Titanium dioxide

Carcinogenicity

Description of key information

The rat is uniquely sensitive to the formation of lung tumours when exposed under conditions of particle overload to titanium dioxide and other poorly soluble low-toxicity particles (Levy, 1995). Although particle overload is observed in other experimental species, such as the mouse, it is only in the rat that a sequence of events is initiated that leads to fibroproliferative disease, septal fibrosis, hyperplasia and eventually lung tumours. Similar pathological changes are not observed in other common laboratory rodents, in non-human primates, or in exposed humans. In addition, detailed epidemiological investigations have shown no causative link between titanium dioxide exposure and cancer risk in humans. At workplace exposure concentrations, no lung cancer hazard has been observed. Thus, a carcinogen rating for titanium dioxide is not warranted.

Key value for chemical safety assessment

Carcinogenicity: via oral route

Link to relevant study records

Reference

Reference 1

Endpoint:

carcinogenicity: oral

Type of information:

experimental study

Adequacy of study:

key study

Study period:

not specified

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

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

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

reference to same study

Qualifier:

no guideline followed

Principles of method if other than guideline:

Groups of 50 male and 50 female B6C3F1 mice each were fed a diet containing 2% corn oil and 25000 or 50000 ppm titanium dioxide for 103 weeks (7 days per week). A control group receiving corn oil in the diet was run concurrently. After the administration period the animals were observed for 1 additional week. The following parameters were assessed and presented: clinical signs, mortality, detailed clinical observations, body weight, and histopathology.

GLP compliance:

no

Specific details on test material used for the study:

not applicable

Species:

mouse

Strain:

B6C3F1

Details on species / strain selection:

not specified

Sex:

male/female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Frederick Cancer Research Center, Maryland
- Age at study initiation: 36 days
- Housing: housed in polycarbonate cages covered with stainless steel cage lids and non-woven fiber filter bonnets; mice were housed 5/cage; bedding material: heat-treated harwood chip bedding (Sani-Chips®)
- Diet (ad libitum): basal diet of Wayne® Lab Blox animal meal (Allied Mills, Inc., Chicago, Ill.)
- Water (ad libitum): well water
- Quarantine period: 15 days

ENVIRONMENTAL CONDITIONS
- Temperature: 20 - 24 °C
- Relative humidity: 45 - 55 %
- Air changes: 12/hour
- Photoperiod (hrs dark / hrs light): 12/12

Route of administration:

oral: feed

Vehicle:

corn oil

Remarks on MMAD:

not applicable

Details on exposure:

DIET PREPARATION
- Mixing appropriate amounts with basal diet of Wayne® Lab Blox animal meal: a quantity of the bulk chemical was sifted to remove any large particles, and the amount required for each dose mixture was weighed out under a hood. This quantity was then incorporated into the diet by thorough mixing in a Patterson-Kelly twin-shell blender equipped with an intensifier bar. Corn oil was added to the dosed diets and to the diets for the matched controls to give a final concentration of 2 %.
- Rate of preparation of diet (frequency): once per week
- Storage temperature of food: room temperature

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

As a quality control measure, selected samples from freshly prepared mixtures were stored at 4 °C and aliquots from these samples, containing approximately 50 micrograms of titanium dioxide were later analyzed for titanium dioxide by the method described by the Association of Official Analytical Chemists (1975)*.
Duplicate 100 mg subsamples of feed were ashed, and the residues fused with 2 g of potassium pyrosulfate. The fusion mixture was quantitatively transferred to a 100 mL volumetric flask using a 1:1 mixture of sulfuric acid and water, and diluted to volume with water. With a Tiron indicator, the transmittance of this solution was read at 410 nm. Concentrations of titanium dioxide were determined by comparison with standard solutions.
Recoveries were also determined from duplicate analyses of spiked samples worked up simultaneously with each set of dosed feed samples. The average recovery from the 2.5 % spiked samples was 97.5 %, and from the 5.0 % spiked samples, 100.3 %.

Results:
At each dietary concentration, the mean value obtained by the analytical method was within 4% of the theoretical value, although the coefficient of variation was nearly 30%. This variation appears to be due to the difficulty in obtaining a homogeneous mix of a fine powder in feed.

Theoretical concentrations in diet: 2.5 and 5.0 % in diet
Sample analytical mean: 2.4 and 4.9 % in diet (coefficient of variation: 26.3 and 29.5 %, respectivley)
Range: 2.2 - 2.9* and 4.79 - 6.85*
*Ranges exclude the two samples at each level during weeks 35 and 45 which analysed at only 40 - 50 % of the theoretical; these samples were included in the Number of Samples, Sample Analytical Mean, and Coefficient of Variation.

*Reference:
- Association of Official Analytical Chemists, Official Methods of Analysis of the Association of Official Analytical Chemists, 12th edition, Horwitz, W., ed., Association of Offical Analytical Chemists, Washington, B.C., 1975, p. 7.ebrc09

Duration of treatment / exposure:

103 weeks

Frequency of treatment:

7 days/week (ad libitum)

Post exposure period:

1 week

Dose / conc.:

25 000 ppm (analytical)

Remarks:

equivalent to 3750 mg/kg/day (recalculated from ppm value (factor (mouse): 0.150)

Dose / conc.:

50 000 ppm (analytical)

Remarks:

equivalent to 7500 mg/kg/day (recalculated from ppm value (factor (mouse): 0.150)

No. of animals per sex per dose:

50 males / 50 females

Control animals:

yes, concurrent vehicle

Details on study design:

- Dose selection rationale: subchronic feeding study was conducted to estimate the maximum tolerated doses of titanium dioxide, on the basis of which two concentrations were selected for administration in the chronic study. On the basis of results from a 14-day (repeated dose) oral range-finding study, doses of 6250, 12500, 25000, 50000, or 100000 ppm were administered in the diet in the subchronic study. Ten male and 10 female mice were administered the test chemical at each dose, and 10 males and 10 females received basal diets. Dosed animals received the test compound for 13 consecutive weeks.
There were no deaths, and dosed animals had mean body weight gains that were comparable to those of the controls. No gross or microscopic pathology was found that could be related to the administration of the test chemical. On the basis of these results, the high dose for mice in the chronic study was set at 50000 ppm and the low dose was set at 25000 ppm.

- Rationale for animal assignment: animals were assigned to the dosed or control groups based on initial individual body weight, so that the mean body weights per group were approx. equal.

Positive control:

not examined

Observations and examinations performed and frequency:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: twice daily
- Cage side observations checked: signs of toxicity and survival

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: every week

DERMAL IRRITATION: No

BODY WEIGHT: Yes
- Time schedule for examinations: every 2 weeks for the first 12 weeks and every month thereafter.

FOOD CONSUMPTION: Yes
- Time schedule for examinations: every 2 weeks for the first 12 weeks and every month thereafter.

FOOD EFFICIENCY:
- Body weight gain in kg/food consumption in kg per unit time X 100 calculated as time-weighted averages from the consumption and body weight gain data: No

WATER CONSUMPTION AND COMPOUND INTAKE: No
OPHTHALMOSCOPIC EXAMINATION: No
HAEMATOLOGY: No
CLINICAL CHEMISTRY: No
URINALYSIS: No
NEUROBEHAVIOURAL EXAMINATION: No

Sacrifice and pathology:

GROSS PATHOLOGY: Yes
HISTOPATHOLOGY: Yes

Animals that were moribund and those that survived to the termination of the study were killed. The pathologic evaluation consisted of gross and microscopic examination of major tissues, major organs, and all gross lesions from killed animals and from animals found dead. The tissues were preserved, embedded in paraffin, sectioned, and stained. The following tissues were examined microscopically: brain (frontal cortex and basal ganglia, parietal cortex and thalamus, and cerebellum and pons), pituitary, spinal cord (if neurologic signs were present), eyes (if grossly abnormal), oesophagus, trachea, salivary glands, mandibular lymph node, thyroid, parathyroid, heart, thymus, lungs and mainstem bronchi, liver, gallbladder, pancreas, spleen, kidney, adrenal, stomach, small intestine, colon, urinary bladder, prostate or uterus, testes or ovaries, sternebrae, femur, or vertebrae including marrow, mammary gland, tissue masses, and any gross lesion.
A few tissues from some animals were not examined, particularly from those animals that died early. Also, some animals may have been missing, cannibalized, or judged to be in such an advanced state of autolysis as to preclude histopathologic evaluation.

Statistics:

Product-limit procedure of Kaplan and Meier, method of Cox, Tarone's extensions of Cox's methods, linearity test, one-tailed Fisher exact test, Bonferroni inequality, Cochran-Armitage test for linear trend in proportions with continuity correction, and life-table methods

Clinical signs:

no effects observed

Dermal irritation (if dermal study):

not examined

Mortality:

mortality observed, treatment-related

Description (incidence):

- female mice: the result of the Tarone test for dose-related trend in mortality shows a significant (P = 0.001) positive dose-related trend.
- female mice: 33/50 (66%) of the 50000 ppm dose group, 39/50 (78%) of the 25000 ppm dose group, and 45/50 (90%) of the matched controls were alive at week 104.

Body weight and weight changes:

no effects observed

Food consumption and compound intake (if feeding study):

not specified

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:

not specified

Neuropathological findings:

not specified

Histopathological findings: non-neoplastic:

no effects observed

Histopathological findings: neoplastic:

no effects observed

Other effects:

not examined

Details on results:

CLINICAL SIGNS AND MORTALITY
- clinical signs observed in the dosed groups were comparable with those of the control group and included protrusion of the eyes, bloody crust surrounding the eyes, palpable nodules, tissue masses and/or wart-like lesions, localized sores, irritation and swelling of the testes, hunched appearance, and/or
thinness.
- alopecia (localized or generalized) was noted in all the control and dosed groups. Alopecia was more often observed in the control females than in the dosed females. The areas of alopecia were primarily located around the nose and head and progressed to generalized alopecia in some of the animals.
- animals in all of the dosed groups had white faeces.

MORTALITY
- male mice: the result of the Tarone test for dose-related trend in mortality is not significant.
- male mice: forty out of fifty (80%) of the 50000 ppm dose males, 40/50 (80%) of the 25000 ppm dose males, and 32/50 (64%) of the matched-control males were still alive at week 104.

BODY WEIGHT AND WEIGHT GAIN
- administration of titanium dioxide had no appreciable effect on the mean body weights of either the male or the female mice

HISTOPATHOLOGY: NON-NEOPLASTIC (data from the study were put in context with historical control data by the evaluator of this study)
The study describes several non-neoplastic findings observed in mice without however attributing any adversity to these. By putting these into context with historical control data, the effects can be considered to lack toxicological significance. The findings can be summarised briefly as follows:

- spleen haematopoiesis: no effects in mice (6-10% M / 4-8% F)

- kidney chronic inflammation: male mice only; incidence 8% (low) and 10% (high) vs. 6% in control; females not affected (0-6%)
historical controls (HC)* male mice (lymphatic infiltration used as correlate for chronic inflammation): 24.5%
Conclusion: the findings observed in male mice can be considered to be within the historical control data of this strain taken from a publication.

- liver necrosis: male mice only; incidence 16% (high) vs. 0% in control and low dose; females (0-2%) not affected
HC 7% M 5.9%F
Conclusion: the incidence in male mice is higher than historical control data for this strain of mice, but it needs to be noted that the incidence is already much higher than the HC in the concomitant control group, and the incidence in the low and high dose group is not dose related. It is therefore considered implausible that treatment with the test item is responsible for these findings observed in male mice.

- Uterus/endometrium cystic hyperplasia: mice only; incidence 86% (low) and 78% (high) vs. 35% in control (no dose relation)
Conclusion: the incidence observed in the control, low and high dose groups is within the range of historical control data.

HISTOPATHOLOGY: NEOPLASTIC
- sufficient numbers of mice of each sex were at risk for the development of late-appearing tumors.
- a low incidence of neoplasia was observed in both the control mice and dosed mice. These neoplasms were of the usual number and type observed in mice of this age and strain.
- a slightly increased number of hepatocellular carcinomas was observed in the 50000 ppm dose males. The incidence of tumours was not increased over that observed in historical-control groups of mice of this age and strain.
- degenerative, proliferative, and inflammatory lesions were also of the usual number and kind observed in aged B6C3F1 mice.

- the results of the Cochran-Armitage test for positive dose-related trend in incidences of tumors and those of the Fisher exact test for higher incidences of tumors in dosed groups than in control groups are not significant for any type of tumor occurring in either sex.
- a significant trend (P = 0.037) in the negative direction is observed in the incidence of follicular-cell adenomas of the thyroid in female mice, in which the incidence in the control group exceeds the incidences in the dosed groups.
- results of the Fisher exact test (P = 0.035 in the negative direction) for the comparison of the incidence of combined lymphomas and leukemias in the female 25000 ppm dose group with that in the corresponding controls are above that of 0.025 required for significance in multiple comparisons. This negative result may be accounted for by the difference in survival, since the dosed animals did not live as long as the control animals.
- in each of the 95% confidence intervals of relative risk the value of one is included; this indicates the absence of significant positive results. It should also be noted that each of the intervals has an upper limit greater than one, indicating the theoretical possibility of the induction of tumors by titanium dioxide, which could not be detected under the conditions of this test.

*Sources for historical control data:
Hirouchi Y., et al. (1994): Historical Data of Neoplastic lesions in B6C3F1 mice, J. Toxicol Pathol 7: 153-177
Goodman D.G., et al. (1978): Neoplastic and Non-neoplastic Lesions in Aging F344 Rats, Tox and Appl Pharmacol 48, 237-248
Coleman G.L. (1977): Pathological Changes During Aging in Barrier-reared Fischer 344 Male Rats, J Gerontol 32 (3):258-278

Relevance of carcinogenic effects / potential:

Increased incidences of neoplastic lesions were not observed.

Key result

Remarks on result:

not determinable due to absence of adverse toxic effects

Critical effects observed:

no

Conclusions:

NOEL (tumourogenicity; mice): 50000 ppm (equivalent to 7500 mg/kg/day)

According to the study authors, there was no clinical sign that was judged to be related to titanium dioxide exposure, with the exception of white faeces. In male and female mice, no tumours occurred in dosed groups at incidences that were significantly higher than those for corresponding control groups. It can therefore safely be concluded that under the conditions of this bioassay, titanium dioxide was not carcinogenic by the oral route for B6C3F1 mice.

Reference 2

Endpoint:

carcinogenicity: oral

Type of information:

experimental study

Adequacy of study:

key study

Study period:

not specified

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

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

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

reference to same study

Qualifier:

no guideline followed

Principles of method if other than guideline:

Groups of 50 male and 50 female Fischer 344 rats each were fed a diet containing 2% corn oil and 25000 or 50000 ppm titanium dioxide for 103 weeks (7 days per week). A control group receiving corn oil in the diet was run concurrently. After the administration period the animals were observed for 1 additional week. The following parameters were assessed and presented: clinical signs, mortality, detailed clinical observations, body weight, and histopathology.

GLP compliance:

no

Specific details on test material used for the study:

not applicable

Species:

rat

Strain:

Fischer 344

Details on species / strain selection:

not specified

Sex:

male/female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Frederick Cancer Research Center, Maryland
- Age at study initiation: 64 days
- Housing: housed in polycarbonate cages covered with stainless steel cage lids and non-woven fiber filter bonnets; rats were initially housed 5/cage, but starting at week 48 the males were divided into groups of 2 or 3/cage; bedding material: heat-treated harwood chip bedding (Sani-Chips®)
- Diet (ad libitum): basal diet of Wayne® Lab Blox animal meal (Allied Mills, Inc., Chicago, Ill.)
- Water (ad libitum): well water
- Quarantine period: 30 days

ENVIRONMENTAL CONDITIONS
- Temperature: 20 - 24 °C
- Relative humidity: 45 - 55 %
- Air changes: 12/hour
- Photoperiod (hrs dark / hrs light): 12/12

Route of administration:

oral: feed

Vehicle:

corn oil

Remarks on MMAD:

not applicable

Details on exposure:

DIET PREPARATION
- Mixing appropriate amounts with basal diet of Wayne® Lab Blox animal meal: a quantity of the bulk chemical was sifted to remove any large particles, and the amount required for each dose mixture was weighed out under a hood. This quantity was then incorporated into the diet by thorough mixing in a Patterson-Kelly twin-shell blender equipped with an intensifier bar. Corn oil was added to the dosed diets and to the diets for the matched controls to give a final concentration of 2 %.
- Rate of preparation of diet (frequency): once per week
- Storage temperature of food: room temperature

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

As a quality control measure, selected samples from freshly prepared mixtures were stored at 4 °C and aliquots from these samples, containing approximately 50 micrograms of titanium dioxide were later analyzed for titanium dioxide by the method described by the Association of Official Analytical Chemists (1975)*.
Duplicate 100 mg subsamples of feed were ashed, and the residues fused with 2 g of potassium pyrosulfate. The fusion mixture was quantitatively transferred to a 100 mL volumetric flask using a 1:1 mixture of sulfuric acid and water, and diluted to volume with water. With a Tiron indicator, the transmittance of this solution was read at 410 nm. Concentrations of titanium dioxide were determined by comparison with standard solutions.
Recoveries were also determined from duplicate analyses of spiked samples worked up simultaneously with each set of dosed feed samples. The average recovery from the 2.5 % spiked samples was 97.5 %, and from the 5.0 % spiked samples, 100.3 %.

Results:
At each dietary concentration, the mean value obtained by the analytical method was within 4% of the theoretical value, although the coefficient of variation was nearly 30%. This variation appears to be due to the difficulty in obtaining a homogeneous mix of a fine powder in feed.

Theoretical concentrations in diet: 2.5 and 5.0 % in diet
Sample analytical mean: 2.4 and 4.9 % in diet (coefficient of variation: 26.3 and 29.5 %, respectivley)
Range: 2.2 - 2.9* and 4.79 - 6.85*
*Ranges exclude the two samples at each level during weeks 35 and 45 which analysed at only 40 - 50 % of the theoretical; these samples were included in the Number of Samples, Sample Analytical Mean, and Coefficient of Variation.

*Reference:
- Association of Official Analytical Chemists, Official Methods of Analysis of the Association of Official Analytical Chemists, 12th edition, Horwitz, W., ed., Association of Offical Analytical Chemists, Washington, B.C., 1975, p. 7.

Duration of treatment / exposure:

103 weeks

Frequency of treatment:

7 days/week (ad libitum)

Post exposure period:

1 week

Dose / conc.:

25 000 ppm (analytical)

Remarks:

equivalent to 1250 mg/kg/day (recalculated from ppm value (factor (older rat): 0.050)

Dose / conc.:

50 000 ppm (analytical)

Remarks:

equivalent to 2500 mg/kg/day (recalculated from ppm value (factor (older rat): 0.050)

No. of animals per sex per dose:

50 males / 50 females

Control animals:

yes, concurrent vehicle

Details on study design:

- Dose selection rationale: subchronic feeding study was conducted to estimate the maximum tolerated doses of titanium dioxide, on the basis of which two concentrations were selected for administration in the chronic study. On the basis of results from a 14-day (repeated dose) oral range-finding study, doses of 6250, 12500, 25000, 50000, or 100000 ppm were administered in the diet in the subchronic study. Ten male and 10 female rats were administered the test chemical at each dose and 10 males and 10 females received basal diets. Dosed animals received the test compound for 13 consecutive weeks.
There were no deaths, and dosed animals had mean body weight gains that were comparable to those of the controls. No gross or microscopic pathology was found that could be related to the administration of the test chemical. On the basis of these results, the high dose for rats in the chronic study was set at 50000 ppm and the low dose was set at 25000 ppm.

- Rationale for animal assignment: animals were assigned to the dosed or control groups based on initial individual body weight, so that the mean body weights per group were approx. equal.

Positive control:

not examined

Observations and examinations performed and frequency:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: twice daily
- Cage side observations checked: signs of toxicity and survival

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: every week

DERMAL IRRITATION: No

BODY WEIGHT: Yes
- Time schedule for examinations: every 2 weeks for the first 12 weeks and every month thereafter.

FOOD CONSUMPTION: Yes
- Time schedule for examinations: every 2 weeks for the first 12 weeks and every month thereafter.

FOOD EFFICIENCY:
- Body weight gain in kg/food consumption in kg per unit time X 100 calculated as time-weighted averages from the consumption and body weight gain data: No

WATER CONSUMPTION AND COMPOUND INTAKE: No
OPHTHALMOSCOPIC EXAMINATION: No
HAEMATOLOGY: No
CLINICAL CHEMISTRY: No
URINALYSIS: No
NEUROBEHAVIOURAL EXAMINATION: No

Sacrifice and pathology:

GROSS PATHOLOGY: Yes
HISTOPATHOLOGY: Yes

Animals that were moribund and those that survived to the termination of the study were killed. The pathologic evaluation consisted of gross and microscopic examination of major tissues, major organs, and all gross lesions from killed animals and from animals found dead. The tissues were preserved, embedded in paraffin, sectioned, and stained. The following tissues were examined microscopically: brain (frontal cortex and basal ganglia, parietal cortex and thalamus, and cerebellum and pons), pituitary, spinal cord (if neurologic signs were present), eyes (if grossly abnormal), oesophagus, trachea, salivary glands, mandibular lymph node, thyroid, parathyroid, heart, thymus, lungs and mainstem bronchi, liver, pancreas, spleen, kidney, adrenal, stomach, small intestine, colon, urinary bladder, prostate or uterus, testes or ovaries, sternebrae, femur, or vertebrae including marrow, mammary gland, tissue masses, and any gross lesion.
A few tissues from some animals were not examined, particularly from those animals that died early. Also, some animals may have been missing, cannibalized, or judged to be in such an advanced state of autolysis as to preclude histopathologic evaluation.

Statistics:

Product-limit procedure of Kaplan and Meier, method of Cox, Tarone's extensions of Cox's methods, linearity test, one-tailed Fisher exact test, Bonferroni inequality, Cochran-Armitage test for linear trend in proportions with continuity correction, and life-table methods

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):

not specified

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:

not specified

Neuropathological findings:

not specified

Histopathological findings: non-neoplastic:

no effects observed

Histopathological findings: neoplastic:

no effects observed

Other effects:

not examined

Details on results:

CLINICAL SIGNS
- clinical signs observed in the dosed groups were generally comparable to those of the control group and included alopecia, sores, and lacrimating, protruding, and/or pale eyes.
- from weeks 88 through 104, hunched appearance and thinness were noted more frequently in the dosed males and females than in their respective controls (not test-item related).
- urine stains were noted on the dosed rats of each sex (not test-item related).
- animals in all of the dosed groups had white faeces.

MORTALITY
- result of the Tarone test for dose-related trend in mortality is not significant in either sex.
- male rats. 36/50 animals (72%) of the 50000 ppm dose group, 37/50 animals (74%) of the 25000 ppm dose group and 31/50 animals (62%) of the matched controls were alive at week 104.
- females rats: 34/50 animals (68%) of the 50000 ppm dose group, 36/50 animals (72%) of the 25000 ppm dose group and 36/50 (72%) of the matched controls were alive at week 104.

BODY WEIGHT AND WEIGHT GAIN
- administration of titanium dioxide had no appreciable effect on the mean body weights of either the male or the female rats.

HISTOPATHOLOGY: NON-NEOPLASTIC (data from the study were put in context with historical control data by the evaluator of this study)
The study describes several non-neoplastic findings observed in rats without however attributing any adversity to these. By putting these into context with historical control data, the effects can be considered to lack toxicological significance. The findings can be summarised briefly as follows:

- spleen haematopoiesis: observed in male rats only; incidence 8% (high) vs 0% in control and 2% in low dose; no effects in females (0-4%);
historical controls (HC)* for extramedullary haematopoiesis rat (18-24 month): 7.5 %
Conclusion: the finding in male rats is most probably not related to treatment with the test item but spontaneous in nature, because the incidence in males is generally low. It was observed in females only with a very low incidence. Information from a publication on the pathology of aging male Fischer rats indicates a percentage for haematopoiesis of 7.5% which is very close to the rate observed in this study (8%). Overall, it is not very likely that the finding of haematopoiesis in spleen observed in male Fischer rats is related to treatment with the test item.

- kidney chronic inflammation: male rat only; incidence 90% and 86% in low dose and high dose vs 59% in control (no dose relation); no effects in female rats (38-53%)
HC rats (chronic interstitial nephritis used as correlate for chronic inflammation): 62.5 % (18-24 months), 70.2 % (24-30 months)
Conclusion: the higher incidence of chronic inflammation in male rats observed in the low and high dose groups is not dose-related, although higher than in control males. In general, the incidence of this lesion is high in aged Fischer rats (18-30 months) based on published data (62.5-70.2%). However, the incidences observed in male rats may reflect a chronic infection in these animals and is most probably not related to treatment with the test item.

- seminal vesicle (SV) and testis atrophy: SV rat: incidence 12% (low) and 20% (high) vs 0% in controls
Testis rat: incidence 10% (low) and 14% (high) vs 6% in control
HC seminal vesicles and testis: 1.7% and 12.4%
HC seminal vesicles and testis: 2.3% (24-30%) and 80-100% (18-30 months)
Conclusion: the incidence observed in seminal vesicles of male rats is indeed higher than described in two publication on historical control data observed in studies (1.7%) conducted by the National Cancer Institute or described in a publication about pathology in aging Fischer rats (2.3%). However, the toxicological relevance of this observation is not clear. In addition, the findings in testes are within the HC in the low dose group and only slightly above the HC data (12.4%) in the high dose group. In general, the findings observed in seminal vesicle and testis of Fischer rats should not be considered as related to treatment with the test item.

- liver necrosis: rats (0-2%) not affected

- uterus/endometrium cystic hyperplasia: female rats (0-6%)

HISTOPATHOLOGY: NEOPLASTIC
Male rats:
- pheochromocytomas of the adrenal medulla (matched control: 7/49 (14 %); 25000 ppm dose: 9/49 (18 %); 50000 ppm dose: 14/50 (28 %)) and fibromas of the subcutaneous tissue (matched control: 1/49 (2 %); 25000 ppm dose: 5/50 (10 %); 50000 ppm dose 5/50 (10 %) were observed with slightly greater frequency in dosed groups. The number of neoplasms was compatible with incidences of these tumours in historical-control rats of this age and strain. Thus, these lesions are not considered to be related to administration of the test chemical.
- three keratoacanthomas of the skin were observed in the 50000 ppm dose group, but none in the other two groups studied. Although the result of the Fisher exact test for direct comparison of the incidence in the 50000 ppm group with that in the control group is not significant, the result of the Cochran-Armitage test for positive dose-related trend in the incidence of these tumors is significant (P = 0.038).
- significant results in the negative direction are observed in the incidence of leukemia, in which the incidence in the control group exceeds the incidences in the dosed groups.

Female rats:
-endometrial stromal polyps were observed more frequently in dosed groups (matched control: 6/50 (12 %); low dose: 15/50 (30 %); high dose 10/49 (20 %)) than in control groups, but the incidence of lesions is comparable with that in historical controls. Thus, these lesions are not considered to be related to administration of the test chemical.
- C-cell adenomas or carcinomas of the thyroid occurred at incidences that were dose related (P = 0.013), but were not high enough (P = 0.043 for direct comparison of the 50000 ppm dose group with the control group) to meet the level of P = 0.025 required by the Bonferroni criterion (controls 1/48, 25000 dose 0/47,50000 ppm dose 6/44). Thus, these tumors of the thyroid were not considered to be related to the administration of the test chemical.
- Fisher exact comparison of the incidence of endometrial stromal polyps of the uterus/endometrium in the 25000 ppm dose females with that in the corresponding controls indicates a P value of 0.045, which is above the 0.025 level required for significance when the Bonferroni inequality criterion is used for multiple
comparison. The incidence of these tumors in the 50000 ppm dose group is not significant when compared with that in the control group, and the result of the Cochran-Armitage test for dose-related trend also is not significant.

Inflammatory, degenerative, and hyperplastic lesions that occurred were similar in number and kind to those naturally occurring lesions found in aged Fischer 344 rats.

*Sources for historical control data:
Hirouchi Y., et al. (1994): Historical Data of Neoplastic lesions in B6C3F1 mice, J. Toxicol Pathol 7: 153-177
Goodman D.G., et al. (1978): Neoplastic and Non-neoplastic Lesions in Aging F344 Rats, Tox and Appl Pharmacol 48, 237-248
Coleman G.L. (1977): Pathological Changes During Aging in Barrier-reared Fischer 344 Male Rats, J Gerontol 32 (3):258-278

Relevance of carcinogenic effects / potential:

Increased incidences of neoplastic lesions were not observed.

Key result

Remarks on result:

not determinable due to absence of adverse toxic effects

Critical effects observed:

no

Conclusions:

NOEL (tumourogenicity; rats): 50000 ppm (equivalent to 2500 mg/kg/day)

According to the study authors, there was no clinical sign that was judged to be related to titanium dioxide exposure, with the exception of white faeces. In female rats, C-cell adenoma or carcinoma of the thyroid occurred at incidences that were dose related (P = 0.013), but were not high enough (P = 0.043 for direct comparison of the 50000 ppm dose group with the control group) to meet the level of P = 0.025 required by the Bonferroni criterion (controls 1/48, 25000 ppm dose 0/47, 50000 ppm dose 6/44). Thus, these tumours of the thyroid were not considered to be related to the administration of the test chemical. It can therefore safely be concluded that under the conditions of this bioassay, titanium dioxide was not carcinogenic by the oral route for Fischer 344 rats.

Reference 3

Endpoint:

carcinogenicity: oral

Type of information:

experimental study

Adequacy of study:

supporting study

Study period:

not specified

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

other: This study was conducted with a test item, rating this study not relevant. However, since the material contained a significant amount of titanium dioxide (28% TiO2), animals received a relevant dose of that substance, rendering this a supporting study.

Remarks:

This study was rated not reliable, since the conduct of the histopathology was not fully described.

Qualifier:

no guideline followed

Principles of method if other than guideline:

In this study groups of 60 male and 60 female Fischer 344 rats were given TiO2-coated mica (28% TiO2, 72% mica; flat platelets 10-35 µm) via diet ad libitum for 103 weeks (vehicle: 1% corn oil). The following dietary levels of 0, 1.0, 2.0 and 5.0% (equivalent to 10000, 20000, and 50000 ppm) were administered. A vehicle control group was run concurrently. Clinical signs, mortality, body weight, fodd consumption, ophthalmoscopic examination, haematology, clinical chemistry, urinalysis, organ weights, and histopathology were recorded as parameters.

GLP compliance:

not specified

Specific details on test material used for the study:

not applicable

Species:

rat

Strain:

Fischer 344

Details on species / strain selection:

Fischer 344 rats were chosen as the test species because of the existence of a large historical data base to aid in the interpretation of spontaneous tumour incidences in aged animals (Haseman et al., 1984; Solleveld et al., 1984)*.
*References:
- Haseman, J.K., Huff, J.E. and Boorman, G.A. 1984. Use of historical control data in carcinogenicity studies in rodents. Toxicol. Pathol. 12: 126 - 135.
- Solleveld, H., Haseman, J.K., and McConnell, E.E. 1984. Natural history of body weight gain, survival, and neoplasia inthe F344 rat. J. NCI 72: 929 - 940.

Sex:

male/female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River Laboratories, Kingston, N.Y.
- Age at study initiation: 6 weeks old
- Weight at study initiation: males: 104 to 166 g; females: 91 to 125 g
- Housing: housed individually in elevated stainless steel mesh cages
- Diet (ad libitum): Purina rodent laboratory chow supplemented with 1 % corn oil
- Water (ad libitum): tap water
- Acclimation period: 2 weeks

ENVIRONMENTAL CONDITIONS
- Temperature: 21.1 ± 4°C
- Relative humidity: 50 ± 20 %
- Photoperiod (hrs dark / hrs light): 12/12

Route of administration:

oral: feed

Vehicle:

corn oil

Remarks on MMAD:

not applicable

Details on exposure:

DIET PREPARATION
- Rate of preparation of diet (frequency): weekly
- Mixing appropriate amounts with Purina rodent laboratory chow

VEHICLE
- Amount of vehicle: 1 %

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

Samples were assayed periodically to verify concentration, homogeneity, and stability.

Duration of treatment / exposure:

130 weeks

Frequency of treatment:

ad libitum

Post exposure period:

not applicable

Dose / conc.:

10 000 ppm

Dose / conc.:

20 000 ppm

Dose / conc.:

50 000 ppm

No. of animals per sex per dose:

60 male rats / 60 female rats

Control animals:

yes, concurrent vehicle

Details on study design:

not applicable

Positive control:

not specified

Observations and examinations performed and frequency:

CAGE SIDE OBSERVATIONS/PHYSICAL EXAMINATION: Yes
- Time schedule: weekly for the first 14 weeks, then every 4 weeks thereafter.

DETAILED CLINICAL OBSERVATIONS: No data
DERMAL IRRITATION: No data

BODY WEIGHT: Yes
- Time schedule for examinations: weekly for the first 14 weeks, then every 4 weeks thereafter.

FOOD CONSUMPTION: Yes
- Time schedule for examinations: weekly for the first 14 weeks, then every 4 weeks thereafter.

FOOD EFFICIENCY:
- Body weight gain in kg/food consumption in kg per unit time X 100 calculated as time-weighted averages from the consumption and body weight gain data: No data

WATER CONSUMPTION AND COMPOUND INTAKE: No data

OPHTHALMOSCOPIC EXAMINATION: Yes
- Time schedule for examinations: prior to the start of the study and during week 52 and 104
Animals exhibiting ocular anbormalities during the prestudy examination were replaced.

HAEMATOLOGY: Yes
- Time schedule for collection of blood: blood was obtained from the orbital sinus in week 27 and 53.
- Anaesthetic used for blood collection: No data
- Animals fasted: Yes, overnight (water reamained available)
- How many animals: 10 animals/sex/group

CLINICAL CHEMISTRY: Yes
- Time schedule for collection of blood: blood was obtained from the orbital sinus in week 27 and 53.
- Animals fasted: Yes, overnight (water reamained available)
- How many animals: 10 animals/sex/group

URINALYSIS: Yes
- Time schedule for collection of urine: collection during week 27 and 53
- Metabolism cages used for collection of urine: Yes
- Animals fasted: Yes, overnight (water reamained available)

NEUROBEHAVIOURAL EXAMINATION: No data

Sacrifice and pathology:

Following blood and urine sample collection during week 53, the animals in the clinical pathology subgroup were killed and necropsied. Organ/body weight ratios were determined for brain, heart, liver, spleen, kidney, testes (with epididymides), both adrenals, thyroid, ovaries, and pituitary. The control and high dose groups were subjected to complete histopathologic evaluation.

All surviving animals were killed and subjected to complete necropsy after 130 weeks of treatment. Complete necropsies were also performed on animals found dead or killed in a moribund state. Paraplast-embedded tissues and visible lesions were stained wiith hematoxylin and eosin and evaluated microscopically from all animals in the control and high dose groups. All tissues and visible lesions from animals that died during the study, as well as grossly visible lesions from low and mid dose animals that survived to the scheduled termination of the study were also examined microscopically.

Statistics:

Bartlett's test (Bartlett, 1937)* was used to confirm homogeneity of variance, and then one-way analyses of variance (ANOVA)(Snedecor and Cochran, 1967)* were performed. If significant F-values were obtained, Scheffe's multiple pairwise comparison procedure (Scheffe, 1953)* was used to compare treated group means to those of the control group. In certain instances the data were transformed to achieve homogeneity of variance. Where transformations were unsuccessful, heterogeneous data were tested by ANOVA, and then by Games and Howell's multiple pairwise comparison procedure (Games and Howell, 1976)* if the F-value was signifcant. Cataract incidence was evaluated by the Cochran-Armitage test for linear trend (Cochran, 1954; Armitage, 1955)*, followed by Fisher's exact test (Fisher, 1950)* to compare treatment to control incidences. The incidence of mononuclear cell leukemia was evaluated by Fisher's exact test with Bonferroni adjustment (Miller, 1966)* to enable multiple group comparisons. Survival rate was evaluated by a National Cancer Institute package (Thomas et al., 1978)*. In all cases, significant differences were determined to exist a probability level of p ≤ 0.05 (one-tail).

*References:
- Bartlett, M.S. 1937. J. R. Stat. Soc. Suppl. 4: 137 - 170.
- Snedecor, G.W. and Cochran, W.G. 1967. Statistical Methods, 6th ed. Ames: Iowa State University Press.
- Scheffe, H. 1953. Biometrika 40: 87 - 104.
- Games, P. A., and Howell, J. F. 1976. Pairwise multiple comparison procedures with unequal N's and/or variances: A Monte Carlo study. J. Ed. Stat. 1: 113 - 125.
- Cochran, W. G. 1954. Biometrics 10: 417 - 451.
- Armitage, P. 1955. Biometrics 11: 375 - 385.
- Fisher, R. A. 1950. Statistical Methods for Research Workers. Edinburgh: Oliver and Boyd.
- Miller, R. G., Jr. 1966. Simultaneous Statistical Inference, Chap. 61. New York: McGraw-Hill.
- Thomas, D. G., Breslow, N., and Gart, J. J. 1978. Comput. Biomed. Res. 10: 373 - 381.

Clinical signs:

effects observed, treatment-related

Description (incidence and severity):

- 5 % concentration: from week 2 of the study, animals administered the 5 % diets had silver-coloured faeces (treatment-related effect).

Dermal irritation (if dermal study):

not specified

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 specified

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

not specified

Ophthalmological findings:

no effects observed

Haematological findings:

no effects observed

Clinical biochemistry findings:

no effects observed

Urinalysis findings:

no effects observed

Behaviour (functional findings):

not specified

Immunological findings:

not specified

Organ weight findings including organ / body weight ratios:

no effects observed

Gross pathological findings:

not specified

Neuropathological findings:

not specified

Histopathological findings: non-neoplastic:

no effects observed

Histopathological findings: neoplastic:

no effects observed

Other effects:

not specified

Details on results:

CLINICAL SIGNS.
- during week 42, males from all groups exhibited periorbital oedema, keratoconjunctivitis, and weight loss due to a sialodacryoadenitis infection. All signs of infection cleared by week 46.

MORTALITY:
- 1.0 % concentration: statistical analysis of survival curves revealed increased mortality among females (week 130 (number survived): controls: 24/50 vs. test item: 12/50; p < 0.05). The survival of the animals of this group was equivalent to that of controls as late as week 102 when 78 % of the controls and 80 % of the low dose animals were still alive.
- survival of the females of the 2.0 % and 5.0 % concentration groups and all dosed male groups was not affected by the administration of the test item.

BODY WEIGHT AND WEIGHT CHANGES:
- 5 % concentration: at week 25 the mean body weights of the male and female groups were significantly less than those of control (p < 0.05). The rate of body weight gain in the female group was reduced relative to controls.
- at termination of the study, absolute body weights and body weight gains were comparable across groups.

FOOD CONSUMPTION:
- 5 % concentration: during the initial 50 weeks of the study the food consumption of the groups was significantly greater than that of the controls.
- during the final 78 weeks food consumption was comparable across groups within sexes.

OPHTHALMOLOGICAL FINDINGS:
- examinations during week 52 revealed an evenly distributed incidence of common ocular lesions in all groups of rats, with the most frequent being cataract.
- after 104 weeks there was a dose-related increase in grossly observed cataracts in dosed males. The incidence in the 5 % concentration group (12/46, 26 %) was significantly greater than the control incidence (2/43, 2.3 %; p < 0.05).
- no evidence of increased cataract formation in dosed female rats at either examination period.

- histopathologic evaluation of ocular tissues showed an even distribution of microscopic cataracts in both male and female rats. Microscopic characterization of cataracts was based on the presence of subcapsular and/or cortical mineralization and degeneration of lens fibers. Using these criteria, the overall incidences of cataracts were unchanged across groups within sexes.

HAEMATOLOGY:
- determinations conducted at weeks 27 and 53 revealed occasional variations from control values but none were consistent or dose related.

CLINICAl CHEMISTRY:
- determinations conducted at weeks 27 and 53 revealed occasional variations from control values but none were consistent or dose related.

URINALYSiS:
- results of urine analyses were unremarkable.

ORGAN WEIGHTS:
- comparisons of absolute and relative organ weights revealed lower mean absolute and relative thyroid weights for the 5 % concentration-treated males and the 2 % concentration-treated females (not dose-related).
- single occurrences of statistically significant differences in absolute and/or relative liver, brain, and pituitary weights were noted (not dose-related; not treatment-related).

HISTOPATHOLOGY
- long term administration of the test item was associated with a significant increase in the overall incidence of adrenal medullary hyperplasia in the 5 % concentration-treated males. There was no evidence of progression of this change to either benign or malignant pheochromocytoma, and a similiar proloferative effect was not observed in females. Malignant pheochromocytoma was not detected in female rats.
- high incidence of mononuclear cell leukemia was found in all groups of both sexes, and the overall incidence in 5 % concentration-treated males was somewhat elevated relative to controls (dose-related, but also age-related).
- overall rates of mononuclear cell leukemia in control and dosed female rats were comparable. Incidences in females that survived to the scheduled termination of the study were variable, but statistically significant differences between groups were not detected.
- other microscopic changes found in the animals of this study were typical of aged F344 rats and were considered to be unrelated to the administration of the test item.

Remarks on result:

not determinable due to absence of adverse toxic effects

Critical effects observed:

not specified

Conclusions:

In this study, TiO2-coated mica (28% TiO2, 72% mica; flat platelets 10-35 µm) was administered for 103 weeks (vehicle: Purina rat chow supplemented with 1% corn oil) at dietary levels of 0, 1.0, 2.0 and 5.0% (equivalent to 10000, 20000, and 50000 ppm). There were no effects at any dose on body weight and survival, no changes in haematological or clinical chemistry parameters, no treatment-related histopathological observations. The only observations commented on by the study authors were (i) reduced mean absolute and relative thyroid weights for high dose males and mid-dose females which were obviously without dose-relation, and (ii) an increased incidence of adrenal medullary hyperplasia in high dose males which however did not progress to either benign or malignant pheochromocytoma, such effect being absent in females. A high incidence of mononuclear cell leukaemia was found in all groups of both sexes is considered an age-related effect of doubtful relevance. Overall, under the conditions of this study, there was no evidence for the induction of carcinogenic effects. The NOEL from this study can be established at the highest dietary exposure level of 5% (corresponding to 1.4% TiO2, equivalent to 14000 ppm).

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed

Study duration:

chronic

Species:

rat

Carcinogenicity: via inhalation route

Link to relevant study records

Reference

Reference 1

Endpoint:

carcinogenicity: inhalation

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

not specified

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

significant methodological deficiencies

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

reference to same study

Qualifier:

no guideline followed

Principles of method if other than guideline:

Female NMRI mice were exposed 18 hours/day, 5 days/week for 13.5 months to ultrafine titanium dioxide (P25, Degussa; MMAD: 0.80 µm) and were subsequently kept in clean air for 9.5 months. The type of inhalation was whole body and the average particle exposure concentrations for the test substance was 10 mg/m³. A control group receiving clean air was run concurrently. The following observations were performed: clinical signs, mortality, body weight, lung wet weight, lung retention of inhaled particles, and histopathology.

GLP compliance:

not specified

Specific details on test material used for the study:

not applicable

Species:

mouse

Strain:

other: NMRI (Crl:NMRI BR)

Details on species / strain selection:

not specified

Sex:

female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River Wiga GmbH, Sulzfeld, Germany
- Age at study initiation: 7 weeks ± 3 days
- Housing: housed in groups of 8 - 10 per cage; kept in wire mesh cages during the exposure period, and in Makrolon cages (18.7 x 21 x 15 cm) during the subsequent clean air period. Softwood bedding (H3/4) was used.
- Diet (ad libitum): "1324 N spec. prepared" (Altromin, Lage, Germany)
- Water (ad libitum): drinking water

ENVIRONMENTAL CONDITIONS
- Temperature: 23 - 25 °C
- Relative humidity: 50 - 70%
- Photoperiod: 12/12

Route of administration:

inhalation: aerosol

Type of inhalation exposure (if applicable):

whole body

Vehicle:

clean air

Mass median aerodynamic diameter (MMAD):

0.8 µm

Geometric standard deviation (GSD):

1.8

Details on exposure:

GENERATION OF TEST ATMOSPHERE / CHAMBER DESCRIPTION
- Exposure apparatus: special whole-body exposure chambers (6 or 12 m³) of the horizontal flow type (Heinrich et al., 1985)*.
The mice were exposed together with Wistar rats in the same exposure chamber (for information on the rats please refer to Section 7.7 Carcinogenicity: NANO_s_Heinrich_1995_rats).

- System of generating particulates/aerosols: aerosols of the test item were generated by a dry dispersion technique using a screw feeder and a pressurized air dispersion nozzle. The mass median aerodynamic diameter of the aerosol was about 1.5 µm. In order to increase the deposition efficiency of the test aerosol in the deep lung, the particle size distribution was shifted toward smaller particles in the submicrometer regime by removing the coarse particles using a cyclone (50% cut-off diameter ≈ 1 mm for a flow rate of 100 m³/h).

- Method of particle size determination: the mass median aerodynamic diameter (MMAD) and the geometric standard deviation of the particles in the exposure chambers were measured every month (n = 24) with a 10-stage Berner impactor (LPI 0.01525; range 15 nm to 16 µm).

*Reference:
- Heinrich, U., Muhle, H., Koch, W., and Mohr, U. 1985. Long-term inhalation studies with rodents. In Safety evaluation and regulation of chemicals 2, ed. F. Homburger, pp. 239 - 250. Basel: Karger.

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

The particle concentration in the exposure chamber was determined continuously, using aerosol photometers (Koch et al., 1986)*. For calibration of the photometer, the aerosol concentration in each exposure chamber was determined gravimetrically at weekly intervals (Heinrich et al., 1986)*.
The following mean particle mass exposure concentrations were measured:
7.2 mg titanium dioxide/m³ for the first 4 months, followed by 14.8 mg titanium dioxide/m³ for 4 months and 9.4 mg titanium dioxide/m³ for 5.5 months

To compare the different exposure concentrations, the cumulative particle exposure (g/m³ x h) was calculated by multiplying the mean particle mass exposure concentration by the actual exposure time per day and subsequently summarizing for the whole exposure period: 51.5 g/m³ x h.

*References:
- Koch, W. Lödding, H., Oenning, G., and Muhle, H. 1986. The generation and measurement of dry aerosols in large scale inhalation experiments., J. Aerosol Sci. 19: 1453 - 1457.
- Heinrich, U., Muhle, H., Takenaka, S., Ernst, H., Fuhst, R., Mohr, Z., Pott, F., and Stöber, W. 1986. Chronic effects on the respiratory tract of hamsters, mice and rats after long-term inhalation of high concentrations of filtered and unfiltered diesel engine emissions. J. Appl. Toxicol. 6: 383 - 395.

Duration of treatment / exposure:

13.5 months

Frequency of treatment:

18 hours/day, 5 days/week

Post exposure period:

After a total exposure time of 13.5 months, the exposure for all groups was stopped and the animals were kept in clean air for another 9.5 months, at the most. The total experimental time for all groups was 23 months.

Dose / conc.:

10 mg/m³ air (analytical)

Remarks:

Standard deviation: 2.9

No. of animals per sex per dose:

Treatment group (total: 160 female mice):
- carcinogenicity: 80 mice
- histology (serial sacrifice): 40 mice
- Particle mass/lung (serial sacrifice): 40 mice

Control group (total: 160 female mice):
- carcinogenicity: 80 mice
- histology (serial sacrifice): 40 mice
- Particle mass/lung (serial sacrifice): 40 mice

Control animals:

yes, concurrent vehicle

Positive control:

not specified

Observations and examinations performed and frequency:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: once daily

DETAILED CLINICAL OBSERVATIONS: No data
DERMAL IRRITATION: No data

BODY WEIGHT: Yes
- Time schedule for examinations: every fourth week

FOOD CONSUMPTION AND COMPOUND INTAKE:
- Food consumption for each animal determined and mean daily diet consumption calculated as g food/kg body weight/day: No data
- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: No data

FOOD EFFICIENCY:
- Body weight gain in kg/food consumption in kg per unit time X 100 calculated as time-weighted averages from the consumption and body weight gain data: No data

WATER CONSUMPTION AND COMPOUND INTAKE: No data
OPHTHALMOSCOPIC EXAMINATION: No data
HAEMATOLOGY: No data
CLINICAL CHEMISTRY: No data
URINALYSIS: No data
NEUROBEHAVIOURAL EXAMINATION: No data

Sacrifice and pathology:

GROSS PATHOLOGY / HISTOPATHOLOGY: Yes
Necropsies of dead or moribund animals were done 7 days/week.
The lung wet weight and lung retention of inhaled particles were determined at different time points during the study. The lung wet weight was determined at 3, 6, 12, 18, and 21 months after the start of exposure. Titanium dioxide samples were determined by atomic absorption spectroscopy after ashing the lungs.

For the histopathological investigations, the organs of scheduled or moribund sacrifices were fixed in 10% neutral buffered formalin or lto­ Karnovsky fixative (Ito & Karnovsky, 1968)*. The tissues were embedded in Paramat-Wax, sectioned at 5 µm, and stained with hematoxylin and eosin (Lilly-Meyer). Histopathological investigations of the following organs were conducted for all animals: nasal and paranasal cavities (four sections; localization according to Popp & Monteiro-Riviere, 1985*), larynx, trachea, and lung (five sections; localization: the left lobe, right caudal lobe, and right middle lobe were sectioned longitudinally, and the right cranial lobe and accessory lobe were sectioned transversely to main bronchus). Graduation of the findings was done with four grades: very slight, slight, moderate, and high.

*References:
- Ito, S., and Karnovsky, M. J. 1968. Formaldehyde-glutaraldehyde fixatives containing trinitro compounds. J. Cell Biol. 39:168a-l69a.
- Popp, J. A., and Monteiro-Riviere, N. A. 1985. Macroscopic, microscopic, and ultrastructural anatomy of the nasal cavity, rat In ILSI monographs on pathology of laboratory animals. Respiratory system, eds. T.C. Jones, U. Mohr, and R. D. Hunt, pp. 3-10. New York: Springer.

Statistics:

Differences between groups were considered casewise as statistically significant for p < .05. Body weight and data of lung weight and lung retention of particles were analyzed using analysis of variance. If the group means differed significantly by the analysis of variance, the means of the treatment groups were compared with the means of the control group, using Dunnett's modification of the t-test. For comparison of histopathological data, Fisher's exact test was used.
Survival data of the animals of the carcinogenicity study were analyzed by the Kaplan-Meier method (Kaplan & Meier, 1958)* using the Lifetest program (SAS Institute, Inc., 1985). For animals with significantly different survival times, the tumor rates were compared within three time periods (days 200-450, 450-600, >600) using the prevalence method of Hoel and Walburg (1972)*.

*References:
- Kaplan, E. L., and Meier, P. 1958. Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53: 457-481.
- Hoel, D. G., and Walburg, H. E. 1972. Statistical analysis of survival experiments. ]. Natl. Cancer Inst. 49:361-372.

Clinical signs:

not specified

Dermal irritation (if dermal study):

not specified

Mortality:

mortality observed, treatment-related

Description (incidence):

- mortality rate was 33% in the TiO2 group compared to 10% in the clean air control group 13.5 months after the start of exposure.
- mortality rate of 50% was reached 17 months from birth in the TiO2 group and 20 months in the control group.

Body weight and weight changes:

effects observed, treatment-related

Description (incidence and severity):

- after 8 months in the TiO2 group the body weight of the mice was significantly lower compared to the clean air control group.

Food consumption and compound intake (if feeding study):

not specified

Food efficiency:

not specified

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

not specified

Ophthalmological findings:

not specified

Haematological findings:

not specified

Clinical biochemistry findings:

not specified

Urinalysis findings:

not specified

Behaviour (functional findings):

not specified

Immunological findings:

not specified

Organ weight findings including organ / body weight ratios:

effects observed, treatment-related

Description (incidence and severity):

- the measurements after 3 and 12 months of exposure to TiO2 (0.3 g, 0.9 g) showed a substantial increase in lung wet weight compared to the controls (0.2 g, 0.2 g), processing with study duration.
- in the recovery phase, after 13.5 months of exposure, a slight decrease in lung wet weight was found in the TiO2 (0.7 g) exposed groups.

Gross pathological findings:

not specified

Neuropathological findings:

not specified

Histopathological findings: non-neoplastic:

not specified

Histopathological findings: neoplastic:

no effects observed

Other effects:

not specified

Details on results:

CLINICAL SIGNS/MORTALITY/BODY WEIGHT
- after 4 months of exposure the TiO2 particle concentrations were increased from 7.0 mg/m³ to 15 mg/m³. Because of some signs of toxicity (individual loss of body weight, bad general condition) and an increased mortality, the particle concentration was reduced after 4 months of exposure from 15 mg/m³ to 10 mg/m³.
- during the last months of exposure there was no significant difference in the body weight between the control and exposed group.

HISTOPATHOLOGY: NEOPLASTIC
- only the lung tumor types adenomas and adenocarcinomas were observed in mice
- percentages of adenomas/adenocarcinomas were 11.3%/2.5% for the TiO2 group and 25%/15.4% for the clean air group
- the lung tumor rates (adenomas and adenocarcinomas) of TiO2-exposed (13.8%) animals were not significantly different from the tumor rate of the control animals (30%).

PARTICLE LUNG BURDEN
- the particle lung burden of the mice found after 3, 6, and 12 months of exposure was 0.8, 2.5, 5.2 (TiO2) mg/lung. Expressed as milligrams particles per gram clean air control lung (wet weight of control lung 0.2 g), the particle lung loads after 1 year of exposure to TiO2 were 26 mg.

Relevance of carcinogenic effects / potential:

Overall, the only the lung tumor types found in the mice were adenomas and adenocarcinomas. The percentages of adenomas/adenocarcinomas were 11.3%/2.5% for the TiO2 group and 25%/15.4% for the clean air group. The lung tumor rates (adenomas and adenocarcinomas) of TiO2-exposed (13.8%) animals were not significantly different from the tumor rate of the control animals (30%).

Remarks on result:

other: Due to the anomalous study design (frequency, no dose response paradigm, out-dated criteria for tumour classification) an effect level cannot be derived.

Critical effects observed:

not specified

Conclusions:

This study by Heinrich et al. (single exposure concentration: 7.2 mg/m³ 1-4 months, 14.8 mg/m³ 5-8 months, 9.4 mg/m³ 9-13.5 months, 18h/d, 5d/w) was noted as a reliability 3 study because it was a satellite group used for another study. The study included only female mice and did not have a dose response paradigm. In addition, the mice were exposed for 18 hours/day, 5 days per week for 13.5 months. The tumour response of mice was not significantly different from controls. Moreover, the evaluation of tumours (including malignant tumours) was assessed and did not consider the two international lung pathology workshops – which have reassessed the criteria for describing malignant vs. benign tumours.
Consequently, the diagnosis should have changed after reconsideration of the revised criteria.

Reference 2

Endpoint:

carcinogenicity: inhalation

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

not specified

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

significant methodological deficiencies

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

reference to same study

Qualifier:

no guideline followed

Principles of method if other than guideline:

Female Wistar rats were exposed 18 hours/day, 5 days/week for 2 years to titanium dioxide (P25, Degussa; MMAD: 0.80 µm) and were subsequently kept in clean air for 6 months. The type of inhalation was whole body and the average particle exposure concentrations for the test substance was 10 mg/m³. A control group receiving clean air was run concurrently. The following observations were performed: clinical signs, mortality, body weight, lung wet weight, lung retention of inhaled particles, and histopathology.

GLP compliance:

not specified

Specific details on test material used for the study:

not applicable

Species:

rat

Strain:

Wistar

Details on species / strain selection:

not specified

Sex:

female

Details on test animals or test system and environmental conditions:

TEST ANIMALS - Crl:(WI)BR)
- Source: Charles River Wiga GmbH, Sulzfeld, Germany
- Age at study initiation: 7 weeks ± 3 days
- Housing: housed 2 per cage; kept in wire mesh cages during the exposure period, and in Makrolon cages (18.7 x 21 x 15 cm) during the subsequent clean air period. Softwood bedding (H3/4) was used.
- Diet (ad libitum): "1324 N spec. prepared" (Altromin, Lage, Germany)
- Water (ad libitum): drinking water

ENVIRONMENTAL CONDITIONS
- Temperature: 23 - 25 °C
- Relative humidity: 50 - 70%
- Photoperiod: 12/12

Route of administration:

inhalation: aerosol

Type of inhalation exposure (if applicable):

whole body

Vehicle:

clean air

Mass median aerodynamic diameter (MMAD):

0.8 µm

Geometric standard deviation (GSD):

1.8

Details on exposure:

GENERATION OF TEST ATMOSPHERE / CHAMBER DESCRIPTION
- Exposure apparatus: special whole-body exposure chambers (6 or 12 m³) of the horizontal flow type (Heinrich et al., 1985)*.
The rats were exposed together with NMRI mice in the same exposure chamber (for information on the mice please refer to Section 7.5.2 Repeated dose toxicity: inhalation: NANO_s_Heinrich_1995_mice).

- System of generating particulates/aerosols: aerosols of the test item were generated by a dry dispersion technique using a screw feeder and a pressurized air dispersion nozzle. The mass median aerodynamic diameter of the aerosol was about 1.5 µm. In order to increase the deposition efficiency of the test aerosol in the deep lung, the particle size distribution was shifted toward smaller particles in the submicrometer regime by removing the coarse particles using a cyclone (50% cut-off diameter ≈ 1 mm for a flow rate of 100 m³/h).

- Method of particle size determination: the mass median aerodynamic diameter (MMAD) and the geometric standard deviation of the particles in the exposure chambers were measured every month (n = 24) with a 10-stage Berner impactor (LPI 0.01525; range 15 nm to 16 µm).
The following MMAD was measured:
MMAD: 0.80 µm (geometric standard deviation: 1.80 µm)

*Reference:
- Heinrich, U., Muhle, H., Koch, W., and Mohr, U. 1985. Long-term inhalation studies with rodents. In Safety evaluation and regulation of chemicals 2, ed. F. Homburger, pp. 239 - 250. Basel: Karger.

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

The particle concentration in the exposure chamber was determined continuously, using aerosol photometers (Koch et al., 1986)*. For calibration of the photometer, the aerosol concentration in each exposure chamber was determined gravimetrically at weekly intervals (Heinrich et al., 1986)*.
The following mean particle mass exposure concentrations were measured:
7.2 mg titanium dioxide/m³ for the first 4 months, followed by 14.8 mg titanium dioxide/m³ for 4 months and 9.4 mg titanium dioxide/m³ for 16 months

To compare the different exposure concentrations, the cumulative particle exposure (g/m³ x h) was calculated by multiplying the mean particle mass exposure concentration by the actual exposure time per day and subsequently summarizing for the whole exposure period: 88.1 g/m³/h

*References:
- Koch, W. Lödding, H., Oenning, G., and Muhle, H. 1986. The generation and measurement of dry aerosols in large scale inhalation experiments., J. Aerosol Sci. 19: 1453 - 1457.
- Heinrich, U., Muhle, H., Takenaka, S., Ernst, H., Fuhst, R., Mohr, Z., Pott, F., and Stöber, W. 1986. Chronic effects on the respiratory tract of hamsters, mice and rats after long-term inhalation of high concentrations of filtered and unfiltered diesel engine emissions. J. Appl. Toxicol. 6: 383 - 395.

Duration of treatment / exposure:

24 months

Frequency of treatment:

18 hours/day, 5 days/week

Post exposure period:

Following the exposure period, the rats were removed from the inhalation chambers and kept under clean air conditions for an additional 6 months.

Dose / conc.:

10 mg/m³ air (analytical)

Remarks:

Standard deviation: 2.9

No. of animals per sex per dose:

Treatment group (total: 260 female rats):
- carcinogenicity: 100 rats
- histology (serial sacrifice): 80 rats
- DNA adducts (24 months): 14 rats (from Gallagher et al., 1994)*
- Particle mass/lung (serial sacrifice): 66 rats

Control group (total: 380 female rats):
- carcinogenicity: 220 rats
- histology (serial sacrifice): 80 rats
- DNA adducts (24 months): 14 rats (from Gallagher et al., 1994)*
- Particle mass/lung (serial sacrifice): 66 rats

*Reference:
- Gallagher, J., Heinrich, U., George, M., Hendee, l., Phillips, D.H., and Lewtas, J. 1994. Formation of DNA adducts in rat lung following chronic inhalation of diesel emissions, carbon black and titanium dioxide particles. Carcinogenesis 15(7): 1291-1299.

Control animals:

yes, concurrent vehicle

Details on study design:

not applicable

Positive control:

not specified

Observations and examinations performed and frequency:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: once daily

DETAILED CLINICAL OBSERVATIONS: No data
DERMAL IRRITATION: No data

BODY WEIGHT: Yes
- Time schedule for examinations: every fourth week

FOOD CONSUMPTION AND COMPOUND INTAKE:
- Food consumption for each animal determined and mean daily diet consumption calculated as g food/kg body weight/day: No data
- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: No data

FOOD EFFICIENCY:
- Body weight gain in kg/food consumption in kg per unit time X 100 calculated as time-weighted averages from the consumption and body weight gain data: No data

WATER CONSUMPTION AND COMPOUND INTAKE: No data
OPHTHALMOSCOPIC EXAMINATION: No data
HAEMATOLOGY: No data
CLINICAL CHEMISTRY: No data
URINALYSIS: No data
NEUROBEHAVIOURAL EXAMINATION: No data

Sacrifice and pathology:

GROSS PATHOLOGY / HISTOPATHOLOGY: Yes
Necropsies of dead or moribund animals were done 7 days/week.
The lung wet weight and lung retention of inhaled particles were determined at different time points during the study. The lung wet weight was determined 3, 6, 12, 18, 22, and 24 months after starting the exposure. Titanium dioxide samples were determined by atomic absorption spectroscopy after ashing the lungs.

For the histopathological investigations, the organs of scheduled or moribund sacrifices were fixed in 10% neutral buffered formalin or lto­ Karnovsky fixative (Ito & Karnovsky, 1968)*. The tissues were embedded in Paramat-Wax, sectioned at 5 µm, and stained with hematoxylin and eosin (Lilly-Meyer). Histopathological investigations of the following organs were conducted for all animals: nasal and paranasal cavities (four sections; localization according to Popp & Monteiro-Riviere, 1985*), larynx, trachea, and lung (five sections; localization: the left lobe, right caudal lobe, and right middle lobe were sectioned longitudinally, and the right cranial lobe and accessory lobe were sectioned transversely to main bronchus). Graduation of the findings was done with four grades: very slight, slight, moderate, and high.

Tumors were classified according to the International Classification of Rodent Tumours (IARC. 1992)*.

*References:
- Ito, S., and Karnovsky, M. J. 1968. Formaldehyde-glutaraldehyde fixatives containing trinitro compounds. J. Cell Biol. 39:168a-l69a.
- Popp, J. A., and Monteiro-Riviere, N. A. 1985. Macroscopic, microscopic, and ultrastructural anatomy of the nasal cavity, rat In ILSI monographs on pathology of laboratory animals. Respiratory system, eds. T.C. Jones, U. Mohr, and R. D. Hunt, pp. 3-10. New York: Springer.
- IARC. 1992. International classification of rodent tumours. Part I: The rat. IARC Sci. Publ. no. 122. Lyon: International Agency for Research on Cancer.

Statistics:

Differences between groups were considered casewise as statistically significant for p < .05. Body weight, data of lung weight and lung retention of particles were analyzed using analysis of variance. If the group means differed significantly by the analysis of variance, the means of the treatment groups were compared with the means of the control group, using Dunnett's modification of the t-test. For comparison of histopathological data, Fisher's exact test was used.
Survival data of the animals of the carcinogenicity study were analyzed by the Kaplan-Meier method (Kaplan & Meier, 1958)* using the Lifetest program (SAS Institute, Inc., 1985). For animals with significantly different survival times, the tumor rates were compared within three time periods (days 400-700, 700-800, >800) using the prevalence method of Hoel and Walburg (1972)*.

*References:
- Kaplan, E. L., and Meier, P. 1958. Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53: 457-481.
- Hoel, D. G., and Walburg, H. E. 1972. Statistical analysis of survival experiments. ]. Natl. Cancer Inst. 49:361-372.

Clinical signs:

not specified

Dermal irritation (if dermal study):

not specified

Mortality:

mortality observed, treatment-related

Description (incidence):

- after 24 months of exposure, the mortality found was 60% in the TiO2 group compared to 42% in the clean air control group.
- at the end of the 130 week experimental time (exposure time and clean air period), the mortality reached 90% in the TiO2 group and 85% in the control group.
- compared to the controls, the mean lifetime of the rats exposed to TiO2 was significantly shortened.

Body weight and weight changes:

effects observed, treatment-related

Description (incidence and severity):

- body weight of the exposed animals was significantly lower from day 400 (TiO2) compared to control.
- at the end of the 2-year exposure, the body weight of the animals exposed to TiO2 (body weight: 365 g) was significantly lower compared to the control rats (body weight: 417 g).

Food consumption and compound intake (if feeding study):

not specified

Food efficiency:

not specified

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

not specified

Ophthalmological findings:

not specified

Haematological findings:

not specified

Clinical biochemistry findings:

not specified

Urinalysis findings:

not specified

Behaviour (functional findings):

not specified

Immunological findings:

not specified

Organ weight findings including organ / body weight ratios:

effects observed, treatment-related

Description (incidence and severity):

- exposure to TiO2 led to a substantial increase in lung wet weight, processing with study duration.

Gross pathological findings:

not specified

Neuropathological findings:

not specified

Histopathological findings: non-neoplastic:

effects observed, treatment-related

Description (incidence and severity):

- moderate to high grade bronchioloalveolar hyperplasia was observed in the TiO2 (99/100 rats) group.
- very slight to slight interstitial fibrosis in the lungs was found after 6 months of exposure.
- slight to moderate interstitial fibrosis in the lungs was observed in all animals exposed for 2 years.
- particle-laden macrophages and particles in the alveolar region were also observed in the lungs of all exposed rats.

Histopathological findings: neoplastic:

effects observed, treatment-related

Description (incidence and severity):

- lung tumors were found in serial sacrificed animals after 18 months of exposure to TiO2 (5/20 rats; p≤0.05). The following tumor were observed:
benign keratinizing cystic squamous-cell tumor: 2/20 rats
squamous-cell carcinoma: 3/20 rats (sometimes together with adenocarcinoma and benign keratinizing cystic squamous-cell tumor)
adenocarcinoma: 2/20 rats
- lung tumors were found in serial sacrificed animals after 24 months of exposure to TiO2 (4/9 rats; p≤0.05). The following tumor were observed:
benign keratinizing cystic squamous-cell tumor: 2/9 rats
squamous-cell carcinoma: 2/99 rats (sometimes together with benign squamous-cell tumor)
adenocarcinoma: 1/9 rats

- after an exposure time of 24 months followed by 6 months of clean air, lung tumor rates of 32% were observed in rats exposed to TiO2. 8 animals showed 2 tumors in their lungs.
- The following tumour types were observed after an experimental time of 30 months (24 months TiO2-exposure plus 6 months clean air):
benign keratinizing cystic squamous-cell tumor: 20/100 rats
squamous-cell carcinoma: 3/100 rats
adenoma: 4/100 rats
adenocarcinoma: 13/100 rats
Number of rats with tumors: 32/100 (19/100 rats: count without benign keratinizing cystic squamous-cell tumors)
1/217 control animal (clean air exposure) showed adenocarcinoma

- lung tumor rate increased with increasing particle exposure concentration.
- lung tumor incidences of the TiO2 exposed group was significantly increased compared to the control group.

Other effects:

not specified

Details on results:

MORTALITY:
- the various exposure groups did not differ significantly in their mean lifetime among themselves.

HISTOPATHOLOGY: NEOPLASTIC
- no lung tumors were observed in the TiO2 satellite group of 20 animals each after 6 and 12 months of exposure.

PARTICLE LUNG BURDEN
- during the second year of exposure, the particle lung load of the TiO2 exposed animals increased by only 13%.
- the retained particle mass in the lung-associated lymph nodes (LALN) of the TiO2-exposed rats after 22 months amounted to about 14%.
- expressed as milligrams particles per gram clean air control lung (wet weight of control lung 1.2 g), the particle lung loads after 1 year of exposure to TiO2 were 29 mg.

Relevance of carcinogenic effects / potential:

According to the authors following findings were made:
No lung tumors were observed in the TiO2 satellite group of 20 animals each after 6 and 12 months of exposure. On the other hand lung tumors were found in serial sacrificed animals after 18 months of exposure to TiO2 (5/20 rats; p≤0.05). Also, lung tumors were found in serial sacrificed animals after 24 months of exposure to TiO2 (4/9 rats; p≤0.05). Furthermore, after an exposure time of 24 months followed by 6 months of clean air, lung tumor rates of 32% were observed in rats exposed to TiO2. 8 animals showed 2 tumors in their lungs.

Overall, lung tumor rate increased with increasing particle exposure concentration and lung tumor incidences of the TiO2 exposed group was significantly increased compared to the control group.

Remarks on result:

other:

Remarks:

Due to the anomalous study design (frequency, no dose response paradigm, out-dated criteria for tumour classification) an effect level cannot be derived.

Critical effects observed:

not specified

Conclusions:

The Heinrich et al (1995) study was reported as a lower reliability study. Rats were exposed for 18 hours per day. The characterisation of tumours preceded the re-evaluation of rat tumour lung pathology workshops and would need to be reconsidered before determining whether these tumours were malignant. Heinrich (rated R3 in the CLH report) reported that 32/100 rats developed lung tumours after exposures to ultrafine TiO2. These included benign squamous tumours, 3 squamous cell carcinomas, adenomas and 13 adenocarcinomas. However, the publication of this report preceded the revised criteria based on 2 workshops for classification of cystic keratinizing squamous lesion of the rat lung. (Carlton, 1994; Levy 1994 and Boorman et al., 1996). As a consequence, the determination that these are malignant tumours should have been properly re-evaluated according to the new criteria.

Reference 3

Endpoint:

carcinogenicity: inhalation

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

1979-06-04 to 1981-07-09

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

significant methodological deficiencies

Qualifier:

no guideline followed

Principles of method if other than guideline:

Groups of 100 male and 100 female Crl:CD(SD)BR rats each were exposed to titanium dioxide (10, 50, and 250 mg/m³). The test item was administrated via whole body inhalation for 6 hours/day, 5 days/week for 24 months. A concurrent control group was run concurrently. The following parameters were assessed:clinical signs, mortality, body weights, haematology, clinical chemistry, urinalysis. gross pathology, and histopathology.

GLP compliance:

no

Specific details on test material used for the study:

not specified

Species:

rat

Strain:

other: Crl:CD(SD)BR

Details on species / strain selection:

Selection of the CD rat was based on extensive experience with the strain and its suitability relative to longevity, hardiness, sensitivity and low incidence of spontaneous disease.

Sex:

male/female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River Breeding Laboratories, Wilmington, Mass.
- Age: 3 weeks
- Housing: housed pairwise in stainless steel wire-mesh cages.
- Diet (ad libitum): Purina Laboratory Chow Checkers '5001
- Water: ad libitum
- Acclimation period: approx. 17 days

ENVIRONMENTAL CONDITIONS
- Temperature: 23 ± 2 °C
- Relative humidity: 50 ± 10 %
- Photoperiod (hrs dark / hrs light): 12/12
- rooms had laminar flows of filtered and recirculated air

Route of administration:

inhalation

Type of inhalation exposure (if applicable):

whole body

Vehicle:

air

Mass median aerodynamic diameter (MMAD):

>= 1.54 - <= 1.93 µm

Remarks on MMAD:

GSD: 2.52 - 3.14
Mean respirable fraction of TiO2: 93.7% or greater (values for each determination ranged from 84.0 to 99.95 %)

Details on exposure:

GENERATION OF TEST ATMOSPHERE / CHAMBER DESCRIPTION
- Exposure apparatus: inhalation chambers made of material not reactive with TiO2
- Chamber volume: 3.85 m³
- System of generating particulates/aerosols: atmospheres of TiO2 were generated by metering the dust into an apparatus containing a vertical elutriator connected in series to a settling chamber. An ACCU-RATE, Model 502, variable-speed screw-feeder was used to meter TiO2 dust into a Plexiglas sample-delivery tube attached perpendicularly to the vertical axis of the elutriator. The dust was dispersed by an air jet directed along the sample delivery tube axis and passed into the elutriator. Initial settling of the heavier nonrespirable dust particle took place in the elutriator; the lighter particles passed into the settling chamber from which the respirable particles were diverted into the exposure chamber. Chamber concentrations were maintained by controlling the TiO2 delivery rate into the generation apparatus and by diluting the dust particle stream as it entered the chamber.
- Temperature and humidity: temperature and relative humidity of the exposure chambers were targeted at 23 ± 2°C and 50 ± 10%, respectively. These were measured at least once daily.
- Air flow rate: >800 L/min

TEST ATMOSPHERE
- Particle size distribution: aerodynamic particle sizing was performed for at least seven exposures in each chamber over the course of the study. Two types of in-stack cascade impactors were used for these determinations: Monsanto 5-Stage Impactor with Cyclone Preseparator and Sierra, Model 210, 8-Stage Impactor with Cyclone Preseparator. The mass median aerodynamic diameter, geometric standard deviation and the fraction of respirable particles were determined graphically. Those particles with a MMD of 10 µm or less were considered respirable.

TEST ATMOSPHERE
- Brief description of analytical method used: chamber concentrations were determined gravimetrically. Approximately every half hour and from each exposure chamber, a predetermined volume of chamber atmosphere was drawn through a preweighed Gellman, Type-A/E, glass-fibre filter, 47 mm diameter. Each chamber concentration was calculated from the net weight of TiO2 collected on the filter. The mean daily chamber concentrations were calculated as the time-weighted averages (TWA) over each 6-hour exposure period.

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

see above ("Details on inhalation exposure")

Duration of treatment / exposure:

24 months

Frequency of treatment:

6 hours/day, 5 days/week

Post exposure period:

none

Dose / conc.:

10.55 mg/m³ air (analytical)

Remarks:

SD: 2.12 mg/m³

Dose / conc.:

50.68 mg/m³ air (analytical)

Remarks:

SD: 6.65 mg/m³

Dose / conc.:

250.1 mg/L air (analytical)

Remarks:

SD: 24.70 mg/m³

No. of animals per sex per dose:

100 male rats / 100 female rats

Control animals:

yes, concurrent vehicle

Details on study design:

- Dose selection rationale: results from previous inhalation studies.
- Rationale for animal assignment: rats of each sex were divided by computerised, stratified randomisation into groups of 100 males and groups of 100 females such that the mean of body weights of each group of rats within a sex were approximately equal.

Positive control:

not specified

Observations and examinations performed and frequency:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: at least twice daily throughout the study
- Cage side observations checked: morbundity/mortality, abnormal behaviour and appearance

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: at least once weekly during the first 3 months and at least every other week during the remainder of the study

BODY WEIGHT: Yes
- Time schedule for examinations: once weekly during the first 3 months of the study followed by approx. once every other week for the remainder of the study.

FOOD CONSUMPTION AND COMPOUND INTAKE : No
FOOD EFFICIENCY: No
WATER CONSUMPTION AND COMPOUND INTAKE: No
OPHTHALMOSCOPIC EXAMINATION: No

HAEMATOLOGY: Yes
- Time schedule for collection of blood: approx. 3, 6, 12, 15 and 18 months after the study's initiation
- How many animals: 10 males / 10 females (same rats were evaluated at each interval throughout the study)
- Parameters checked: basophil count, eosinophil count, erythrocyte count, haematocrit (Ht), haemoglobin, leukocyte count, lymphocyte count, mean cell haemoglobin, mean cell volume, mean corpuscular haemoglobin concentration, monocyte count and neutrophil count.

CLINICAL CHEMISTRY: Yes
- Time schedule for collection of blood: approx. 3, 6, 12, 15 and 18 months after the study's initiation
- How many animals: 10 males / 10 females (same rats were evaluated at each interval throughout the study)
- Parameters checked: alanine aminotransferase activity, alkaline phosphatase activity, bilirubin, calcium, phosphorus, total protein and urea nitrogen.

URINALYSIS: Yes
- Time schedule for collection of urine: approx. 3, 6, 12, 15 and 18 months after the study's initiation.
- Parameters checked: volume, osmolality, pH, bilirubin, blood, protein, sugar, urobilinogen, appearance and sediment

NEUROBEHAVIOURAL EXAMINATION: No

Sacrifice and pathology:

GROSS PATHOLOGY: Yes
HISTOPATHOLOGY: Yes

Gross and histopathological examinations were conducted on 5 rats/sex/treatment group after 3 and 6 months exposure as well as on 10 rats/sex/treatment group after 12 months exposure and on all rats alive after 24 months exposure.
Rats which had been designated for clinical chemical evaluation were not included among those selected for the interim sacrifices.
All rats found dead or sacrificed in extremis (integrity of tissue permitting), were examined grossly and histopathologically.

All rats were sacrificed and the respiratory tract was prepared for fixation.
All other tissues were removed for gross examination and weighed (lungs, trachea, heart, liver, stomach (3 and 12 months sacrifice only), kidneys, spleen, testes, pituitary, brain, thymus and adrenals) at each sacrifice. The tissues were fixed in either Bouin's (nasal cavitiy (turbinates), trachea, luings, oesophagus, kidneys, sternal bone marrow, testes, epididymides, mammary gland,pituitary, thyroid - parathyroids, adrenals, eyes, bone (sternum, femur, and vertebrate, ear (zymbal gland), and skin (neck)) or formalin at the 6- through 24-month sacrifices. At the 3-month sacrifice only, adipose tissue and ovaries were fixed in Bouin's fixative and the adrenals, oesophagus, mammary and Zymbal's glands were fixed in formalin.

Representative specimens of the following organs and tissues were taken from all rats: heart, thoracic aorta, nasal cavity (turbinates), trachea, lungs, liver, pancreas, small intestine (duodenum, jejunum, and ileum), tongue, oesophagus, stomach, salivary glands, large intestine (cecum and colon), rectum, kidneys, bladder, sternal bone marrow, spleen, lymph nodes (cervical, mesenteric and tracheobronchial), thymus, testes, epididymes, prostate, seminal vesicles, uterus, ovaries, mammary gland, pituitary, adrenals, thyroid - parathyroids, brain, spinal cord, sciatic nerve, skeletal muscle, bone (sternum, femur, and vertebrate), eyes, Harderian's gland, exorbital lacrimal glands, ear (Zymbal's gland), skin (neck), adipose tissue, and all gross lesions

.

Other examinations:

not specified

Statistics:

Body weight and weight gain data were evaluated with a one-way analysis of variance and the least significant difference test. Organ weight and particle size data were evaluated with a one-way analysis of variance with pairwise comparison being made with the LSD and/or Dunnett's tests, and a test for linear trend. Clinical laboratory data were evaluated by a partially nested and crossed analysis of variance and by the LSD test. Outliners within the clinical laboratory and organ weight data were evaluated and excluded from calculations of the means by using the Dixon Criterion. The Bartlett test was also used to evaluate the organ weight data.
Significance was judged at the 5 % level of probability.

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):

not examined

Food efficiency:

not examined

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

not examined

Ophthalmological findings:

not examined

Haematological findings:

effects observed, treatment-related

Description (incidence and severity):

There were several changes in haematological parameters related to exposure of rats to TiO2. When compared to the controls over course of this study:
- haematocrits (both sexes) and heamoglobins (males only) for rats in the 250 mg/m³ treatment groups were greater (up to 12 and 11%, respectively).
- mean cell volumes and mean cell haemoglobins in male rats from all exposed groups were greater (up to 14%); there were only minimal differences among these groups.
- neutrophil counts in all treatment groups were greater (up to 80% in male and 117% in female rats in the 250 mg/m³ treatment groups). A dose-related trend was observed among these treatment groups.
- lymphocyte counts in all treatment groups were lower (up to 45% lower for rats in the 250 mg/m³ treatment groups). A dose-related trend was observed among these treatment groups.

Clinical biochemistry findings:

effects observed, treatment-related

Description (incidence and severity):

There were several statistically significant changes in clinical chemical parameters measured in the serum from rats exposed to TiO2. When compared to the controls over the course of the study:
- bilirubin content in female rats were greater and were more pronounced in the 50 and 250 mg/m³ treatment groups at the 18-months evaluation (90 and 117%, respectively).
- calcium concentrations in all treatment groups were generally lower (up to 7% less).
- phosphorus concentrations were lower in male (up to 36% lower) and higher in female rats (up to 61%). These effects were consistent in males throughout the study but observed in females at only the 3-, 15- and 18-month evaluations.

Urinalysis findings:

no effects observed

Behaviour (functional findings):

not examined

Immunological findings:

not examined

Organ weight findings including organ / body weight ratios:

effects observed, treatment-related

Description (incidence and severity):

Exposure of rats to TiO2 resulted in changes in mean absolute and/or relative weights of several organs over the course of the study. The organs affected were as follows:
Lungs:
- clear dose- and time-dependent lung weight increases were observed
- mean absolute and relative lung weights for rats in the 250 mg/m³ treatment group were significantly greater than controls throughout the study. The range was from approximately 1.52- to 2.59-fold and from 1.53- to 3.38-fold greater for male and female rats, respectively.
- mean absolute and relative lung weights for rats in the 50 mg/m³ treatment group were also greater than controls throughout the study. With exception of the mean absolute male lung weights at six months, these effects were significant at the 6-months and subsequent sacrifices. The lung weights ranged from approx. 1.2- to 1.4-fold and 1.46- to 1.7-fold greater for male and female rats, respectively.
markedly heavier at 50 mg/m³, and were more than two times control lung weights at 250mg/m³.

Thymus:
- A dose-related thymus weight effect was observed in rats in the 50 and 250 mg/m³ treatment groups. Mean absolute and relative thymus weights were as high as 1.37-fold greater than controls.

Gross pathological findings:

effects observed, treatment-related

Description (incidence and severity):

During the gross pathological examinations, TiO2 deposits were observed on skin and the mucosa of the nasal cavity, trachea, bronchus and gastrointestinal tract of rats exposed to this compound. The pleural surfaces of the lungs contained scattered white foci which were present in greater numbers and larger sizes in rats exposed to the higher TiO2 concentrations. Subpleural cholesterol granulomas appeared on the lungs of rats in the 50 and 250 mg/m³ treatment groups as slightly elevated gray nodules. The lungs of rats in the 250 mg/m³ treatment groups were white in appearance, voluminous, of rubbery consistency and failed to collapse upon opening the chest cavity at necropsy.
The tracheabronchial lymph nodes were markedly swollen and appeared as chalky masses in all exposure groups. Most of these gross observations were apparent at six months with the severity and frequency of occurrence increasing over time.

Neuropathological findings:

no effects observed

Histopathological findings: non-neoplastic:

effects observed, treatment-related

Description (incidence and severity):

The respiratory tract was the primary deposition site for TiO2 as well as the primary site in which a tissue response to this compound was observed. Over the course of this study, the following dose-related effects were observed:
- increased incidence of rhinitis with accompanying squamous metaplasia in the anterior nasal cavity;
- greater numbers of TiO2 –laden alveolar macrophages (dust cells), and of aggregated, foamy alveolar macrophages;
- deposits of TiO2 particles within the lymph nodes where the relative severity for tracheobronchial > peribronchial and perivascular > mesenteric lymph nodes.
- alveolar proteinosis in approximately 51 and 97% of the rats in the 50 and 250 mg/m³ treatment groups, respectively, but it was not found in all other groups;
- markedly increased incidences of cholesterol granulomas that contained large numbers of collagen fibers in the lungs of rats in the 50 and 250 mg/m³ treatment groups (approximately 75 and 97% of these rats, respectively; 13% or less for rats in all other groups);
- markedly increased incidences of collagenized fibrosis within alveoli of rats in the 50 and 250 mg/m³ treatment groups (in approximately 60 and 99% of these rats, respectively; 14% less for rats in all other groups);
- greater incidences of focal pleurisy associated with subpleural cholesterol granulomas and focal dust cell infiltration in the lungs of TiO2-exposed rats;
- bronchiolarization of alveoli adjacent to terminal bronchioles, which occurred more frequently in female than male rats

Histopathological findings: neoplastic:

effects observed, treatment-related

Other effects:

not examined

Details on results:

CLINICAL SIGNS
Exposure to TiO2 resulted in no abnormal clinical sign in any exposed group.
The following observations were associated with the TiO2 exposures:
- coloured dischanges around the eyes and nose were observed approximately 1.4- to 8-fold less frequently among TiO2-exposed rats than the controls.
- irregular respiration and abnormal lung noise were observed with a greater incidence and earlier in the study among the TiO2-exposed rats than the controls. The incidence was greater for TiO2-exposed females than males. A clear dose-response relationship was not observed.
- stained and/or wet perineum was observed with a greater incidence among female rats exposed to TiO2. At the higher TiO2 concentrations, the incidence was greater and the observation was made earlier in the study than at the lower concentrations.

MORTALITY
- exposure to TiO2 resulted in no excess mortality in any exposed group.
- mortality rates for all groups within a sex were similar. The mortality rates for females were 1.2- to 2-fold higher than those for male rats.

BODY WEIGHT AND WEIGHT CHANGES:
Mean body weights of all TiO2-exposed groups were generally lower that those of the controls. Although there was no dose-related trend for these body weigh effects, the mean weights for male and female rats exposed to 250 mg/m³ was consistently lower than those for rats exposed to 10 or 50 mg/m³. The differences in the mean body weight gain among all treatment groups parallelled the mean body weight effects.

HAEMATOLOGY:
- erythrocyte counts for rats in the 250 mg/m³ treatment groups were slightly greater (up to 9%; not statistically significant).
- leukocyte counts in all treatment groups were greater (up to 80% for the high-dose group). The values for rats in the 250 mg/m³ treatment group were generally higher than those of the other treatment groups.

URINALYSIS
Urological parameters were within the expected range of biological variability for this species.

ORGAN WEIGHTS
Lungs:
- lung weights at 10mg/m³ were comparable to those of control group

Liver:
- mean absolute and/or relative liver weights of TiO2-exposed male and female rats were generally lower than those of controls over the course of the study.
- these differences ranged from approximately 100 to 66% of the control weights. A clear dose-response was not observed.

Kidney:
- mean absolute and/or relative kidney weights for TiO2 exposed male rats were approx. 77 to 88% of the controls at the 3-, 12- and 24-months sacrifices. A clear dose-response was not demonstrated.

Significant but transient effects were observed for other organ weights over the course of this study. These effects were observed for other organ weights over the course of this study. These effects were either related to body weight differences or were not considered to be biologically significant.

HISTOPATHOLOGY-NEOPLASTIC
Exposures to titanium dioxide at concentrations of 10, 50, or 250 mg/m³ produced lung tumors (bronchioalveolar adenoma and squamous cell carcinoma) only at the highest concentration.

To further characterise the bronchoalveolar lesions, in 1992, a group of pathologists from North America and Europe examined lung lesions produced by para-aramid RFP (respirable fibre-shaped particles) and titanium dioxide. This panel diagnosed the lesion as a "proliferative keratin cyst" (PKC). Additionally, the pathologists agreed that the lesion was not a malignant neoplasm and is most likely not neoplastic. A minority opinion was that the lesion is probably a benign tumour (Carlton 1994; Levy 1994). Another subsequent international pathology workshop was convened to develop standardized histological criteria for classifying pulmonary keratin lesions (Boorman et al., 1996). As a consequence, most of the lesions that had originally been diagnosed as “cystic keratinizing squamous cell carcinomas” were re-classified by the consensus panels as non-tumorous “proliferative keratin cysts” (Warheit and Frame, 2006).
In the aftermath of two international pathology workshops designed, in part, to establish histological criteria for classifying pulmonary keratin lesions, these lesions were evaluated by four pathologists using current diagnostic criteria. Microscopic review of 16 proliferative squamous lesions, previously diagnosed as cystic keratinizing squamous cell carcinoma in the lungs of rats from the 2-year inhalation study was performed. Unanimous agreement was reached as to the diagnosis of each of the lesions. Two of the lesions were diagnosed as squamous metaplasia and 1 as poorly-keratinizing squamous cell carcinoma. Most of the remaining 13 lesions were diagnosed as non-neoplastic pulmonary keratin cysts (Warheit and Frame, 2006).
Consequently, the diagnosis of many of these lesions has changed after the development of revised criteria.
Please also refer for background information on the re-evaluation of the neoplastic findings to the field "Attached background material" below.

There were several effects associated with exposure to TiO2 which were not clearly dose- related. These include:
- hyperplasia of the Type II pneumocytes lining alveolar walls;
- increased incidences of chronic tracheitis; and
- Increased incidences of broncho/bronchiolar pneumonia.
Over the course of the study, most of the inhaled TiO2 particles that were observed within the alveoli were phagocytized by dust cells. These particles were retained primarily within the alveolar ducts and adjoining alveoli of rats in the 10 and 50 mg/m³ treatment groups which also had some dust-free alveoli in the peripheral acini. However, for rats in the 250 mg/m³ treatment group, the number of dust cells was markedly greater than that for rats in other treatment groups. The dust cells were distributed throughout the alveoli with larger concentrations in the alveolar duct region. This region served as a focal point for several of the tissue responses previously listed, e.g. collagenized fibrosis, bronchiolarization, squamous cell metaplasia and tumor formation. Within each treatment group, the lesions gradually increased in frequency and/or severity with time on test. By the end of this study, rats in the 50 and 250 mg/m³ treatment groups had irreversible effects whereas all of the lesions observed in rats exposed to 10 mg/m³ were considered to be reversible.
Although most of the biological effects associated with TiO2 exposure occurred within the respiratory tract, other tissues were also affected. The incidences of degenerative retinopathy in exposed rats was increased significantly compared to those of controls and were greater among female than male rats; however, a dose-response relationship was not observed. Female rats exposed to TiO2 also had significantly greater incidences of endometritis than the controls (dose-related).
Exposure of rats to TiO2 resulted in a dose-related transmigration of dust particles from the lung through the lymhatics to the lymph nodes, liver, and spleen. The accumulation of TiO2 in the tracheobronchial, cervical, and mesenteric lymph nodes did not result in a significant tissue response. In the spleen, dust particles aggregates were primarily located in the lymphnode tissue of the white pulp. In the liver, TiO2 accumulated in the portal riads and Kupfter cells. There were no cellular responses to these deposits in either the spleen or liver.

A wide variety of spontaneous neoplastic and non-neoplastic lesions that were considered to be age related occurred with similar incidence and severity among control and all TiO2-exposed groups

REFERENCES
- Carlton, W.W. (1994) "Proliferative Keratin Cyst," a Lesion in the Lungs of Rats Following Chronic Exposure to Para-aramid Fibrils. Fundam. Appl. Toxicol. 23, 304-307
- Levy L.S. (1994). Squamous Lung Lesions Associated with Chronic Exposure by Inhalation of Rats to p-Aramid Fibrils (Fine Fiber Dust) and to Titanium Dioxide: Findings of a Pathology Workshop. In: Mohr, U. (Ed.): Toxic and carcinogenic effects of solid particles in the respiratory tract, ILSI Press, 473-478
Boorman, G.A: et al. (1996). Classification of Cystic Keratinizing Squamous Lesions of the Rat Lung: report of a Workshop. Toxicol. Pathol. 24, 564-573

Relevance of carcinogenic effects / potential:

Please refer to the field "Details on results" above.

Dose descriptor:

NOEC

Effect level:

50.68 mg/m³ air (analytical)

Based on:

test mat.

Sex:

male/female

Basis for effect level:

histopathology: neoplastic

Dose descriptor:

LOAEC

Effect level:

250 mg/m³ air (analytical)

Based on:

test mat.

Sex:

female

Basis for effect level:

histopathology: neoplastic

Remarks on result:

other:

Remarks:

As demonstrated by the accumulating lung burden and the extended clearance time, the highest concentration used (250mg/m³) is clearly exceeding the maximum tolerated dose.

Critical effects observed:

not specified

Conclusions:

NOEC (tumourogenicity; rats): 50.68 mg/m³ (analytical conc.)

The findings of lung tumours at 250 mg/m³ described in the Lee et al. (exposure concentrations used: 0; 10; 50; 250 mg/m³, 6 hours/day, 5 days/week) study clearly exceeded the maximum tolerable dose (MTD) and therefore it would be inappropriate to be considered as a positive tumour response.
According to the NIOSH Executive Summary of their Current Intelligence bulletin (NIOSH 2011, pages VI-VII)*, the 250 mg/m³ concentration in the Lee et al., 1985 study was an excessive dose and is not relevant for human risk assessment: ”However, exposure concentrations greater than 100 mg/m³ are generally not considered acceptable inhalation toxicology practice today. Consequently, in a weight-of-evidence analysis, NIOSH questions the relevance of the 250 mg/m³ dose for classifying exposure to TiO2 as a carcinogenic hazard to workers and therefore, concludes that there are insufficient data at this time to classify tine TiO2 as a potential occupational carcinogen.”

Lee et al (1985) noted that, due to excessive loading in the lungs of rats exposed chronically at 250 mg/m³, the lung tumours were different from common human lung cancers in terms of tumour type, anatomic location, tumorigenesis and were devoid of tumour metastasis. Therefore, the biological relevance of these lung tumours were negligible.

Lung burdens of 118 and 130 mg/lung of TiO2 in male and female rats exposed to 50 mg/m³ TiO2 for 2 years did not result in lung tumours. In their Current Intelligence Bulletin document, NIOSH (2011)* questioned/dismissed the relevance of the results following the exposures to 250 mg/m³ of the 2-year study. Moreover, Lee et al. commented that the relevance of these rat tumours for humans was negligible, due to anatomic type, location, etc. Moreover, at the 50 mg/m³ exposure levels, there were no tumours in either male or female rats and this represented 118 mg/lung in the males and 130 mg/lung in the females. Therefore this study should not be considered a positive study for carcinogenicity, but instead a negative study for carcinogenicity in rats.

The dosimetric analysis of the 2-year inhalation study by Lee et al. shows that all three TiO2 exposure concentrations resulted in significant lung particle overload, i.e., an impaired alveolar macrophage-mediated particle clearance function. Per g lung weight, the retained normalized lung burden observed in the study of 28 mg/g exposed lung and 39 mg/g control lung for the 50 mg/m3-exposed rats (shown at the beginning of this analysis, p.7) is obviously greatly exceeding a retained lung burden of 1 mg/g lung which - according to Morrow (1988) - signals the beginning of lung overload in rats.This result also shows that despite a significantly PSP overloaded lung in this 2- year inhalation study lung tumors were only induced at the most excessive lung overload resulting from 250
mg/m3 exposure which is beyond any relevant realistic exposure scenario. While such excessively overloaded lung – as expected – resulted in the induction of lung tumors in the rat, the result also demonstrates that overload conditions below that highly unrealistic excessive level did not induce lung tumors in rats after chronic 2-year inhalation exposure.

For the evaluation of that study in the context of a carcinogenicity classification, its adequacy for this purpose needs to be evaluated. Adequacy defines the usefulness of information for the purpose of hazard identification and risk characterisation; in other words whether the available information allows clear decision-making about whether the substance meets the criteria for classification. Since the adverse findings in rats were obtained at doses exceeding the maximum tolerated dose, the adequacy of that study for classification purposes is explicitly questioned – to demonstrate this, the study design of Lee et al. was checked against the relevant OECD and ECHA guidance documents:
1) The guideline OECD 451 for the conduct of carcinogenicity studies, in conjunction with the relevant OECD guidance document 116, highlights on various occasions that inhalation concentrations overwhelming physiological mechanisms are in exceedance of the MTD:

a) The guidance explicitly states that “inhalation of doses that overwhelm pulmonary clearance may lead to tissue responses that are specific to the species being tested” (section 94, p.54).
b) “The robustness of a carcinogenicity or chronic toxicity study, in particular the former, is dependent on a demonstration that the dose levels selected in the study are adequate to show an effect or effects of the test substance, without producing either false negative results (because the doses selected were too low) or false positive results (because metabolic/homeostatic mechanisms are overwhelmed, etc.), which may be problematic in assessing risk in humans” (section 101, page 55).
c) In the selection of the maximum concentration, it should be considered that “disturbances of physiology or homeostasis that would compromise the validity of the study should be considered in the dose-selection process. Examples include hypotension, inhibition of blood clotting, overwhelming normal pulmonary clearance mechanisms, immune system effects, and in some cases hormonal imbalance” (p.63).
d) “For substances likely to accumulate in the lung over time due to poor solubility or other properties, the degree of lung-overload and delay in clearance needs to be estimated based on adequately designed pre-studies; ideally a 90-day study with post-exposure periods long enough to encompass at least one elimination half-time. The use of concentrations exceeding an elimination half-time of approximately 1 year due to lung-overload at the end of study is discouraged” (section 135, p.71).

2) The ECHA guidance on the Application of the CLP Criteria (Version 4.1, June 2015) highlights that “Tumours occurring only at excessive doses associated with severe toxicity generally have a more doubtful potential for carcinogenicity in humans. In addition, tumours occurring only at sites of contact and/or only at excessive doses need to be carefully evaluated for human relevance for carcinogenic hazard” (section 3.6.2.3.2., p.379-380).

The highest concentration used in the Lee et al. study clearly exceeded the MTD, since lung overload conditions were already attained at the mid concentration of 50 mg/m³, as cited in the publication of the study director (Lee et al. 1986):

“Lung response at 10 mg/m³ satisfied the biological criteria for a "nuisance dust," while adverse effects resulting from gradually accumulated particles (8.1%, 67.7 mg per lung) were found after 1 year of exposure to 50 mg/m³. An early pulmonary response indicating an overloaded lung clearance mechanism was manifested by massive accumulation of dust-laden macrophages (dust cells), foamy dust cells, free particles or cellular debris derived from disintegrated foamy dust cells in the alveoli adjacent to the alveolar ducts. Alveolar proteinosis also appeared to be an important marker of an overloaded lung clearance mechanism and was observed at 50 and 250 mg/m³ after 1 year of exposure.”

To further characterise the bronchoalveolar lesions, in 1992, a group of pathologists from North America and Europe examined lung lesions produced by para-aramid RFP (respirable fibre-shaped particles) and titanium dioxide .This panel diagnosed the lesion as a "proliferative keratin cyst" (PKC). Additionally, the pathologists agreed that the lesion was not a malignant neoplasm and is most likely not neoplastic. A minority opinion was that the lesion is probably a benign tumour (Carlton 1994; Levy 1994)*. Another subsequent international pathology workshop was convened to develop standardized histological criteria for classifying pulmonary keratin lesions (Boorman et al., 1996)*. As a consequence, most of the lesions that had originally been diagnosed as “cystic keratinizing squamous cell carcinomas” were re-classified by the consensus panels as non-tumorous “proliferative keratin cysts” (Warheit and Frame, 2006)*.

In the aftermath of two international pathology workshops designed, in part, to establish histological criteria for classifying pulmonary keratin lesions, these lesions were evaluated by four pathologists using current diagnostic criteria. Microscopic review of 16 proliferative squamous lesions, previously diagnosed as cystic keratinizing squamous cell carcinoma in the lungs of rats from the 2-year inhalation study was performed. Unanimous agreement was reached as to the diagnosis of each of the lesions. Two of the lesions were diagnosed as squamous metaplasia and 1 as poorly-keratinizing squamous cell carcinoma. Most of the remaining 13 lesions were diagnosed as non-neoplastic pulmonary keratin cysts (Warheit and Frame, 2006)*.
Consequently, the diagnosis of many of these lesions has changed after the development of revised criteria.

References:
- NIOSH (2011): Current Intelligence Bulletin 63 – Occupational Exposure to Titanium Dioxide, NIOSH Dept of Health and Human Services
- Carlton W.W. (1994): “Proliferative keratin Cyst", a lesion in the lungs of rats following chronic exposure to para-aramid fibrils, Fundamental Appl. Toxicol. 23, 304-307
- Levy L.S. (1994): Squamous Lung Lesions Associated with Chronic Exposure by Inhalation of Rats to p-Aramid Fibrils (Fine Fiber Dust) and to Titanium Dioxide: Findings of a Pathology Workshop, Mohr, U. (Ed.): Toxic and carcinogenic effects of solid particles in the respiratory tract, ILSI Press, 473-478
- Boorman G.A. et al. (1996): Classification of cystic keratinizing squamous lesions of the rat lung: report of a workshop, Toxicol Pathol. 24, 564–72
- Warheit, D.B. & Frame, S.R. (2006): Characterization and reclassification of titanium dioxide-related pulmonary lesions, J Occup Environ Med 48, 1308-1313

Reference 4

Endpoint:

carcinogenicity: inhalation

Type of information:

experimental study

Adequacy of study:

weight of evidence

Study period:

not specified

Reliability:

3 (not reliable)

Rationale for reliability incl. deficiencies:

significant methodological deficiencies

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 453 (Combined Chronic Toxicity / Carcinogenicity Studies)

Deviations:

yes

Remarks:

only one concentration tested

Principles of method if other than guideline:

It is highlighted that the study was conducted in compliance with OECD 453. However, titanium dioxide was not selected as primary test item (which was carbon black), but as negative control substance. Consequently, only one high titanium dioxide concentration was used, being the maximum tolerated dose (see OECD guidance document 116). Although no dose-response relationship can be derived for the titanium dioxide part of this study, the results can be used for hazard assessment purposes since it can be considered as limit-dose study.

GLP compliance:

yes

Specific details on test material used for the study:

not applicable

Species:

rat

Strain:

Fischer 344

Details on species / strain selection:

The study was done using Fischer-344 rats due to their wide use by the National Toxicology Program and the extensive data base available on their health status and background tumor incidence.

Sex:

male/female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Charles River, Wiga GmbH
- Age: 4 weeks old
- Housing: individually housed in metal wire mesh cages (18.7 x 21 x 15 cm) in horizontal air flow chambers throughout the study.
- Acclimation period: 4 weeks

ENVIRONMENTAL CONDITIONS
- Photoperiod (hrs dark / hrs light): 12/12

Route of administration:

inhalation: aerosol

Type of inhalation exposure (if applicable):

whole body

Vehicle:

air

Remarks on MMAD:

not specified

Details on exposure:

GENERATION OF TEST ATMOSPHERE / CHAMBER DESCRIPTION
- Exposure apparatus: whole-body exposure chambers were used
- System of generating particulates/aerosols: a dry aerosol dispersion technique was used. The aerosol generator consisted of a commercially available feeding system connected to a two-stage pressurised air ejector.
To maximize mixing and uniform distribution of the aerosol, the chamber inlets were equipped with diffusers and perforated plates.
- Temperature, humidity, and air flow: chambers were maintained at 23.5 ± 1 °C and 40 - 60 % relative humidity with an air flow of 3.8 m³/min.

- Method of particle size determination: particle size distribution in each chamber was measured 13 times during the study using Berner impactor.

TEST ATMOSPHERE
- Brief description of analytical method used: at the inlet side of the chamber photometric determination of the aerosol concentration was done and gravimetric samples were collected.

Analytical verification of doses or concentrations:

yes

Details on analytical verification of doses or concentrations:

Please refer to the field "Details on analytical verification of doses or concentrations" above.

Duration of treatment / exposure:

24 months

Frequency of treatment:

6 hours/day, 5 days/week

Post exposure period:

1.5 months

Dose / conc.:

5 mg/m³ air (analytical)

Remarks:

SD: 0.7 mg/m³ air

No. of animals per sex per dose:

144 males / 144 females (total: 288 animals)

Control animals:

yes, concurrent vehicle

Details on study design:

not applicable

Positive control:

not specified

Observations and examinations performed and frequency:

CAGE SIDE OBSERVATIONS: Yes

DETAILED CLINICAL OBSERVATIONS: Yes
- Time schedule: on arrival, 2 and 4 weeks subsequently, and thereafter at 6-month intervals

DERMAL IRRITATION: No data

BODY WEIGHT: Yes
- Time schedule for examinations: every 2 weeks during the first 14 weeks and then once every 4 weeks

FOOD CONSUMPTION:
- Time schedule for examinations: weekly during the first 13 weeks and then once every 3 months.

FOOD EFFICIENCY:
- Body weight gain in kg/food consumption in kg per unit time X 100 calculated as time-weighted averages from the consumption and body weight gain data: No data

WATER CONSUMPTION AND COMPOUND INTAKE: No data
OPHTHALMOSCOPIC EXAMINATION: No data

HAEMATOLOGY: Yes
- Time schedule for examinations: 6, 12, 18 and 25.5 months
- Parameters checked: leukocyte count

CLINICAL CHEMISTRY: Yes
- Time schedule for examinations: 6, 12, 18 and 25.5 months

URINALYSIS: Yes
- Time schedule for examinations: 6, 12, 18 and 25.5 months

NEUROBEHAVIOURAL EXAMINATION: No data

Sacrifice and pathology:

Rats sacrificed during and after scheduled exposure were anesthetized and killed. The abdominal cavity was opened and the diaphragm was cut allowing the lungs to collapse. Organs (brain, liver, kidneys, adrenals, and gonads) were weighed. Tissues and organs were fixed and processed for routine histology. Bones were decalcified. The larynx, trachea, oesophagus, thymus, heart, and lungs were removed. The larynx and upper part of the trachea were separated and placed in formalin. The lungs scheduled for histopathology were fixed. Complete histopathological examination of organs, tissues, and gross lesions of animals of the serial sacrifices and of all animals of the “basic study” was performed (please refer to the table 1 in the field “Any other information on materials incl. tables” below).

Bronchoalveolar lavage: the method of Henderson et al. (1987)* was used with minor modifications. Following preparation of the lungs, they were lavaged with saline without massage. The cell concentration was determined using a counting chamber. The lavagate was centrifuged and the supematant used for the determination of the biochemical parameters (lactic dehydrogenase (LDH), β-glucuronidase, total protein). Cytoslides were prepared for differential cell count.

*Reference:
- Henderson, R.F., Mauderly, J. L., Pickrell, J. A., Hahn, R. F., Muhle, H., and Rebar, A. H. (1987). Comparative study of bronchoalveolar lavage fluid: Effect of species, age and method of lavage. Exp. Lung Res. 13, 329-342.

Statistics:

Parametric data were examined by analysis of variance (ANOVA) followed by Dunnett test to compare various treatment groups with controls. Survival data were analysed by the Kaplan-Meier method using the lifetest program of SAS. For necropsy and tumour occurrance data, simple tests for homogeneity of contingency tables (using qui square statistics or Fisher's exact method) were used.

Clinical signs:

no effects observed

Dermal irritation (if dermal study):

not specified

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 specified

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

not specified

Ophthalmological findings:

not specified

Haematological findings:

no effects observed

Clinical biochemistry findings:

no effects observed

Urinalysis findings:

no effects observed

Behaviour (functional findings):

not specified

Immunological findings:

not specified

Organ weight findings including organ / body weight ratios:

no effects observed

Gross pathological findings:

not specified

Neuropathological findings:

not specified

Histopathological findings: non-neoplastic:

no effects observed

Histopathological findings: neoplastic:

no effects observed

Other effects:

no effects observed

Details on results:

CLINICAL SIGNS
Inhalation of the test material did not cause overt signs of toxicity.
Clinical appearance of the rats was found to be within normal limits. Measured virologic, bacteriologic, and parasitologic parameters were also within normal limits or negative during the study.

MORTALITY:
The mean survival in the basic study (100 animals, 50 of each gender) was 40% at the termination of the experiment. The data analyzed by Kaplan-Meier method, separately for each gender, indicated no difference among various exposure groups.

BODY WEIGHT AND WEIGHT GAIN
Body weight development appeared to be normal when compared to concurrent and historical control.

FOOD CONSUMPTION AND COMPOUND INTAKE
Food consumption was normal in the TiO2-exposed groups.

HAEMATOLOGY, CLINICAL CHEMISTRY and URINALYSIS
Clinical laboratory test results were essentially within normal range. The results were consistent with healthy animals.

ORGAN WEIGHTS
No changes in organ weights were observed at the terminal sacrifice.

HISTOPATHOLOGY: NON-NEOPLASTIC
- TiO2 group: extent of particle-laden macrophages increased with exposure time.
- a small but statistically insignificant incidence of fibrosis was seen in the TiO2 and air-only control groups as shown below:
Control (mild degree of fibrosis): 1.2 (date of sacrifice: 21 - 25.5 months)
TiO2 (moderate degree of fibrosis): 35.7 (date of sacrifice: 21 months); 19.1* (date of sacrifice: 21 - 25.5 months)
TiO2 (minimal degree of fibrosis): 16.7 (date of sacrifice: 15 months); 4.5 (date of sacrifice: 21 - 25.5 months)
TiO2 (mild degree of fibrosis): 1.1 (date of sacrifice: 21- 25.5 monthsmonths)
(*p <0.001; approx. 14 animals/group were examined at each serial sacrifice point and about 90 animals/ group were evaluated at 21 - 26 months of the study)
- no significant difference from air-only control in the extent of various upper respiratory system lesions was noted regarding the TiO2 group.

HISTOPATHOLOGY: NEOPLASTIC
- two adenomas and one adenocarcinoma (3/100 animals) were observed in the air-only control group while one tumour of each type was detected in the TiO2 group (2/100 animals).

PARTICLE RETENTION
TiO2 accumulated progressively in the lungs of the rats. The mean quantity of TiO2 retained was 2.72 mg/lung.

BRONCHOALVEOLAR LAVAGE (BAL) EXAMINATION
The number of lavagable leukocytes at 15 months of exposure was unaffected by treatment. In the lavage fluid a substantial amount of fragments of macrophages were observed. The results from the TiO2 exposure showed a significant decrease in macrophages at 15 months. This cytologic pattern persisted throughout the rest of the study as the results at 21, 24, and 25.5 months were quite similar. A linear relationship was observed between the fraction of polymorphonuclear leukocytes (PMN) in the lavagate and the retention half-time of the polystyrene tracer (Bellmann et al., 1991)* at TiO2 concentration.
The levels of cytoplasmic and lysosomal enzymes and total protein in lavage fluid were comparable to those of air-only controls in the TiO2 groups.
Please also refer to table 1 in the field "Any other information on resutls incl. tables" below.

*Reference:
- Bellmann, B., Muhle, H., Creutzenberg, O., Dasenbrock, C., Kilpper, R., Mackenzie, J., Morrow, P., and Mermelstein, R. (1991). Lung clearance and retention of toner, utilizing a tracer technique during chronic inhalation exposure in rats. Fundam. Appl. Toxicol. 17,300-3 13.

Key result

Dose descriptor:

NOAEC

Effect level:

5 mg/m³ air (analytical)

Based on:

test mat.

Sex:

male/female

Remarks on result:

not determinable due to absence of adverse toxic effects

Critical effects observed:

not specified

Table 1: a) Cytology results in bronchoalveolar lavage fluid and b) mean levels of controls and levels normalized to control of LDH, β-glucuronidase, and protein in lavagate

	Exposure (months)
	15	21	24	24 + 6 weeks clean air
	Number of lavaged leukocytes (10³ cells/mL)
Control	100	133	221	194
TiO2	102	156	211	265
Macrophages (%)
Control	97.2	97.0	98.0	97.4
TiO2	90.8*	93.2	95.7	92.8
PMN (%)
Control	1.1	1.9	0.5	0.8
TiO2	4.9	4.2	2.3	4.3
Lymphocytes
Control	1.7	1.2	1.6	1.9
TiO2	4.4*	2.7	2.1	3.0
LDH
Control (U/liter	25	35	39	33
Control	1.00	1.00	1.00	1.00
TiO2	1.16	0.71	1.26	1.15
β-glucuronidase
Control (U/liter	0.15	0.20	0.18	0.15
Control	1.00	1.00	1.00	1.00
TiO2	1.00	0.70	1.89	1.33
Protein
Control (U/liter	108	114	146	144
Control	1.00	1.00	1.00	1.00
TiO2	1.04	0.99	1.86	1.04

Conclusions:

Inhalation of titanium dioxide showed no signs of overt toxicity. Body weight, clinical chemistry values, food consumption, and organ weights were normal. Fibrosis was present in the controls at a comparable rate to that of titanium dioxide exposed rats, being minimal to mild and not statistically significantly different from controls. There were no significant increases in lung tumours vs. control rats exposed for up to 24 months by whole body inhalation to titanium dioxide in this study.
Therefore a NOAEC(rat) for inhalation tumourigenicity of 5.0 ± 0.7 mg/m³ air (analytical) can be derived on the basis of a complete lack of tumour findings after 24 months inhalation exposure.

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed

Study duration:

chronic

Species:

rat

Carcinogenicity: via dermal route

Endpoint conclusion

Endpoint conclusion:

no study available

Justification for classification or non-classification

Commission Regulation (EU) 2020-217, the 14th adaption to technical progress (ATP) of the CLP introduced a new harmonised classification for certain forms of TiO2 as a category 2 carcinogen by inhalation. The ATP was published in the Official Journal of the EU on 18 February 2020, it came into force on 9 March 2020 and applies from 1 October 2021 onwards. The new harmonised classification specifies two criteria to determine whether a substance is to be classified:

 

 

1. the substance must be in powder form


2. the substance must contain 1% or more of particles with aerodynamic diameter ≤ 10 μm

 

There are no methods officially recognized at the EU level to measure compliance with the second classification criteria mentioned above. There are however internationally recognised methods to measure dustiness. In a dedicated analytical project, TDMA investigated the suitability of all available dustiness test methods for the purpose of determining the content of particles with aerodynamic diameter ≤ 10 μm. As a result, method EN15051-2 provided the most consistent and repeatable results.

 

Further details on the outcome of the dustiness analytical project and the results of the tests using representative forms of titanium dioxide (nano and pigment grade forms) are reported under IUCLID section “4.5 Particle size distribution (Granulometry)” and “4.28.8 Nanomaterial dustiness”, and a summary report is attached to the endpoint summaries of both afore mentioned IUCLID sections.

 

The dustiness of different titanium dioxide forms is highly dependent on inter alia the shape, the particle size, the surface modification of the respective sample. Consequently, the dustiness of titanium dioxide forms needs to be assessed on an individual, company-specific basis due to the above given reasons. The 1% mass-based fraction of particles with an aerodynamic diameter ≤ 10 μm of any specific titanium dioxide sample and the subsequent classification as a category 2 carcinogen by inhalation will be included in the accompanying material safety data sheet and thus any required risk management measures are communicated to the downstream users.

Additional information

Overall Conclusions

The rat has been found to be uniquely sensitive to the formation of lung tumours when exposed under conditions of severe particle overload to titanium dioxide and other poorly soluble low-toxicity particles (Levy, 1995). Although particle overload is observed in other experimental species such as the mouse, it is only in the rat that a sequence of events is initiated that leads to fibroproliferative disease, septal fibrosis, hyperplasia and eventually lung tumours. Similar pathological changes are not observed in other common laboratory rodents, in non-human primates or in exposed humans. In addition, detailed epidemiological investigations have shown no causative link between titanium dioxide exposure and cancer risk in humans. At workplace exposure concentrations, no lung cancer hazard has been observed

 

Animal data

Chronic Studies: pigment-grade titanium dioxide

In 2-year chronic inhalation studies, male and female CD rats were exposed to titanium dioxide (pigment-grade) at concentrations of 10, 50, or 250 mg/m³ for six hours a day, five days a week (Lee et al., 1985). The majority of the titanium dioxide was of respirable size (95% of the particles were less than 10 microns and average particle diameter was about 1.5 microns for high dose group).

The findings of lung tumours at 250 mg/m³ described in the Lee et al. (exposure concentrations used: 0; 10; 50; 250 mg/m³, 6h/d, 5d/w) study clearly exceeded the maximum tolerable dose (MTD), with complete cessation of alveolar clearance. It would therefore appear inappropriate to be considered as a positive tumour response. In their evaluation, ECHA RAC (2017) discounted this study for hazard characterisation purposes primarily because of the above limitations, assuming an exceedance of 60% volumetric alveolar macrophage loading:

“The experimental schedule of the Lee et al. (1985) study resulted in a complete cessation of alveolar clearance already at the non-carcinogenic exposure level of 50 mg/m³. Alveolar clearance half-times measured in different studies at the exposure level of 250 mg/m³ reached and exceeded 1 year. RAC takes the view, that these exposure conditions represent excessive exposure which invalidates the results of the Lee et al. (1985) study on their own for classification purposes.”

Likewise, the NIOSH Executive Summary in their Current Intelligence bulletin (NIOSH 2011, pages VI-VII), designated the 250 mg/m³ concentration in the Lee et al., 1985 study as an excessive dose not relevant for human risk assessment:

”However, exposure concentrations greater than 100 mg/m³ are generally not considered acceptable inhalation toxicology practice today. Consequently, in a weight-of-evidence analysis, NIOSH questions the relevance of the 250 mg/m³ dose for classifying exposure to TiO2 as a carcinogenic hazard to workers and therefore concludes that there are insufficient data at this time to classify fine TiO2 as a potential occupational carcinogen.”

To further characterise the bronchio-alveolar lesions, a group of pathologists from North America andEurope examined lung lesions produced by para-aramid RFP (respirable fibre-shaped particles) andtitanium dioxidein 1992.This panel diagnosed these lesions as "proliferative keratin cysts"(PKC). Additionally, the pathologists agreed that this lesion was not a malignant neoplasm and also most likely not neoplastic. A minority opinion was that the lesion is probably a benign tumour(Carlton 1994; Levy 1994).Another subsequent international pathology workshop was convened to develop standardised histological criteria for classifying pulmonary keratin lesions (Boorman et al., 1996).

In the aftermath of these two international pathology workshops designed, in part, to establish histological criteria for classifying pulmonary keratin lesions, these lesions from the Lee ta l. (1995) study were re-evaluated by four pathologists using current diagnostic criteria. Microscopic review of 16 proliferative squamous lesions, previously diagnosed as cystic keratinizing squamous cell carcinoma in the lungs of rats from the 2-year inhalation study was performed. Unanimous agreement was reached as to the diagnosis of each of the lesions. Two of the lesions were diagnosed as squamous metaplasia and 1 as poorly-keratinizing squamous cell carcinoma. Most of the remaining 13 lesions were diagnosed as non-neoplastic pulmonary keratin cysts (Warheit and Frame, 2006).

Lee et al (1985) noted that, due to excessive loading in the lungs of rats exposed chronically at 250 mg/m³, the lung tumours were different from common human lung cancers in terms of tumour type, anatomic location, tumorigenesis and were devoid of tumour metastasis. Therefore, the biological relevance of these lung tumours was deemed negligible.

 

Table 1: Lung Tissue Analysis of TiO2

Exposure time (months)	Sex	0mg/m³	10mg/m³	50mg/m³	250mg/m³
		TiO2in dried lung tissue (%)
3	M	ND	0.60	3.07	12.6
	F	ND	0.60	3.12	13.7
6	M	ND	1.28	6.97	17.9
	F	ND	0.88	7.02	18.0
12	M	ND	1.97	8.37	16.6
	F	ND	2.14	7.84	17.2
24	M	ND	3.2	10.7	31.5
	F	ND	3.0	8.41	24.4

 

Table 2: Statistical analysis of Lung Weights (average TiO₂weights per lung at different exposure intervals)

Exposure time (months)	Sex	0mg/m³	10mg/m³	50mg/m³	250mg/m³
		TiO2weight (mg/lung)
3	M	ND	2.5	21.7	180.8
	F	ND	2.8	16.6	136.8
6	M	ND	4.8	57.3	275.3
	F	ND	4.4	54.0	238.6
12	M	ND	10.1	75.6	361.7
	F	ND	8.7	59.7	381.5
24	M	ND	20.7	118.3	784.8
	F	ND	32.3	130.0	545.8

 

Further, lung burdens of 118 and 130 mg/lung of TiO2 in male and female rats exposed to 50 mg/m³ TiO2 for 2 years did not result in lung tumours. In their Current Intelligence Bulletin document, NIOSH (2011) questioned/dismissed the relevance of the results following the exposures to 250 mg/m³ of the 2-year study. Moreover, Lee et al. themselves commented that the relevance of these rat tumours for humans was negligible, due to anatomic type, location, etc. Moreover, at the 50 mg/m³ exposure levels, there were no tumours in either male or female rats representing lung burdens of 118 mg/lung in males and 130 mg/lung in females.

The accumulation of particles in the lung was caused by overwhelming the normal pulmonary clearance mechanisms of the respiratory system. Such extreme disturbance of physiological mechanisms are considered by internationally accepted guidelines to be in exceedance of the maximum tolerated dose (MTD). The guideline OECD 451 for the conduct of carcinogenicity studies, in conjunction with the relevant OECD guidance document 116, highlights on various occasions that inhalation concentrations overwhelming physiological mechanisms are in exceedance of the MTD:

a)     The guidance explicitly states that “inhalation of doses that overwhelm pulmonary clearance may lead to tissue responses that are specific to the species being tested” (section 94, p.54).

b)     “The robustness of a carcinogenicity or chronic toxicity study, in particular the former, is dependent on a demonstration that the dose levels selected in the study are adequate to show an effect or effects of the test substance, without producing either false negative results (because the doses selected were too low) or false positive results (because metabolic/homeostatic mechanisms are overwhelmed, etc.), which may be problematic in assessing risk in humans” (section 101, page 55).

c)      In the selection of the maximum concentration, it states that it should be considered that “disturbances of physiology or homeostasis that would compromise the validity of the study should be considered in the dose-selection process. Examples include hypotension, inhibition of blood clotting, overwhelming normal pulmonary clearance mechanisms, immune system effects, and in some cases hormonal imbalance” (p.63).

d)     “For substances likely to accumulate in the lung over time due to poor solubility or other properties, the degree of lung-overload and delay in clearance needs to be estimated based on adequately designed pre-studies; ideally a 90-day study with post-exposure periods long enough to encompass at least one elimination half-time. The use of concentrations exceeding an elimination half-time of approximately 1 year due to lung-overload at the end of study is discouraged” (section 135, p.71).

As demonstrated above, all concentrations applied in the Lee et al. study clearly exceeded the maximum tolerated dose, since lung overload and overwhelmed lung clearance mechanisms were already achieved at the lowest concentration. Consequently, the Lee et al. study is considered unsuitable for hazard and risk assessment purposes due to an inadequate dose regime, being not compliant with current guidelines. However, due to an absence of adverse effects at the lowest concentration, it is considered that titanium dioxide shows no adverse effects up to the limit concentration, limited by the maximum tolerated dose.

 

In a study reported by Muhle et al. (1991), titanium dioxide was used as a negative control dust in a two-year inhalation study with toner particles. Male and female rats were exposed (6 hr/day, 5 days/week) to 5 mg/m³ titanium dioxide (rutile form) of 1.1μm MMAD2 with a respirable fraction of 78%. Inhalation of titanium dioxide showed no signs of overt toxicity. Body weight, clinical chemistry values, food consumption, and organ weights were normal. Fibrosis was present in the controls at a comparable rate to that of titanium dioxide exposed rats, being minimal to mild and not statistically significantly different from controls. There were no significant increases in lung tumours vs. control rats exposed for up to 24 months by whole body inhalation to titanium dioxide in this study.

 

Chronic Studies: ultrafine grade titanium dioxide

Heinrich et al. (1995) exposed female Wistar rats by whole body inhalation to ultrafine titanium dioxide (80% anatase: 20% rutile) at an average concentration of 10 mg/m³ (single exposure concentration: 7.2 mg/m³ 1-4 months, 14.8 mg/m³ 5-8 months, 9.4 mg/m³ 9-24 months, 18h/d, 5d/w) for 24 months followed by 6 months without exposure. The primary particle size of the titanium dioxide used ranged from 15 to 40 nm with a MMAD of 0.8 mm (agglomerates of ultrafine particles). The mean lifetime of rats was shortened, with 90% mortality in the TiO2 exposed animals at the end of the 130-week experiment compared with the control animals with 85% mortality. At the end of 2-year exposures, body weights of exposed animals were significantly lower than controls. Wet lung weights increased substantially during exposure, progressing with study duration. This was reflected in retained test material in the lungs, which increased from 5 mg/lung to 39 mg/lung from 3 months to 24 months of exposure. This value is higher than the retained material in the Lee et al. study (26.5 mg) at the identical exposure concentration of 10mg/m³, which similarly showed impairment of lung clearance and therefore an exceedance of the maximum tolerated dose. This is also reflected in the measured alveolar lung clearance rates, which were significantly compromised after 3 months of exposure and a 3-month recovery period following 18 months of exposure showed no reversibility of the effect.

Table: Half-times of pulmonary tracer clearance (59Fe oxide, taken from Heinrich et al. 1995)

Exposure	3 months	12 months	18 months	18 months + recovery
Control	61	72	96	93
TiO2	208*	403*	357*	368*

* significant at p<0.01 (Dunnett’s test)

Bronchioalveolar hyperplasia of moderate to severe grade and slight to moderate interstitial fibrosis of the lungs was present after 2 years of exposure in rats. There were no lung tumours observed in TiO2, satellite groups of about 20 animals each after 6 and 12 months of exposure. First lung tumours were found in serially sacrificed animals after 18 months of exposure to TiO2. Heinrich reported that 32/100 rats developed lung tumours after exposures to ultrafine TiO2. These included benign squamous tumours, 3 squamous cell carcinomas, adenomas and 13 adenocarcinomas. However, the publication of this report preceded the revised criteria based on 2 workshops for classification of cystic keratinizing squamous lesion of the rat lung. (Carlton, 1994; Levy 1994 and Boorman et al., 1996). As a consequence, the determination that these are malignant tumours should be properly re-evaluated according to the new criteria, before a definitive conclusion can be made. Dueing the RAC discussion it was noted that the pathological diagnosis of lung tumours in the study of Heinrich et al. (1995) was also adapted according to the revised diagnostic terms (Rittinghausen et al. 1997).

“Two histological types of rat lung tumours were observed: there was a 13/100 incidence of adenocarcinoma in test animals (versus 1/217 in controls). The original findings of cystic keratinizing lesions were re-evaluated based on the already mentioned revised diagnostic criteria and classification: following this re-evaluation, the incidence in female rats with cystic keratinizing epitheliomas (considered to be benign tumours) was 16/100, the incidence of squamous cell carcinomas was 4/100 (including 3/100 cystic keratinizing squamous cell carcinoma; Rittinghausen et al 1997).”

 However, as already discussed above for the Lee et al. study, the concentration applied in this study clearly exceeded the maximum tolerated dose, since lung overload and overwhelmed lung clearance mechanisms were determined and reported by the authors themselves. Theuse of a single dose-regime and the significantly extended daily exposure duration to 18 hrs per day renders this study non-guideline compliant and of limited relevance for hazard- and risk assessment purposes.A detailed critique of the study by Heinrich et al. (1995) is provided in the Appendix. Nevertheless, ECHA RAC (2017) considered this study as “acceptable” on the grounds that the impairment of alveolar clearance was not complete, attaining 40% of volumetric alveolar macrophage loading (details/results of underlying calculation not provided).In the same study, the exposure of female NMRI mice to ultrafine titanium dioxide under the same conditions as for rats resulted in a significantly decreased lifespan at an inhalation concentration of approximately 10 mg/m³ (Heinrich et al., 1995). Lung wet weights and particle lung burdens increased as for rats. Retained test material in mouse lung was 26 mg/lung after 1 year of exposure as compared with 29 mg/lung for similarly exposed rats. No other pathology-related findings were however reported for titanium dioxide in this study.

 

Subchronic inhalation studies (reported under the endpoint: repeated dose toxicity)

Subchronic inhalation exposures of rats, mice, and hamsters to either pigmentary or ultrafine titanium dioxide particles at concentrations likely to induce particle overload produced a more severe and persistent pulmonary inflammatory response in rats, as compared with either mice or hamsters. Rats were unique among these three species in the development of progressive fibroproliferative lesions and alveolar epithelial metaplasia (Bermudez et al., 2002 and 2004). Thus, female rats, mice or hamsters were exposed to 10, 50 or 250 mg/m³ concentrations of pigmentary (rutile type) titanium dioxide particles for 6 hours/day, 5 days/week for 13 weeks followed by 4, 13, 26 or 52 weeks of post exposure (46 weeks for hamsters) (Bermudez et al., 2002). Lung and associated lymph node loads of titanium dioxide increased in a concentration-related manner. Retained lung burdens were greatest in mice following exposure, with rats and hamsters displaying similar lung burdens immediately following exposure. Particle retention data indicated that particle overload in the lungs was reached in both rats and mice at the 50 and 250 mg/m³ concentrations. Inflammation was observed in all three species at the two highest concentrations. This inflammation persisted in rats and mice throughout the post exposure recovery period at the highest exposure concentration. In hamsters, inflammatory responses were eventually resolved due to the more rapid clearance of particles from the lung. In rats exposed to the highest concentration (250 mg/m³), pulmonary lesions consisted of epithelial proliferative changes manifested by increased alveolar epithelial cell labelling indices, as evidenced by the results of cell proliferation studies. Associated with these proliferative changes in the rat were increased interstitial particle accumulations and alveolar septal fibrosis. Although rats exposed to 50 mg/m³ developed minimal alveolar cell hypertrophy, accumulation of particle-laden macrophages, and inflammation, no alveolar septal fibrosis or relevant cell turnover at alveolar sites were observed at this lower exposure concentration. Similar changes to those seen in rats were not observed in either mice or hamsters.

In a study with ultrafine titanium dioxide (80% anatase: 20% rutile; average primary particle size = 21 nm), female rats, mice or hamsters were exposed to aerosol concentrations of 0.5, 2.0 or 10 mg/m³ titanium dioxide for 6 hours/day, 5 days/week for 13 weeks followed by 4, 13, 26 or 52 weeks of post-exposure (49 weeks for hamsters)(Bermudez et al., 2004). Retained lung burdens increased in a concentration-related manner in all three species. Mice and rats had similar lung burdens at the end of exposures but hamsters were significantly lower. Retardation of particle clearance in rats and mice at the highest exposure concentration (10 mg/m³) indicated that pulmonary particle overload had been achieved. Lesions of the lungs in rats consisted of foci of alveolar epithelial proliferation of metaplastic epithelial cells (alveolar bronchiolisation) concomitant with circumscribed areas of heavy, particle-laden macrophages. In rats these changes were manifested by increased alveolar epithelial cell labelling indices, as evidenced by cell proliferation studies. Associated with these foci were areas of interstitial particle accumulation and alveolar septal fibrosis. These lesions observed in the rat became more pronounced with time. Mice developed a less severe inflammatory response without the progressive epithelial and fibroproliferative changes.

 

Due to the experimental design, the 19 dust study (Pott F and Roller M, 2005) data for female rats cannot be interpreted in a manner that makes it useful for human risk assessment and development of airborne dusts limits in workplace environments. Likewise, because of the species (rat) and the exposure conditions (particle overload in the lungs), the data are not applicable to hazard characterisation of granular biopersistent particles (GBP) as to human carcinogenicity. Problems with interpretation of the experiments include:

1.      Dosages used: P25 (5 x 3mg/rat, 5 x 6 mg/rat, or 10 x 6mg/rat); P805 (15 x 0.5 mg/rat, or 30 x 0.5 mg/rat); micro-TiO2 (10 x 6 mg/rat, or 20 x 6 mg/rat)

2.      The high doses and the high dose-rate delivery led to lung overload in the rats. 19-DS lung instillations were performed in a dose range that lacks relevance to the actual exposure that occur at workplaces. [TiO2 anatase – instilled mass = 120 mg into lungs; TiO2 hydrophilic ultrafine (UF) – 60 mg into lungs; TiO2 hydrophobic – 15 mg into lungs].

3.      The responses of rat lungs to overload conditions are unique to this species and not particle-specific.

4.      The study did not include low-dose studies and only female rats were employed. The rat-lung inflammatory response has a mechanistic threshold.

5.      The occupational epidemiology results for workers in dusty trades experiencing historically elevated levels of airborne dust do not bear out the tumorigenicity of GBP, as might be predicted from the study results.

Using the same dusts administered as single doses to rats by intratracheal instillation, but at lower doses corresponding to permissible workplace levels, Rehn et al. (2003; dosages used: P25 and T805 (0.15, 0.3 0.6 and 1.2 mg/rat) evaluated the lungs at 3, 21, or 90 days post-exposure by bronchoalveolar lavage to gauge the lung inflammatory and genotoxic reactions. Quartz particles were used as a positive control. The authors concluded that both types of TiO2 were not different from saline controls.

The intratracheal instillation study by Pott and Roller (2005) utilised excessive doses or non-physiological routes of administration and should therefore not be considered for hazard classification of granular biopersistent particles (GBP) as to human carcinogenicity (Valberg et al., 2009).

 

Summary of Animal Data

Inhalation exposures to titanium dioxide under conditions causing pulmonary overload have produced primarily benign lung tumours in rats, but no tumours in other common laboratory rodents. A number of conclusions can be reached concerning the lack of relevance of the rat lung tumour data for the carcinogenicity classification of titanium dioxide.

The lack of relevance of rat lung tumours following chronic inhalation exposures to PSPs of low cytotoxicity can be summarised by the following five factors (Warheit et al. 2016):

1.      Data and findings from three subchronic, 90-day interspecies rodent inhalation studies provide convincing mechanistic justifications to better understand the distinct differences in cellular responses to particle overload exposures when comparing rats to either mice or hamsters (Bermudez et al., 2002; Bermudez et al., 2004; Elder et al., 2005; Carter et al., 2006). In addition, a conceptual AOP scenario has been developed (ECETOC, 2013) for the rat pulmonary response to particle-overload, leading to lung tumours which is substantively different from pulmonary responses demonstrated in particle-exposed mice or hamsters and/or in either nonhuman primates or coal workers. In chronic inhalation studies to TiO2 and carbon black particles, only rats developed tumours – but not mice exposed to the same particles/concentrations.

 

2.      Several 2-year inhalation studies have compared the effects of similarly or identically exposed rats and monkeys to a variety of low solubility dusts, such as shale dust, petroleum coke dust and diesel exhaust particles (Wagner et al., 1969; Klonne et al, 1987; Lewis et al., 1989; MacFarland et al., 1982; Nikula et al., 1997; Nikula et al., 2000). In every case, the lung cellular responses of rats exposed chronically to particles were considered hyperinflammatory and hyperplastic, while the pulmonary responses in monkeys were limited to general, normal physiological effects (particle accumulation, macrophage responses) to inhaled particles. In addition, morphometric studies reported by Nikula et al., 1997 were developed to investigate the distribution patterns of inhaled particles in both chronically-exposed rats and cynomolgus monkeys. The results demonstrated that the majority of inhaled particles that deposited in the distal regions of the lung had transmigrated to interstitial compartments of the lungs of nonhuman primates. In contrast to the pulmonary responses and particle distribution patterns measured in monkeys, inhaled particles in diesel and coal dust exposed rats after deposition were retained primarily on alveolar surfaces, and subsequently stimulated active inflammatory responses. In another set of morphometric studies assessing the particle disposition pattern in deceased coal miners, particle distribution patterns similar to cynomolgus monkeys were measured. In this regard, most of the coal particles had translocated to interstitial sites (Nikula et al., 2001).

 

3.      The ICRP – Human Respiratory Tract Model has been an internationally recognised standard model to estimate the deposition, clearance and retention patterns for workers in the nuclear and coal dust industry (ICRP, 1994). The model has been updated/revised by Gregoratto et al. (2010; 2011) to demonstrate that a greater proportion of inhaled low solubility dusts translocate from alveolar/respiratory bronchiolar sites of initial particle deposition to interstitial sites. This updated revision has important implications for lung clearance and retention estimates of inhaled particles, and supports species differences in particle distribution patterns, in particular the finding of enhanced translocation of inhaled particles to the interstitium. The impact of the model supports increases in the retention time of particles in the human lung. It is also noteworthy that the finding of enhanced transmigration rates in these models also correlates well with the morphometric findings reported by Nikula et al. (2001) in particle-exposed lungs of nonhuman primates and coal workers.

 

4.      Fundamental differences have been recognised by human and veterinary pathologists when considering the characterisation and location of tumour types in rats chronically exposed to PSPs of low cytotoxicity vs. humans exposed to cigarette smoke or asbestos fibres (Schultz, 1996; Green, 2000). First, many PSP-induced rat neoplasms are unique species-specific entities that are only consistently observed in particle overload instances. Furthermore, there is no known documentation of human production workers developing an increase in lung cancers following exposure to poorly soluble particulates. Moreover, the types of lung tumours characterised in humans exposed to cigarette smoke or asbestos fibres – occur primarily in the bronchiolar regions of the respiratory tract and do not have the “squamous or keratinising” features of rat lung tumours, which are more prominent in this region of the lung following chronic exposures to such PSPs. It is generally acknowledged that comparing asbestos and cigarette smoke-induced tumours in humans to such PSP-induced neoplastic entities in the rat probably does not contribute meaningfully to cancer risk of such PSPs, as the lungs differ in morphological aspects such as the presence (humans) and absence (rodents) of a respiratory bronchiole (Schultz, 1996). Nonetheless, it should be recognised that cystic keratinising tumours of rats arise very differently than squamous lesions in humans and appear to be adaptive versus true neoplastic changes (Carlton, 1994; Levy, 1994). In the Lee et al. study 1985 – in which rats developed tumours after being exposed to 250 mg/m³ (but not at 50 mg/m³), it was noted by Lee that the lung tumours were different from common human lung cancers in terms of tumour type, anatomic location, tumorigenesis and were devoid of tumour metastases. Therefore, these lung tumours were deemed biologically irrelevant.

 

5.      All of the published epidemiological studies on titanium dioxide (Boffetta et al. 2004; Chen and Fayerweather, 1988, Ellis et al., 2010 and 2013; and Fryzek et al. 2003), carbon black and toner production workers demonstrate no association between working life-time exposures to PSPs of low cytotoxicity and lung cancer and/or non-cancer respiratory disease.

 

The relevance of particle-overload related lung tumours in rats for human risk assessment following chronic inhalation exposures to poorly soluble particulates (PSP) of low cytotoxicity has been a controversial issue for more than 30 years. In 1998, an ILSI (International Life Sciences) Working Group of health scientists was convened to address this issue of applicability of experimental study findings of lung neoplasms in rats for lifetime-exposed production workers (ILSI, 2000). A full consensus view was not reached by the Workshop participants, but it was generally acknowledged that the findings of lung tumours in rats following chronic inhalation, particle-overload PSP exposures were unique to rats; and that there was an absence of lung cancers in PSP-exposed production workers. Subsequently following up on this, a further thorough and comprehensive review of the health effects literature on poorly soluble particles/lung overload was published by an ECETOC Task Force in 2013. One of the significant conclusions derived from that technical report specified that the rat represents a uniquely sensitive lung tumour model under chronic inhalation overload exposures to PSPs of low cytotoxicity.

 

Despite the above discussed limitations in the experimental data set on inhalation carcinogenicity of TiO2, ECHA RAC (2017) proposed a cat 2 carcinogenicity classification, with explicit recognition of the deficiencies in the data for TiO2 but at the same time making explicit reference to similar data on other PSLTs. It is also worthy of note that RAC also clearly designated the rat lung tumours as being elicited merely by a “physical, particle effect” and not aTiO2-specific chemically-induced effect.

 

 

 

References

Boffetta, P. et al. (2004): Mortality among workers employed in the titanium dioxide production industry in Europe, Cancer Causes Control. 15, 697-706

Boorman G.A. et al. (1996): Classification of cystic keratinizing squamous lesions of the rat lung: report of a workshop, Toxicol Pathol. 24, 564–72

Carlton W.W. (1994): “Proliferative keratin Cyst", a lesion in the lungs of rats following chronic exposure to para-aramid fibrils, Fundamental Appl. Toxicol. 23, 304-307

Carter, J.M. et al. (2006): A comparative dose-related response of several key pro- and antiinflammatory mediators in the lungs of rats, mice, and hamsters after subchronic inhalation of carbon black, J Occup Environ Med 48, 1265-1278

Chen, J.L. & Fayerweather, W.E. (1988): Epidemiologic study of workers exposed to titanium dioxide, J Occup Med 30, 937-942

ECETOC (2013): Poorly Soluble Particles/Lung Overload, European Centre for Ecotoxicology and Toxicology of Chemicals, Technical Report No. 122

ECHA (2017): Committee for Risk Assessment RAC, Opinion proposing harmonised classification and labelling at EU level of Titanium dioxide, Document: CLH-O-0000001412-86-163/F

Elder, A. et al. (2005): Effects of subchronically inhaled carbon black in three species, I. Retention kinetics, lung inflammation, and histopathology, Toxicol Sci. 88(2), 614-29

Ellis, E.D. et al. (2010): Mortality among titanium dioxide workers at three DuPont plants, J Occup Environ Med. 52(3), 303-9

Ellis, E.D. et al. (2013): Occupational exposure and mortality among workers at three titanium dioxide plants, Am J Ind Med 56, 282-291

Fryzek, J.P. et al. (2003): A Cohort Mortality Study Among Titanium Dioxide Manufacturing Workers in the United States, J Occup. Environ. Med 45, 400-409

Green, F.H.Y. (2000): Pulmonary responses to inhaled poorly soluble particulate in the human, Inhal. Toxicol 12, 59-95

Gregoratto, D. et al. (2010): Modelling particle retention in the alveolar-interstitial region of the human lungs, Journal of radiological protection 30, 491–512

Gregoratto, D. et al. (2011): Particle clearance in the alveolar-interstitial region of the human lungs: model validation, Radiat Prot Dosimetry 144, 353–356

ICRP (1994): Human respiratory tract model for radiological protection, Smith, H. (ed), Annals of the ICRP, ICRP Publication No. 66, International Commission on Radiological Protection, Tarrytown, New York

ILSI Risk Science Institute (2000): The relevance of the rat lung response to particle overload for human risk assessment: a workshop consensus report, ILSI Risk Science Institute, Inhal. Toxicol. 12, 1-17

Klonne, D.R. et al. (1987): Two-year inhalation toxicity study of petroleum coke in rats and monkeys, Am. J Ind. Med 11, 375-389

Levy L.S. (1994): Squamous Lung Lesions Associated with Chronic Exposure by Inhalation of Rats to p-Aramid Fibrils (Fine Fiber Dust) and to Titanium Dioxide: Findings of a Pathology Workshop, Mohr, U. (Ed.): Toxic and carcinogenic effects of solid particles in the respiratory tract, ILSI Press, 473-478

Levy, L.S. (1995): Review: The 'Particle Overload' phenomenon and human risk assessment, Indoor and Built Environment 4, 254-262

Lewis, T.R. et al. (1989): A chronic inhalation toxicity study of diesel engine emission and coal dust, alone and combined, J. Am Coll. Toxicol 8, 345-375

MacFarland, H.N. et al. (1982): Long-term inhalation studies with raw and processed shale dusts, Ann. Occup. Hyg. 26, 213-224

Nikula K.J, et al. (1997): Lung tissue responses and sites of particle retention differ between rats and cynomolgus monkeys exposed chronically to diesel exhaust and coal dust, Fundam Appl Toxicol 37, 37-53

Nikula K.J. (2000): Rat lung tumors induced by exposure to selected poorly soluble nonfibrous particles,Inhalat Toxicol12,97–119

Nikula, K.J. et al. (2001): Influence of exposure concentrations or dose on the distribution of particulate material in rat and human lungs, Environ. Health Perspect 109, 311-318

NIOSH (2011): Current Intelligence Bulletin 63 – Occupational Exposure to Titanium Dioxide, NIOSH Dept of Health and Human Services

Rehn, B. et al. (2003): Investigations on the inflammatory and genotoxic lung effects of two types of titanium dioxide, Toxicol Appl Pharmacol 189, 84-95

Rittinghausen, S. et al. (1997): Pulmonary cystic keratinizing squamous cell lesions of rats after inhalation/installation of different particles. Exp. Toxicol. Pathol. 49, 433-446

Schultz M. (1996): Comparative pathology of dust-induced pulmonary lesions: Significance of animal studies to humans, Inhalation Toxicology 8, 433-456

Valberg, P.A. et al. (2009): Are rats results from intratracheal instillation of 19 granular dusts a reliable basis for predicting cancer risk?, Reg. Toxicol Pharmacol. 54, 72-83

Wagner, W.D. et al. (1969): Comparative chronic inhalation toxicity of beryllium ores, bertrandite and beryl, with production of pulmonary tumors by beryl, Toxicol Appl. Pharmacol. 15, 10-29

Warheit, D.B. et al. (2016): Relevance of the rat lung tumor response to particle overload for human risk assessment—Update and interpretation of new data since ILSI 2000. Toxicology 374, 42-59