesarean delivery, the placental cord is clamped, there are not as many problems with collapsed vessels, but the placentae are still subject to ischemia due to the lack of circulation of oxygenated blood (Kaufmann and Burton, 1994).

Further, the authors do not give a good description of the perfusion method. Firstly, the authors do not state, if they oxygenated the maternal perfusate, as described by Schneider et al. (1972). Impairment of placental barrier function and placental tissue damage result, if perfusion is performed under hypoxic conditions (Bachmaier et al., 2007). Next, the authors stated that the placentae were prepared for perfusion within 1 hour of delivery. However, experts recommend more rapid perfusion preparation. Methods developed for placental perfusion (e.g., Schneider et al., 1972; Mathiesen et al., 2010) specify that placentae should be perfused immediately after delivery to maintain placental integrity.

The perfusion of placenta is technically a challenging procedure. Quality checkpoints have been established for determining placental perfusion success as described by Mathiesen et al., 2010:

-            cannulation and decision to proceed by isolating the cotyledon

-            pre-perfusion with oxygen transfer and no leak of foetal perfusion media

-            successful perfusion by the success criteria stated; antipyrine foetal – maternal ratio above 0.75 and volume loss from foetal reservoir no more than 3 mL/hr

Based on the above success criteria, Mathiesen et al. (2010) determined a perfusion success of only 15% over a 1-year period (n=202 placentae perfused). However, the authors of the current study reported a placental perfusion rate of 68% (15 of 22 placenta were perfused) and, in addition,they used inadequate criteria to confirm perfusion success:

-            no measurement of O2 transfer is noted

-            no assessment of foetal perfusion media leak was conducted

-            antipyrine as a control substance that passively diffuses across the placenta was included in the maternal reservoir, but the maternal-to-foetal transfer value of 20% used in this study as study inclusion/ exclusion criteria is well below the foetal-maternal ratio of 0.75 recommended by Mathiesen et al. (2010)

 

The authors used a concentration of 15 µg/mL of titanium dioxide for maternal perfusate. However, the mean titanium concentrations reported in the literature for human whole blood have ranged from 2.87 to 70.7 μg/L (Koller et al., 2018). Therefore, the concentration used in the current study is not representative of human exposure.

 

Regarding the measurements of total titanium content in human term placenta and meconium by ICP-MS, several experimental deficiencies can be determined. Firstly, a very limited data set (22 placentae; 18 meconium) was used for investigation, thus, it is unclear how these data extrapolate to the general population. Furthermore, it appears that the reported results are based on the analysis of single samples from each placenta or meconium collected rather than mean values from multiple replicates per sample (few exceptions exist), therefore, the within tissue variability of these results is unknown. In addition, results from an interlaboratory validation effort indicate that reliable quantification of titanium by ICP-MS requires tissue levels of >4 mg/kg (> 4 µg/g tissue) (Krystek et al., 2014); all of the placental and meconium samples analysed in the current study, however, provided tissue concentrations below 4 mg/kg and thus are not reliable.

 

In the current study, the samples of the placentae were dry-ashed prior to analysis. For this reason, it is not known whether the titanium detected in these samples was located on the maternal or foetal side of the placenta. Thus, it cannot be discerned, if the presence of these particles represents transfer of titanium across the placenta to the offspring. In addition, the average value of 0.1 mg/kg in the placental samples for titanium content was influenced by the results of two samples with concentrations of 0.47 and 0.48 mg/kg. Moreover, the sample size of the placental tissues used for analysis were not stated.

 

Lastly, experimental deficiencies were detected for the analysis of meconium in the current study. Although samples were collected within 48 hours of parturition, it is not reported whether the infants were breastfed or fed infant formula during that time or whether diaper rash creams were applied. Specific information regarding the potential use of TiO2 in infant formula could not be located; however, E171 is used in dehydrated milk products, dietary foods for special medical purposes, and food supplements supplied in liquid form (EFSA, 2016); therefore, its potential presence in infant formula cannot be excluded, which could account for titanium measured in meconium samples. In addition, titanium has been reported to be present in some diaper rash creams and whether it may be present in the absorbent material in diapers is unknown; however, the potential for meconium contamination via these sources cannot be discounted. Also, since titanium is present in many different materials, including polypropylene (Krystek et al., 2014), its introduction via the diaper sampling process cannot be discounted. Lastly, half of the meconium samples had levels below the limits of quantitation and values provided for 5 samples were based on weighted means from 3 analyses, while those for the other 4 quantifiable samples appear to be single measures; the reason for the discrepancy between single versus mean values is not explained

 

In conclusion, maternal-foetal transfer of food-grade titanium dioxide across the placenta cannot be concluded based on the results of the current study based on the reasons presented above. The conditions of perfusion experiments were not representative of those associated with typical human exposures. Then, the integrity of the placental barrier in the perfusion experiments was not demonstrated and is unlikely. In addition, the titanium present in the unperfused placental samples cannot be identified as having transferred to the foetal circulation. Lastly, multiple alternative sources other than placental transfer may have contributed to the titanium measured in the meconium samples.

 

Due to the shortcomings presented above, this publication was disregarded for the hazard and risk assessment.

 

References:

Bachmaier N, Linnemann K, May K, Warzok R, Kuno S, Niemeyer M, Balk S, Fusch C. 2007. Ultrastructure of human placental tissue after 6 h of normoxic and hypoxic dual in vitro placental perfusion. Placenta 28:861-7.

Burton GJ, Fowden AL. 2015. The placenta: A multifaceted, transient organ. Phil. Trans. R. Soc. B 370:20140066. http://dx.doi.org/10.1098/rstb.2014.0066

Carney EW, Scialli AR, Watson RE, DeSesso JM. 2005. Mechanisms regulating toxicant disposition to the embryo during early pregnancy: An interspecies comparison. Birth Defects Research 72:345-360.

DeSesso JM, Williams AW, Ahuja A, Bowman CJ, Hurtt ME. 2012. The placenta, transfer of immunoglobulins, and safety assessment of biopharmaceuticals in pregnancy. Critical Reviews in Toxicology 42:185-210.

European Food Safety Authority (EFSA). 2016. Re-evaluation of titanium dioxide (E 171) as a food additive. EFSA Journal 14:4545.

Koller D, Bramhall P, Devoy J, Goenaga-Infante H, Harrington CF, Leese E, Morton J, Nuñez S, Rogers J, Sampson B, Powell JJ. 2018. Analysis of soluble or titanium dioxide derived titanium levels in human whole blood: consensus from an inter-laboratory comparison. Analyst 143:5520 - 5529.

Krystek P, Tentschert J, Nia Y, Trouiller B, Noël L, Goetz ME, Papin A, Luch A, Guérin T, de Jong WH. 2014. Method development and inter-laboratory comparison about the determination of titanium from titanium dioxide nanoparticles in tissues by inductively coupled plasma mass spectrometry. Anal. Bioanal. Chem. 406:3853-3861.

Kaufmann P, Burton G. 1994. Anatomy and genesis of the placenta, Chapter 8 in The Physiology of Reproduction, 2nd Edition, Knobil E, O’Neill JD, Eds, Raven Press, New York.

Mathiesen L, Mose T, Mørck TJ, Kolstrup J, Nielsen S, Nielsen LK, Maroun LL, Dziegiel MH, Larsen LG, Kndsen LE. 2010. Quality assessment of a placental perfusion protocol. Reproductive Toxicology 30:138-146.

Schneider H, Panigel M, Dancis J. 1972.Transfer across the perfused human placenta of antipyrine, sodium, and leucine. Am. J. Obstet. Gynecol. 114:822-828.

 

 

Animal toxicity data

Six GLP and guideline-compliant pre-natal developmental toxicities studies in rats are available, with 3 pigment grade and 3 ultra-fine (nano form) titanium dioxide forms, administered via gavage (according to OECD 414 (2001), GLP). The reliable guideline studies are summarized below.

Holalagoudar (2014) evaluated ultrafine and pigment-grade titanium dioxide in two separate developmental toxicity studies according to OECD 414 (2001) and under GLP. Groups of 24 -25 mated female Wistar rats were given titanium dioxide or titanium dioxide nanoparticles in aqua ad iniectabilia (water for injection) at dose levels of 100, 300, and 1000 mg/kg bw/day via gavage once daily during gestation day 5 through gestation day 19. A vehicle control group was run concurrently. Animals were sacrificed on day 20 of gestation and were investigated for general toxicity including examination of ovaries and uterine content. Pups were evaluated for viability (number of alive and dead foetuses), sex distribution and body weights. Foetuses were inspected externally and subjected to skeletal and soft tissue examinations and evaluated for malformations, variations and retardations. There were no adverse test substance-related effects on any maternal or foetal data collected. Therefore, the no-observed-effect level (NOAEL) for maternal and developmental toxicity was considered to be 1,000 mg/kg/day.

 

Munley (2014) evaluated ultrafine and pigment-grade titanium dioxide in three separate developmental toxicity studies according to OECD 414 (2001) and under GLP. Groups of 22 time-mated female Crl:CD(SD) rats were treated with titanium dioxide or titanium dioxide nanoparticles in sterile water for injection at dose levels of 100, 300, and 1000 mg/kg bw/day via gavage once daily during gestation days 6 – 20. A vehicle control group was run concurrently. Animals were sacrificed on day 21 of gestation and were investigated for general toxicity including examinations of ovaries and uterine content. Alive foetuses were inspected externally and subjected to skeletal and soft tissue examinations. Furthermore, alive foetuses were evaluated for sex distribution and body weight. There were no adverse test substance-related effects on any maternal or foetal data collected. Therefore, the no-observed-effect level (NOAEL) for maternal and developmental toxicity was considered to be 1,000 mg/kg/day.

 

Takawale (2014) evaluated pigment-grade titanium dioxide in a developmental toxicity study according to OECD 414 (2001) and under GLP. Groups of 24 -25 mated female Wistar rats were treated with the test item in aqua ad iniectabilia (water for injection) at dose levels of 100, 300, and 1000 mg/kg bw/day via gavage once daily during gestation days 5 through 19. A vehicle control group was run concurrently. Animals were sacrificed on gestation day 20 and were investigated for general toxicity including examinations of ovaries and uterine content. Pups were evaluated for viability (number of alive and dead foetuses) sex distribution and body weight. Furthermore, they were inspected externally and subjected to skeletal and soft tissue examinations. There were no adverse test substance-related effects on any maternal or foetal data collected. Therefore, the no-observed-effect level (NOAEL) for maternal and developmental toxicity was considered to be 1,000 mg/kg/day.

 

Furthermore, during a comprehensive literature search for data on developmental toxicity of TiO2, seven references were identified representing studies with information on developmental toxicity in rats or mice receiving either ultrafine or pigment-grade titanium dioxide via oral, inhalation, intraperitoneal or intravenous administration. After a thorough reliability screening, these references were considered of limited relevance for hazard assessment purposes. The criteria for quality, reliability and adequacy of experimental data under REACH for hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X) are not fulfilled. These references are discussed briefly below, highlighting their findings and the reasons for their exclusion from health hazard assessment:

NIER (2009) describes a study which evaluated the reproductive and developmental toxicity of titanium dioxide according to OECD 421. A group of 10 male and 10 female Sprague-Dawley rats was treated with 1000 mg TiO2/kg bw/day in 1 % methylcellulose solution via gavage. According to the authors, no treatment-related changes were observed in the parental generation regarding general toxicity and reproductive performance. Also, the authors stated that no treatment-related changes were observed in the F1 generation. The NOAEL of titanium dioxide was considered to be above 1,000 mg/kg bw/day. The reference is only available as study summary record without raw data and was therefore rated as not assignable (RL=4). The original study record was not obtainable, however the study was assessed and peer-reviewed within the OECD SIDS initial assessment report (OECD CoCAM4, April 2013) and approved by the OECD procedure on Mutual Acceptance of Data (MAD). Thus, the study is considered useful as supporting information.

Stapleton et al. (2013) investigated the effect of ultrafine titanium dioxide on the maternal and foetal microvascular function. Starting on gestation day 10, 10 pregnant Sprague Dawley rats were exposed to the test item at 11.3 ±0.039 mg/m³ via inhalation for 5 hours/day. The average duration of exposure was 8.2 ±0.85 days (varying between 5 - 13 days). A vehicle control group was run concurrently (n = 6). At gestation day 20, isolation of the microvascular tissue and arteriolar reactivity of the uterine premyometrial as well as of foetal tail arteries were carried out. According to the authors, test item exposures led to significant maternal and foetal microvascular dysfunction which presented as robustly compromised endothelium-dependent and -independent reactivity to pharmacological and mechanical stimuli. Isolated maternal uterine arteriolar reactivity was consistent with a metabolically impaired profile and hostile gestational environment, impacting foetal weight. The foetal microvessels isolated from exposed dams were suggested to demonstrate significant impairments to signals of vasodilation specific to mechanistic signalling and shear stress. However, this reference has several reporting and experimental deficiencies which do not reliable conclusions on developmental toxicity:

-      the number of pregnant rats was too low in this study (treatment: n = 10; control group: n = 6): in comparison, the guideline foresees a total of approx. 20 pregnant females for each dose level and control group. This reduction of pregnant dams represents a significant reduction of statistical power and relevance of findings.

-      the exposure regime was not the same for all exposed animals. The duration of exposure varied between 5 – 13 days. This regime makes it impossible to compare the data between the various females.

-      the guideline foresees that at least three dose level and a concurrent control should be used. In this study, only one concentration level of 11.3 ±0.039 mg/m³ was tested. The usage of one concentration level precludes the possibility to demonstrate any dose-related response and the determination of a No-Observed-Adverse Effect level (NOAEL).

-      the age of dams was not stated. Furthermore, data on clinical observations, body weight measurements (during dosing), food consumption, and gross pathology of dams are missing. Lastly, uterine weight, no. of corpora lutea, no. and % of live and dead foetuses, no. and % of pre- and post implantation losses and resorptions are not reported. Without this kind of information, it is unknown whether the test item had any adverse effect on the pregnant animals, which in turn could have an adverse outcome on the developing offspring.

-      the sex of the offspring was not reported. This precludes the possibility to determine whether the test item had any sex-specific effect.

-      the skeletal examination and soft tissue examination (exception microvascular level) of the foetuses are missing. These examinations are foreseen by the guideline.

Based on the shortcomings given above, this publication was disregarded for the hazard and risk assessment.

 

Stapleton et al. (2015) investigated the effect of ultrafine titanium dioxide on the uterine and coronary microvascular function of female offspring after gestational exposure. Pregnant Sprague-Dawley rats (n = 6) were exposed to test item aerosols (10.6 mg/m³) from gestational day 6 to gestation day 20. A control group (n = 7) was run concurrently, which received filtered air. Exposure duration was 5 hours/day, 4 days/week. Pups were delivered, and the progeny grew into adulthood. Microvascular reactivity, mitochondrial respiration and hydrogen peroxide production of the coronary and uterine circulations of the female offspring were evaluated. According to the authors, the study demonstrated microvascular dysfunction coinciding with mitochondrial inefficiencies in both the cardiac and uterine tissues of adult offspring exposed to the test item in utero, which may represent initial evidence that prenatal test item exposure produces microvascular impairments that persist throughout multiple developmental stages. This reference has several reporting and experimental deficiencies, as follows:

-      the number of pregnant rats was too low in this study (n = 6). The guideline foresees a total of approx. 20 pregnant females for each dose level and control group. This reduction of pregnant dams causes a significant reduction of the statistical power.

-      exposure was not daily from implantation to the day prior to caesarean section (4 days/week from GD6 to GD 20). The guideline foresees that normally the pregnant animals should be exposed to the test item from implantation to caesarean section.

-      the guideline foresees that at least three dose level and a control should be used. In this study, only one concentration level (10.6 mg/m³) was tested. The use of one concentration level precludes the possibility to demonstrate any dose-related response and the determination of a No-Observed-Adverse Effect level (NOAEL).

-      clinical observations, body weight, food consumption and gross pathology of dams were not examined/recorded. Furthermore, dams were sacrificed after weaning but examination of uterine content was not conducted (except: implantation). Without this kind of information, it is unknown if the test item had any adverse effect on the pregnant animals, which in turn could have an adverse outcome on the developing offspring.

-      Only female progeny were examined, which does not allow a conclusion for effect in males.

-      The external, skeletal, and soft tissue examinations of the offspring were missing. These examinations are foreseen by the guideline.

Based on the shortcomings given above, this publication was disregarded for health hazard assessment.

 

Hathaway et al. (2017) investigated the effect of ultrafine titanium dioxide on the cardiomyocytes of foetuses. A group of 13 pregnant Sprague-Dawley rats were exposed to titanium dioxide (P25) nano-aerosols (approx. 10 mg/m³) for 7 – 8 non-consecutive days, beginning at gestational day 5 – 6. A control group was run concurrently. The effects on cardiomyocytes of foetuses were evaluated on gestational day 20. According to the authors, foetal cardiomyocytes showed a decrease in basal respiration and an increase in proton leak in the experimental group following inhalative administration of the test item to pregnant rats. Furthermore, they stated that for the proton leak, qPCR of UCP2 revealed increased mRNA levels in the experimental group compared with the control, with qPCR of fatty acid metabolism constituents PPAR-γ and CPT1A (P = 0.023) exhibiting a similar increase. This reference has the following several reporting and experimental deficiencies, precluding it from being useful for hazard assessment purposes, as follows:

-      test item information is incomplete: the purity and stability of the test item are missing.

-      the number of pregnant rats was too low in this study (n = 13). The guideline foresees a total of approx. 20 pregnant females for each dose level and control group. This reduction of pregnant dams causes a significant reduction of the statistical power.

-      test item administration was not daily (7 - 8 non-consecutive days between gestation day 5 or 6 and gestational day 20) as foreseen by the guideline.

-      the guideline foresees that at least three dose level and a control should be used. In this study, only one concentration level (approx. 10 mg/m³) was tested. The use of one concentration level precludes the possibility to demonstrate any dose-related response and the determination of a No-Observed-Adverse Effect level (NOAEL).

-      the age of the parental animals was missing. Furthermore, examinations of the maternal parent are missing (clinical signs, mortality, food consumption, gross pathology missing, corpora lutea count, no. of implantations, no. and % of live and dead foetuses, resorptions, and no. and % of pre- and post-implantation losses missing). Without this kind of information, it is unknown whether the test item had any adverse effect on the pregnant animals, which in turn could have an adverse outcome on the developing offspring.

-      the examinations of the foetuses are missing (sex ratio, % of live offspring, body weight, external examinations, skeletal examinations, soft tissue examinations missing). These examinations are foreseen by the guideline.

-      The vehicle is not described, which precludes the possibility to determine if an appropriate vehicle was used.

Based on the shortcomings given above, this publication was disregarded for the hazard and risk assessment.

 

Elbastawisy and Almasry (2014) performed a study to determine the potential for histomorphological alterations occurring in maternal and neonatal pulmonary distal airspaces of Wistar rats after maternal administration of ultrafine titanium dioxide. A group of 15 pregnant rats were treated with 5000 mg/kg bw/day of the test item suspended in phosphate-buffered saline. The test item was given orally by gavage once daily from days 6 to 12 of gestation. A control group of 15 pregnant rats receiving the vehicle alone was run concurrently. On gestation day 21, the dams were sacrificed. Maternal and neonatal lungs were collected and processed for energy-dispersive X-ray (EDX) and histological analysis. The mean linear intercept and airspace wall thickness were measured by a stereological procedure with image analysis to assess alveolarisation. According to the authors, titanium dioxide was detected in maternal and neonatal lungs after oral intake by pregnant rats. They stated that the pulmonary response manifested as inflammatory lesions and delayed saccular development in neonates. This reference has the following severe reporting and experimental deficiencies:

- test item was insufficiently characterised, which makes it impossible to verify the identity of the test item.

- furthermore, the results presented in this publication are implausible. No publication has thus far shown systemic availability of titanium dioxide (whether pigment-grade or nano-sized) via the oral route. Furthermore, there is the potential that the titanium dioxide particles translocated via aspiration or mis-gavage to the lungs.

- the number of pregnant female per group is too low (n = 15). The guideline foresees a total of approx. 20 pregnant females for each dose level and control group. This reduction of pregnant dams causes a significant reduction of the statistical power.

- dosing was not conducted up to sacrifice of the dams.

- maternal and foetal examination were incomplete, which makes it difficult to determine whether the effects observed were truly caused by the test item.

Based on the shortcomings given above, this publication was disregarded for the hazard and risk assessment.

 

Fournier (2019) exposed pregnant Sprague-Dawley rats to filtered air or nano-TiO2 aerosols or not exposed (naive) on GD 4,12, or 17 in a whole body chamber. The animals were euthanised on GD 20 and the following parameters were analysed: litter size (number of pups), pup weight, placenta weight and number of “reabsorption sites". Further, the mean arterial pressure (MAP) of the right maternal carotid artery (by a BLPR2 pressure transducer with blood pressure monitor) as well as the vascular reactivity of the maternal thoracic aorta, the uterine artery, the umbilical vein, and the foetal thoracic aorta were evaluated using wire myography. According to the authors impairments were noted at each level of the materno-foetal vascular tree and at each exposure day. The greatest effects were observed within the foetal vasculature (umbilical vein and foetal aorta), wherein effects of a single maternal inhalational exposure to nano-TiO2 on GD 4 modified responses to cholinergic, NO, and a-adrenergic signaling. However, due to the following major study restrictions the study is not relevant for human hazard assessment and therefore disregarded:

-                     

- it is explicitly stated that groups of 17 pregnant rats were placed in a 84-L whole-body exposure chamber for 5 h during a single inhalation exposure and removed when the concentration fell below 1 mg/m3, thereby claiming a 4 h exposure at the target concentration. The aerosol size distribution (SMPS) was reported with 133.73 ± 1.87 nm (not further qualified), and the achieved mass concentration is stated with 9.71 ± 0.22 mg/m3 , from which the authors calculated a pulmonary deposition of 15.8 ± 1.2 μg (no further details). The applied methodology to calculate such deposition is not documented in the publication and therefore casts doubt on the correctness of the reported values.

- OECD TG 414 conventionally requires 20 pregnant animals per dose group; here, group sizes of controls (air) varied between three single exposure on GD 4, 12 and 17 with 5, 6 and 4 animals/group, and for the TiO2-exposed animals on GD 4, 12 and 17 with 6, 8 and 7 animals/group, respectively. The somewhat random group sizes and the discrepancy between the statement that 17 rats were exposed together whole body at one occasion shed doubt on the standardisation of test conditions and suggest that obviously not all animals per group were exposed at the same time or under similar conditions. Also, the variation in group sizes is not explained and casts doubt as to the health status of the treated animals.

- OECD 414 also foresees daily exposure up to day before section, and a minimum of three dose levels plus control to allow for the assessment of dose-response.

- The publication does not provide any information on animals age, acclimation period, environmental conditions or diet and water supply.

- Body weight and food consumption was not recorded, only the final maternal body weight was stated.

- Thyroid hormone levels in maternal animals were not determined.

- The publication does not provide any method description whatsoever for the macroscopic evaluation, except that animals were euthanised on GD20. Only the following parameters were reported: litter size (number of pups), pup weight, placenta weight and number of “reabsorption sites. The authors use the term „reabsorption sites“ but without explaining what this means and how this was assessed.

- No information on skeletal, soft tissue and external alterations are given.

- In contrast, OECD 414 requires weighing of (i) gravid uteri including the cervix, (ii) examination of the uterine contents for numbers of embryonic or foetal deaths and viable foetuses, and (iii) the degree of resorptions and whether they were early or late resorptions. In this study, the authors measured placental weight instead. The lack of using standardised guidance-compliant reporting parameters makes it difficult to compare any alterations to historical background data.

- When the authors use the term „pups“, one must assume that they refer to foetuses, and they report these foetus weights as ranging from approx. 2.5-2.7 grams. When comparing these to historical control data (CRL, see Appendix; the lab themselves do not report any historical control data of their own), the corresponding values for GD20 5.44-6.31 for males, 5.16-6.05 for females (CRL).

 

The methodology used in the statistical analysis is reported in a confusing way:

- for Table 1 it is stated in the table note that a one-way ANOVA was conducted. An indication on the post-hoc comparison of means conducted is however not given. One may assume that the Tukey-Kramer test was conducted that would allow a comparison even in the aforementioned unbalanced study design. This is however not explicitly stated. In contrast, the section on “Statistics” mentions a two-factorial ANOVA for repeated measures with subsequent Tukey-Kramer comparison of means for “point-to-point” differences. From the given explanations on study design and results it remains unclear: 1) which factors has been considered (only 1 factor (i.e. agonist or day of treatment) is obvious from the curve- and column figures in Figure 3, 4, 5 and 6. 2) which “repeated measures” (standard statistical terminology would consider “repeated measures” as subsequent measures of the same end point in the same study object, when the influence of time is to be considered) were considered.

- No information is given how many foetuses were examined, sex of the foetuses, based on which criteria the foetuses were selected and how the vessels of the foetuses were explanted. The isolation and trimming of vessels is described very briefly, especially the consistent use of unbranched vessels is important to obtain comparable results. The concentrations used to induce contraction (phenylephrine, PHE) and relaxation (methacholine, MCH or sodium nitroprusside, SNP) appear somewhat high (up to 10-4 M) compared to concentrations used by other researchers (usually up to 10-6 or 10-5 M).

- Pup weight appears to be significantly altered in comparison to the (air) control, but actually only on GD12. However, the pup weights in the unexposed controls are given 2.56 (GD4), 2.48 (GD12) and 2.65 (GD17) and therefore do not follow a plausible increasing trend. Likewise, the pup weights of the nano-TiO2-exposed animals were 2.50 (GD4), 2.65 (GD12), 2.57 (GD17). Taken altogether, the pup weights vary (min-max) in the air controls by 6.8% between the three observation time points, and by 6.0% (min-max) for the nano-TiO2 exposed animals 5. When comparing these above data to historical control data for this rat strain (SD), it is interesting to note that the same provider of these animals (Charles River) states average foetal weights on GD20 caesarean section in the range of 3.84-4.03 grams (m) and of 3.67-3.81 grams (f); the historical control thus differ between sexes, but Fournier et al. do not report the sex of the foetuses. Most striking however, is the observation that the foetal weights in the Fournier et al study are approx. 40% lower than historical controls for this strain, which renders the findings by the authors as somewhat implausible.

- Placenta weight: the determination of placenta weight is not a standard parameter in developmental toxicity testing, not being required according to OECD TG 414. Instead, the guideline states “gravid uteri including the cervix should be weighed“. Instead of weighing the intact uterus after caesarean section, the investigators chose to undertake the rather complicated procedure of extracting the foetuses first. After removing approx. 12 foetuses form each uterus, one may expect a certain variation in the remaining placenta. weight. However, these were weighed at approx. 0.4-0.5 g, and the variance of weighing is stated with exactly 0.01g for every group, which suggests a level of precision in their sectioning which is rather unrealistic.

- Number of resorptions: the authors speak about “number of reabsorption sites”, which are not further qualified. This terminology is however nowhere defined in developmental toxicity testing. Instead, OECD TG 414 states that “the uterine contents should be examined for numbers of embryonic or foetal deaths and viable foetuses. The degree of resorption should be described in order to estimate the relative time of death of the conceptus”. Assuming that the authors mean “resorption” by “reabsorption”, they nevertheless disregard the aspect of early and late resorptions.

- According to historical control data from CRL for this strain (SD) on GD20 caesarean section, the total of resorptions is on average 0.6% (range: 0.4-1.0%). Fournier et al. state a value of 0.58% for ”naïve” rats, but then give values of 0.0-0.2% for their air controls, which is unrealistic. By then comparing their results for nano-TiO2 exposed animals to these controls, their number of resorptions attain statistical significance. It is however improper to evaluate any developmental findings without consideration of the ranges of historical control data. Assuming that the authors mean “total resorptions/litter” by “number of reabsorption sites”, the values of the air and ENM exposed animals are well within the historical control (1.02 mean, 0.5-1.4 min-max).

In addition, the inconsistent and low number of pregnant animals (5-7 per group) instead of 20 per group as requires by OECD TG 414 renders the statistical significance questionable.

In conclusion, in this investigation, the authors subjected pregnant rats in a non-standard study design to single 4h exposures (single exposure level, ca. 10mg/m3) of nano-TiO2 (P25, Evonik) either on GD4, 12 or 17. They then report inconsistent (with respect to time points) and mild vascular reactivity and some selected findings which appear to suggest developmental effects, but which due to the use of only one singe exposure concentration lack dose-response relations. They finally suggest that the observed moderate vascular dysfunction associated with this single maternal exposure to nano-TiO2 aerosols may have a significant impact on foetal systemic vascular function.

The authors give absolutely no reasoning why, when they are interested in potential developmental effects, they deviate from standardised developmental toxicity practices. Further, (i) their “random” assignment of numbers of animals per exposure group, (ii) the lack of statistical reliability (due to insufficient documentation), (iii) the questionable developmental findings which are almost entirely within the range of historical control data for this rat strain, and (iv) the marginal and inconsistent character of pharmacological vascular observations render this publication of limited relevance and reliability. It can also not be ruled out that the isolated observations may rather be more a reflection of the stress induced by the 4h inhalation exposure.

 

Lee et al. (2019) investigated the potential effects of oral exposure to titanium dioxide nanoparticles and their tissue distribution during pregnancy. Groups of 12 pregnant Sprague-Dawley rats were exposed to 100, 300, and 1000 mg/kg bw/day of titanium dioxide nanoparticles via gavage daily from gestation day 6 to gestation day 19. A control group was run concurrently. No mortality was observed in the dams of any dose group. Furthermore, no treatment-related effects were observed in dams for clinical signs, body weight, food consumption, organ weights, number of implantations, pre- implantation loss, post-implantation loss, number of corpora lutea, number of resorptions (early and late), number of pregnant females, and number of dead or alive foetuses.

 

In addition, no treatment-related effects were observed in foetuses of any dose group for body weight, sex ratio, external examinations, visceral examinations, skeletal examinations, placental weight and placental macroscopic observations. Lastly, titanium concentrations in maternal liver, maternal brain and placenta at 1000 mg/kg bw/day were elevated compared to the concentration in control animals (statistically significant for maternal brain only). In addition, at 300 mg/kg bw/day, titanium concentrations in the maternal brain and placenta were also slightly elevated (not statistically significant). Moreover, there was no titanium concentration change in the maternal blood, foetal liver, foetal brain or foetal blood. Based on these results, the NOAEL for maternal toxicity and developmental toxicity can be determined to be 1000 mg/kg bw/day. However, this reference had reporting and experimental deficiencies, which do not allow an independent review about the exposure-effect correlation:

- the number of pregnant animals per group (n=12) is not sufficient to perform an appropriate statistical analysis.

- the gravid uteri was weighted without cervix, which is not guideline-conform.

- actual concentrations were not measured, which makes it impossible to confirm that the animals received the assumed dose level. 

- percentage of live/dead foetuses, resorptions, number of pre- and post-implantation losses were not reported, as request by the guideline

- individual data was no reported

 

 

 Summary entry – Developmental toxicity

Another 11 references were identified during literature search, representing in vivo mechanistic investigations in rats and mice, following oral, inhalation, intraperitoneal or intravenous administration of nano-sized titanium dioxide. These studies are primarily academic research papers, focussing on isolated organs, tissues or biomolecules, which respond to exposure with titanium dioxide. The study designs are not in accordance with accepted guidelines and are therefore of limited relevance for chemicals hazard assessment. Further, the references usually lack significance due to, e.g., poor test item characterisation, the low number of animals used, missing dose response relationship, using non-physiological route, contradictory results in comparison with highly reliable guideline studies. It is therefore concluded that all references do not fulfil the criteria for quality, reliability and adequacy of experimental data under REACH for hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X). The studies given below were therefore included in the IUCLID for information purposes only:

Ghareeb, A. et al. (2015): the experimental design is insufficiently documented, rat strain used in the study not further characterised, precluding a comparison with strain-specific historical control data, number of test animals/group is not stated, documentation of the results is imprecise (important raw data is missing, grading and localisation of findings were not specified), non-physiological route of administration via intraperitoneal injection (single dose) does not comply with relevant guidelines.

 Mohammadipour, A. et al. (2014): The investigated experimental parameters are not useful for the fulfilment of data requirements under REACH for hazard assessment purposes. The test material used in this study is not sufficiently characterised and the experimental design (e.g. only one dose administered) is not in accordance to any guideline. The repetitions were not sufficient to reach a statistically relevant level and moreover, the number of animals used in the tests were too low. Regarding the data analysis, it is unclear, whether only a positive variation is stated in the graphs.

Yamashita, K. et al. (2011): This publication is not relevant for human health risk assessment due to the irrelevant route of test item application. Furthermore, the publication is not relevant for hazard assessment since data on particle characterisation are poorly described. In addition, the number of animals is not clearly stated, one dose level only was tested, and exposure duration was too short. Lastly, the following parameters were not examined/recorded: clinical observations, mortality, food consumption, full examination of uterine content (resorption only), macroscopic examination of the dams, and no full foetal examination (body weight only).

Hong, F. et al. (2017): The publication is not relevant for human health risk assessment since the test item was self-synthesised and data on particle characterisation are poorly described. The number of animals per group is too low for an appropriate statistical analysis. Besides of these major restrictions, the following information is not included: mating cages, dosing volume, clinical observation, maternal mortality rate, food consumption, macroscopic examination of the dams, weight of gravid uteri including cervix, number of corpora lutea, sex ratio of foetuses, examination of the reproductive tract of the foetuses for signs of altered development.

Stapleton, P.A. et al. (2018): The study design is not compliant with any accepted/validated guideline and the investigations in toxicogenomics have no direct value for fulfilling data requirements under the REACH regulation. Moreover, essential information on exposure (e.g. MMAD and GSD) are missing. Maternal effects and toxicity were not investigated. Aeroxide P25, which although a commercial substance, it represents a unique specialty material that is manufactured at relatively low production volumes and is generally considered to be “different” from the more common forms of ultrafine and pigmentary TiO2. Method description is very brief and lacks essential information.

Zhou, Y. et al. (2019a): Since the self-synthesized test substance is poorly described and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in this experiment, and (ii) whether the effects can really be attributed to titanium dioxide and (iii) the methodical setup is not in accordance with any guideline. Maternal effects and toxicity were not investigated which raises the question whether the effects observed in offspring animals are due to the systemic toxicity observed in maternal animals. The number of animals per group (n=5) is too low for a statistical evaluation and parameters relevant for the identification of developmental toxicity were not analysed. Important information about the general study design, material and methods and results (individual results, etc.) are missing. Apart from that, this publication is part of the publication series of the University of Soochow. The publications of this working group were subject to an investigation of the University of Soochow, resulting in the retraction of several publication. The shortcomings identified were incorrect statistics, experimental errors and missing original data. Thus, due to these major study restrictions this study is disregarded.

Zhou, Y. et al. (2019b): Since the self-synthesized test substance is poorly described and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in this experiment, and (ii) whether the effects can really be attributed to titanium dioxide and (iii) the methodical setup is not in accordance with any guideline. Maternal effects and toxicity were not investigated which raises the question whether the effects observed in offspring animals are due to the systemic toxicity observed in maternal animals. The number of animals per group (n=5-6) is too low for a statistical evaluation and parameters relevant for the identification of developmental toxicity were not analysed. Important information about the general study design, material and methods and results (individual results, etc.) are missing. Apart from that, this publication is part of the publication series of the University of Soochow. The publications of this working group were subject to an investigation of the University of Soochow, resulting in the retraction of several publication. The shortcomings identified were incorrect statistics, experimental errors and missing original data. On top of that, it is also woth mentioning that this publication seems to report at least one part of the already published results by Zhou, Y. et al. (2019a)- please refer to the publication above). Thus, due to these major study restrictions this study is disregarded.

Abukabda, A.B. et al. (2019): This mechanistic study is not compliant with any accepted/validated guideline for testing developmental effects. Maternal effects and toxicity were not investigated which raises the question whether the effects observed in offspring animals are due to the systemic toxicity observed in maternal animals. The fact, that only one dose was administered does not allow a dose response related analysis. The number of animals per group (n=8) is too low for a statistical evaluation and parameters relevant for the identification of developmental toxicity were not analysed. Aeroxide P25, which although a commercial substance, it represents a unique specialty material that is manufactured at relatively low production volumes and is generally considered to be “different” from the more common forms of ultrafine and pigmentary TiO2. It is also noteworthy that the study design is not fully clear. The animals were sacrificed on GD 17 or 20. In the abstract it was mentioned that the animals were sacrificed on GD 20 whereas in the material and method part in was mentioned that the animals were exposed for 6 consecutive days and the observation was performed 24 hrs after the last exposure, which is GD 17. Thus, it remains unclear when the observations were performed.

Zhang, L. et al. (2018): This study is not compliant with any accepted/validated guideline for testing developmental effects. Maternal effects and toxicity were not investigated which raises the question whether the effects observed in offspring animals are due to the systemic toxicity observed in maternal animals. Since the test item is poorly described and also the influence of the other simultaneously administered process chemicals is not addressed, this raises doubts as to whether (i) nano-sized TiO2 particles at all were used in this experiment, and (ii) whether the effects can really be attributed to titanium dioxide. Further, the environmental conditions are uncommon, since animals had a 14 h light and 10 h dark cycle and the concentration of the dose formulations was not quantified. Only two dose levels were tested, which does not allow a dose response related analysis. Overall, in this mechanistic study parameters relevant for the identification of developmental toxicity were not analysed and thus this study is considered disregarded.

Bowdridge, E.C. et al. (2019): This mechanistic study is not compliant with any accepted/validated guideline for testing developmental effects. Maternal effects and toxicity were not investigated which raises the question whether the effects observed in offspring animals are due to the systemic toxicity observed in maternal animals. The fact, that only one dose was administered does not allow a dose response related analysis. The number of animals per group (n=6-8) is too low for a statistical evaluation and parameters relevant for the identification of developmental toxicity were not analysed. Aeroxide P25, which although a commercial substance, it represents a unique specialty material that is manufactured at relatively low production volumes and is generally considered to be “different” from the more common forms of ultrafine and pigmentary TiO2. It is also noteworthy that the study design is not fully clear. The animals were sacrificed on GD 17 or 20. In the abstract it was mentioned that the animals were sacrificed on GD 20 whereas in the material and method part in was mentioned that the animals were exposed for 6 consecutive days and the observation was performed 24 hrs after the last exposure, which is GD 17. Thus, it remains unclear when the observations were performed.

 

Philbrook, N. A. et al. (2011): This study is not compliant with any accepted/validated guideline for testing developmental effects. The test substance was not administered daily from implantation to the day prior to scheduled caesarean section. No measurement of food consumption was carried out. It is also noteworthy that the study design is not fully clear. The number of animals per group is not clearly stated (n = at least 11-13). No information about the measurement of the actual TiO2 concentrations in suspensions.

 

 

Developmental neurotoxicity

Note: animal data on developmental neurotoxicity can be found in IUCLID Section 7.8.1 Toxicity to reproduction.

Two references were identified during a comprehensive literature search, representing experiments with investigation on developmental neurotoxicity. These experiments were conducted with mice receiving nano-form titanium dioxide via intravenous and subcutaneous routes. The study designs are not in accordance with accepted guidelines and are therefore of limited relevance for chemicals hazard assessment. The references usually lack significance due to, e.g., administration via non-physiological route, missing dose-response-relationship, or low number of animals used. It is therefore concluded that all references do not fulfil the criteria for quality, reliability and adequacy of experimental data for the fulfilment of data requirements under REACH and hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X). The information contained therein were included for information purposes only.

Notter, Z. (2018): The non-physiological route of administration via intravenous injection and the single dose administration on gestation day 9 is not guideline conform and not suitable to assess reproduction toxicity and developmental neurotoxicity. Maternal effects and toxicity were not investigated which raises the question whether the effects observed in offspring animals are due to the systemic toxicity observed in maternal animals. Only two dose groups was tested which does not allow a dose-response related analysis. The number of animals per group (n=7-10) is too low for a statistical evaluation and parameters relevant for the identification of developmental neurotoxicity were not analysed. Further, there is no information on whether the statistical unit relates to the litter (or dam) or the pup.

Tarlan, M. et al. (2020): This study is not compliant with any accepted/validated guideline for testing developmental effects. The non-physiological route of administration via subcutaneous injection is not guideline conform and not suitable to assess developmental neurotoxicity. The test substance was not administered daily from implantation to the day prior to scheduled caesarean section. Only depressive-like behaviour was investigated. No clinical observations and measurement of body weight or food consumption of the dams. No gross pathological or neuropathological examinations were carried out. No histopathological examination of the brain from the offsprings. Maternal effects and toxicity were not investigated which raises the question whether the effects observed in offspring animals are due to the systemic toxicity observed in maternal animals. Furthermore, the study design is insufficiently described. The details about humidity and photoperiod in the environmental conditions are not given.

 

 

Conclusion

Developmental toxicity:

Human data:

One reference presenting data on developmental toxicity in humans is available. The study designs of this reference is not in accordance with any accepted guideline and are therefore of limited relevance for chemical hazard assessment. After a thorough reliability screening, this reference was considered of limited relevance for hazard assessment purposes. The criteria for quality, reliability and adequacy of experimental data under REACH for hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X) are not fulfilled. The study was included for information purposes only.

Animal data:

Based on the information from the available six pre-natal developmental toxicity studies via gavage in rats (according to OECD 414 and under GLP) it is concluded that maternal toxicity up to the highest dose level of 1,000 mg/kg bw/d and reproductive performance of dams were not affected by treatment. There was no embryo-toxicity and no effect on the development of foetuses. No external visceral or skeletal malformations were observed, and the incidence of variations was not different between treated and control groups. No incidences of skeletal retardations were observed.

 

One reproductive toxicity screening assay is available, but only as a brief study summary without raw data and was therefore rated as not assignable (RL=4). This study showed no developmental toxicity in the foetuses. Due to its use within the OECD HPV programme, the study is used as supporting information.

 

Further eighteen references report mechanistic studies with a focus on organ-specific investigations, such as uterine and coronary microvascular function, histomorphological alterations in maternal and neonatal pulmonary distal airspaces, effect of ultrafine titanium dioxide on the cardiomyocytes of foetuses, and tissue distribution during pregnancy. The study designs of these eighteen references are not in accordance with any accepted guideline and are therefore of limited relevance for chemical hazard assessment. The references lack significance due to poor test item characterisation, non-physiological route of administration, low number of animals available for testing, one concentration tested only, short exposure duration, wrong dosing regime, and/or incomplete observation data. It is therefore concluded that all references do not fulfil the criteria for quality, reliability and adequacy of experimental data for the fulfilment of data requirements under REACH and hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X). The studies were included for information purposes only.

 

 

Developmental neurotoxicity:

Two references were identified during a comprehensive literature search, representing experiments with investigation on developmental neurotoxicity.The study designs of these references are not in accordance with accepted guidelinesand are therefore of limited relevance for chemicals hazard assessment.The references usually lack significance due to, e.g., administration via non-physiological route, missing dose-response-relationship, or low number of animals used.It is therefore concluded that all references do not fulfil the criteria for quality, reliability and adequacy of experimental data for the fulfilment of data requirements under REACH and hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X). The information contained therein were included for information purposes only.

 

Toxicity to reproduction: other studies

Additional information

Toxicity to reproduction: other studies

 

During a comprehensive literature search for information on reproduction toxicity of titanium dioxide, two references were identified representing studies with special investigation on toxicity to reproduction, conducted in mice receiving pigment-grade or nano-form titanium dioxide via subcutaneous administration. After a thorough reliability screening, both of these references were considered of very limited relevance for hazard assessment purposes. The criteria for quality, reliability and adequacy of experimental data under REACH and for hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X) are not fulfilled. These references are discussed below, highlighting their findings and the reasons for their exclusion from hazard assessment:

Takeda et al. (2009) administered 100 µL of 1 mg/mL TiO2 nanoparticles (25 – 70 nm) in saline plus 0.05% Tween 80 via subcutaneous injections to pregnant mice at 3, 7, 10, and 14 days postcoitum. Control mice were treated on the same schedule with 0.05% Tween 80. The results showed that nano-sized titanium dioxide administered subcutaneously to pregnant mice is transferred to the offspring and affects the genital and cranial nerve system of the male offspring. Furthermore, nanoparticles identified as TiO2 by energy-dispersive X-ray spectroscopy were found in testis and brain of exposed 6-week-old male mice. In addition, in the offspring of TiO2-injected mice, various functional and pathologic disorders, such as reduced daily sperm production and numerous caspase-3 (a biomarker of apoptosis) positive cells in the olfactory bulb of the brain were observed. This reference has several reporting and experimental deficiencies which do not allow a independent review about the exposure-effect correlation:

-         the publication is not relevant for human health risk assessment due to irrelevant route of test item application.

-         TiO2 powder size was confirmed by FE-SEM (results in graphic form).

-         number of animals examined was not clearly stated.

-         frequency of treatment is not guideline conform.

-         one dose level was only tested

-         test item was insufficiently characterised.

 

Shimizu et al. (2009) investigated the effects of maternal exposure to nano-sized anatase titanium dioxide on gene expression in the brain during the developmental period using cDNA microarray analysis combined with Gene Ontology (GO) and Medical Subject Headings (MeSH) terms information. A 100 μL volume of TiO2 suspended at 1 μg/μL was injected subcutaneously into pregnant mice on gestational days 6, 9, 12, and 15 for the exposure group, while 100 μL of vehicle alone was injected into pregnant mice as a control group. Brain tissue was obtained from male foetuses on embryonic day 16 and from male pups on postnatal days 2, 7, 14, and 21. The results showed that maternal exposure to anatase TiO2 nanoparticle caused the changes in the expression of genes associated with brain development, cell death, response to oxidative stress, and mitochondria in the brain during the perinatal period, and those associated with inflammation and neurotransmitters in the later stage. This reference has several reporting and experimental deficiencies which do not allow a independent review about the exposure-effect correlation:

-         the publication is not relevant for human health risk assessment due to irrelevant route of test item application.

 

-         one dose was only tested, and the frequency of treatment is not according a guideline.

 

-         test item was insufficiently characterised, the given particle size of 2570 nm (2.57µm) does not fall under the nano-definition and also does not correspond with the surface area of 2025m²/g (which is also unrealistic high for any commercially available titanium dioxide form).

 

 

Summary entry – Toxicity to reproduction: other studies

Another reference was also identified during the comprehensive literature search, representing an in vivo experiment with special investigation on reproduction toxicity. The experiment was conducted with mice receiving nano-form titanium dioxide via intraperitoneal administration. The study design is not in accordance with accepted guidelines and are therefore of limited relevance for chemicals hazard assessment. The reference lacks significance due to poor test item characterisation and non-physiological route of administration. Therefore, it is concluded that the reference does not fulfil the criteria for quality, reliability and adequacy of experimental data for the fulfilment of data requirements under REACH and hazard assessment purposes (ECHA guidance R4 in conjunction with regulation (EC) 1907/2006, Annexes VII-X). The study given below was included in the IUCLID for information purposes only:

Smith, M. A. et al. (2015): The methodical setup is not adequately designed for risk assessment purposes of the test substance. The non-physiological route of administration via intraperitoneal injection is not guideline conform and not suitable to assess toxicity to reproduction. Furthermore, the test material is not sufficiently characterised for hazard assessment purposes.

Justification for classification or non-classification

Effects on fertility:

Based on the weight of evidence from the available long-term toxicity/carcinogenicity studies in rodents and the relevant information on the toxicokinetic behaviour in rats, it is concluded that titanium dioxide does not present a reproductive toxicity hazard. There is evidence from the animal chronic toxicity/carcinogenicity studies in rats and mice that the intake of high amounts of titanium dioxide or inhalation to high concentrations of titanium dioxide was not associated with adverse effects on the reproductive organs.

The results of a toxicokinetic study demonstrate that no relevant systemic absorption occurs via the oral exposure route as indicated by the titanium concentrations in whole-blood and urine which were below the limit of quantification (<0.04 mg/l). Tissue titanium concentrations in liver, kidney and muscle were below the limit of detection (<0.1 - <0.2 mg/kg wet weight) indicating no substantial accumulation of titanium in the body. Furthermore, any metabolism of the inorganic material whatsoever can be excluded.

Within the scope of the re-evaluation of titanium dioxide (E171) as food additive by the European Food Safety Authority (EFSA) adopted on the 28th June 2016, the conduct of a “multigeneration or extended-one generation reproduction toxicity study according to the current OECD guidelines” was recommended. This study was requested in order to establish a health-based guidance value (ADI) for the food additive titanium dioxide (EFSA Journal 2016;14(9):4545). The experimental phase for an extended one generation reproductive toxicity study is currently ongoing (status January 2020). Upon study completion, a robust study summary will also be included in the REACH registration dossier. Further experimental testing for the endpoint toxicity to reproduction within the scope of REACH is therefore considered not to be required.

 

Effects on developmental toxicity:

In six guideline and GLP compliant studies, titanium dioxide did not show any effects on the developing foetuses of rats when administered up to the limit dose of 1000 mg/kg bw/day. The test items covered titanium dioxide in the crystalline forms rutile and anatase in pigment and ultrafine grade. In conclusion, based on the data observed for titanium dioxide in six developmental toxicity studies in rats, a classification as reproductive toxicant is not justified due to the complete absence of any adverse effects.