Fumes, silica

Carcinogenicity

Description of key information

A two year rat and mouse feeding study with micronized silica gel (Syloid) by Takizawa et al. (1988) showed no increased incidence of cancer. Limited intrapleural studies with synthetic amorphous silica did not show increased incidence of cancer. Epidemiological studies have not raised concerns about lung carcinogenicity.

Key value for chemical safety assessment

Carcinogenicity: via oral route

Link to relevant study records

Reference

Endpoint:

carcinogenicity: oral

Type of information:

experimental study

Adequacy of study:

weight of evidence

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

guideline study with acceptable restrictions

Reason / purpose for cross-reference:

reference to same study

Qualifier:

equivalent or similar to guideline

Guideline:

other: OECD 452 (Chronic toxicity studies)

GLP compliance:

no

Species:

other: mouse and rat

Strain:

other: B6C3F1 mice and Fisher rats

Sex:

male/female

Details on test animals or test system and environmental conditions:

TEST ANIMALS
- Source: Funabashifarm Animal Co. Ltd, Japan
- Age at study initiation: mice 4 weeks, rats 3 weeks
- Weight at study initiation: male-mice 21.0-27.3 g, female-mice 16.0-19.9 g; male-rats 117-150 g, female-rats 92.0- 126 g
- Fasting period before study:
- Housing: in wire-mesh cages, separated according to sex, 5 mice per cage, 2 rats per cage
- Diet (e.g. ad libitum):
- Water (e.g. ad libitum): tap water ad lipidum
- Acclimation period: 1 week for mice; 2 weeks for rats


ENVIRONMENTAL CONDITIONS
- Temperature (°C): 23.1+-1
- Humidity (%): 50+-10
- Air changes (per hr): air-conditioned
- Photoperiod (hrs dark / hrs light): artifical fluorescent lighting daily for a continuous 14-hour period


IN-LIFE DATES: From: To:

Route of administration:

oral: feed

Details on exposure:

PREPARATION OF DOSING SOLUTIONS:


DIET PREPARATION
- Rate of preparation of diet (frequency): prepared weekly
- Mixing appropriate amounts with (Type of food):
- Storage temperature of food:


VEHICLE
- Justification for use and choice of vehicle (if other than water):
- Concentration in vehicle:
- Amount of vehicle (if gavage):
- Lot/batch no. (if required):
- Purity:

Analytical verification of doses or concentrations:

no

Duration of treatment / exposure:

93 weeks (mice), 103 weeks (rats)

Frequency of treatment:

daily

Post exposure period:

no

Remarks:

Doses / Concentrations:
0, 1.25, 2.5, and 5%
Basis:
nominal in diet

Remarks:

Doses / Concentrations:
0, 2500, 5000, 10000 mg/kg/d (mice)
Basis:
nominal in diet

Remarks:

Doses / Concentrations:
0, 625, 1250, 2500 mg/kg/d (rats)
Basis:
nominal in diet

No. of animals per sex per dose:

10-20 animals/sex/dose group

Control animals:

yes

Observations and examinations performed and frequency:

CAGE SIDE OBSERVATIONS: Yes
- Time schedule: daily for survival
- Cage side observations checked in table [No.?] were included.


DETAILED CLINICAL OBSERVATIONS: No
- Time schedule:

BODY WEIGHT: Yes
- Time schedule for examinations: with mice 0, 5, 15, 30, 50, 81 and 93 weeks after feeding; with rats 0, 5, 15, 30, 50, 81 and 103 weeks after feeding


FOOD CONSUMPTION AND COMPOUND INTAKE (if feeding study):
- Food consumption for each animal determined and mean daily diet consumption calculated as g food/kg body weight/day: Yes
- Compound intake calculated as time-weighted averages from the consumption and body weight gain data: Yes


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


OPHTHALMOSCOPIC EXAMINATION: No


HAEMATOLOGY: Yes
- Time schedule for collection of blood: 6, 12 and 21 months for mice and 6, 12 and 24 months for rats
- How many animals: 89/160 male mice and 105/158 female mice; 118/161 male rats and 123/161 female rats
- Parameters checked in table [No.?] were examined: erythrocytes (RBC), hemoglobin (Hb), leucocytes (WBC) and hematocrit (Ht)


CLINICAL CHEMISTRY: Yes
- Time schedule for collection of blood: 6, 12 and 21 (mice) or 24 (rats) months
- How many animals: 89/160 male mice and 105/158 female mice; 118/161 male rats and 123/161 female rats
- Parameters examined: aspartate transaminase (AST), alanine transaminase (ALT), serum inorganic phosphorus (IP), total protein (TP), albumin (ALB), lactic dehydrogenase (LDH), alkali phosphatase (ALP), total bilirubin (TB), total cholesterol (T-Cho), high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), triglyceride (TG), blood urea nitrogen (BUN), uric acid (UA), creatine (Cre), and calcium (CA) on serum separated from the blood after clotting


URINALYSIS: No

NEUROBEHAVIOURAL EXAMINATION: No

Sacrifice and pathology:

10 animals/sex/dose group were killed after 6 and 12 months and the remaining animals (20/sex/dose group) were reserved for 21 months.

Statistics:

The mean and standard deviations of various measured parameters were calculated for each dose group. The significant difference between the control and the compound-treated groups was tested by Student's t-analysis variance test (P<0.05*; P<0.01**). The chi-square test of significance (P<0.05) by Mantel-Hanszel was employed to compare the survival date exclusive of sacrificed specimens. Prevalence rates were cited as percentages of tumor groups and non-tumor groups in cases of post-mortem examination. The significance of differences between the two means of prevalence was tested by Fisher's exact test for fourfold tables. The percentages of the frequencies of tumor in specific tissues were analysed by the Cochrane-Armitage test for linear trend in proportion with continuity correction.

Clinical signs:

no effects observed

Mortality:

no mortality observed

Body weight and weight changes:

no effects observed

Food consumption and compound intake (if feeding study):

no effects observed

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

not examined

Ophthalmological findings:

not examined

Haematological findings:

no effects observed

Clinical biochemistry findings:

no effects observed

Urinalysis findings:

not examined

Behaviour (functional findings):

not examined

Organ weight findings including organ / body weight ratios:

no effects observed

Gross pathological findings:

no effects observed

Histopathological findings: non-neoplastic:

no effects observed

Histopathological findings: neoplastic:

no effects observed

Details on results:

CLINICAL SIGNS AND MORTALITY:
Most of the mice remained in good health, appeared to be active, and showed normal behavior throughout the treatment. No significant difference in survival rates for each group was observed.
In rats, no physical or behavioral signs of pharmacologic effects were observed during the treatment. In male rats over a perioid of 48 weeks, the mean survival rates in treated groups was greatest in the 5%, followed by the control and 1.25 or 2.5% dosage groups. However, the variations were not significant between the control and treated groups. While the female survival rates of 5%, 2.5%, and 1.25% groups were 0.875, 0.80, and 0.65 respectively, these were not statisticall and significantly different from the values observed in the control group.

BODY WEIGHT AND WEIGHT GAIN:
In mice, during the initial 61-week period, the control and treated groups grew at essentially the same rate. No significant variation in body weight were observed throughout this study between the control and treated groups of 1.25% and 2.5% dosages. However, at the end of the initial 10-week period, the 5% dosage group showed lower growth rate as compared to the control group. At 81 weeks, an increase in food consumption in the control groups was evident in the treated male groups of 2.5% and 5% dosages. Increased food consumption in the treated group of 5% dosage was accompanied by decreased body weight.
In rats, no consistent compound- or dose-related changes in growth rates were evident.

FOOD CONSUMPTION AND COMPOUND INTAKE (if feeding study):
In mice, the mean cumulative intake of SYLOID at the end of 93 weeks in the 1.25, 2.5 and 5% dietary levels was 38.45, 79.78 and 160.23 g/mouse in males, and 37.02, 72.46 and 157.59 g/mouse in females, respectively.
In rats, the mean cumulative intake of SYLOID at the end of 103 weeks for the 1.25, 2.5 and 5% dietary levels was 143.46, 179.55 and 581.18 g/rat in males, and 107.25, 205.02 and 435.33 g/rat in females, respectively.


FOOD EFFICIENCY

HAEMATOLOGY:
In mice, the mean HCT and MCV at 12 months in female mice showed a somewhat lower level in comparison with the normal group. However, there was no evidence of dose-related alteration of hematological profiles at the end of the 12- and 21-month treatments.
In rats, occasional erratic variations in hematologic profiles were observed in the treated groups: high WBC at 24 months in male groups of 1.25% dosage, and low RBC, HGB, and HCT at 24 months in the female 2.5% dosage group. The very high value for WBC in male rats of the 1.25% group looks as if it has one or two values, perhaps because of technical errors. However, no significance can be attached to the difference.

CLINICAL CHEMISTRY: In rats, no biologically meaningful changes in TA, ALB, AST, ALT, ALP, T-BL and LDH were observed, although transient differences reaching statistical significance were frequently present. No noteworthy changes related to compound ingestion were observed in any parameters of renal analyses, such as BUN, CRE and UA.

ORGAN WEIGHTS: No noted atrophy or hypertrophy of the organs in each group was sex- or dose-related.


GROSS PATHOLOGY


HISTOPATHOLOGY: NON-NEOPLASTIC:
In mice, non-neoplastic lesions were observed in the subcutis, lungs, kidneys, and liver in the treated groups. These were considered to be of no toxicological significance.

HISTOPATHOLOGY: NEOPLASTIC (if applicable):
In mice, tumours attributed to the treatment of Syloid were found in the hematopoietic organs, particularly malignant lymphoma/leuiemia, which occurred in 7/20 (35%) in the female groups of the 2.5% dosage groups (not scientifically significant). In the lungs, the frequency of adenoma/adenocarcinoma was 1/16 (6.25%) for the control, 2/17 (11.8%) for the 1.25%, 3/14 (21.4%) for the 2.5% and 3/16 (18.8%) for the 5% dosage groups of males. The incidence of the lung adenomas in females was greater than that of males. However, none of these findings were sex- or dose-related. In the liver, the correlation of hyperplasic nodules/hepato cellular carcinoma/hemangioma/fibrosarcoma in the treated groups, as compared with the control group, was relatively low.
In rats, the incidence of tumors was the greatest in the genital organs, next in the skin. The other organs showed relatively low incidence.
The occasional presence of some neoplasms did not reveal any consistent, dose-related trends in any group.


HISTORICAL CONTROL DATA (if applicable)


OTHER FINDINGS

Relevance of carcinogenic effects / potential:

The occasional presence of some neoplasms did not reveal any consistent, dose-related trends in any group when tested with Cochran-Armitage linear trend test.

The main shortcoming of the study was too small group sizes (only 20 animals/sex/group were kept to the end of the study) to discriminate small carcinogenic effects.

Conclusions:

Amorphous silica is not carcinogenic when administered orally.

Executive summary:

Takizawa et al. (1988)conducted a chronic oral study with food grade micronized silica. Micronized silica gel (Syloid) was given in a feed to B6C3F1 mice and Fisher rats at dose levels of 0, 1.25, 2.5, and 5% for 93 and 103 weeks, respectively. Each group consisted of 40 male and 40 female animals. 10 animals/sex/group were killed after 6 and 12 months and the remaining animals were reserved for 21 months. Measurements included physical examinations and observations, clinical chemistry, and post-mortem examination including histopathology. There were no biological or any other meaningful alterations in body weight, food consumption or physical features. No differences were seen in the survival of the animals between the groups. No significant dose-related effects were seen at any dose level upon clinical laboratory examinations. No gross or microscopic changes in the tissues examined. The occasional presence of some neoplasms did not reveal any consistent, dose-related trends in any group when tested with Cochran-Armitage linear trend test. The main shortcoming of the study was too small group sizes (only 20 animals/sex/dose group were kept to the end of the study) to discriminate small carcinogenic effects.

Endpoint conclusion

Endpoint conclusion:

no adverse effect observed

Carcinogenicity: via inhalation route

Endpoint conclusion

Endpoint conclusion:

no study available

Carcinogenicity: via dermal route

Endpoint conclusion

Endpoint conclusion:

no study available

Justification for classification or non-classification

Based on available information on synthetic amorphous silica, amorphous silica, including silica fume, is not carcinogenic. The impurities of silica fume include quartz, which may be present in silica fume at levels of <0.1% of respirable quartz. Respirable quartz is more relevant than total quartz in this respect. In addition, quartz is currently not classified as a carcinogen within the EU. Silicon carbide does not exist in silica fume in its fibrous, possibly carcinogenic, form.

Additional information

There are no animal studies on the carcinogenicity of silica fume. Human epidemiological data from the ferrosilicon/silicon metal industry do not show an increased incidence of cancer attributed to ultra-fine silica fumes present in furnace work.

Some data is available on the carcinogenicity of synthetic amorphous silica. A critical study is the two‑year rat and mouse feeding study with micronized silica gel (Syloid) by Takizawa et al. (1988). No increased incidence of cancer was observed in the study. Although the group sizes were small, causing uncertainty in the interpretation, it suggests a non-carcinogenicity of synthetic amorphous silica via the oral route. This conclusion is supported by negativein vitroandin vivomutagenicity data (see Chapter 'Mutagenicity').

Regarding the inhalation carcinogenesis of amorphous silica, there are no proper studies available. An early study by(1940) is not reliable due to technical deficiencies and especially due to the failure of substance identification. However, intrapleural implantation of two different preparations of pyrogenic (Cab-O-Sil) silica did not increase the incidence of tumours.

The mechanisms of quartz carcinogenicity have been widely studied, but not yet fully understood. The surface reactivity of the crystalline structure and biopersistence are important factors in the toxicology of quartz. Compared to amorphous silica, crystalline silica is practically insoluble in body fluids and its reml from lungs by macrophages is also limited (see, for example, SCOEL 2003). Persistent pulmonary inflammation is characteristic for crystalline silica exposure (see, for example, Warheit et al. 1995). Synthetic amorphous silica, on the other hand, is eliminated from lung tissue during and after prolonged inhalation exposure. It produces a transient pulmonary inflammatory response with a return of inflammatory markers to control levels after the exposure (see, for example, Warheit et al. 1995). An older comparative study on silica fume and synthetic amorphous silica (Swensson et al. 1967) suggested that silica fume resembles amorphous silica in its lung effects: pyrogenic silica and silica fume caused a tissue reaction (detected as lung and lymph node weight change and collagen formation) within one month after the administration of the particles; but the reaction did not progress in a follow-up assessment, whereas quartz and quartz glass caused a progressive tissue reaction.In vitrodissolution data show that the dissolution of synthetic pyrogenic silica and silica fume in artificial lung fluid is rather similar (see Chapter 'Toxicokinetics'). This supports the conclusion that the biopersistence of silica fume is likened to synthetic amorphous silica.

The epidemiological data from the ferrosilicon/silicon metal industry supports the non-carcinogenicity of silica fume, although it should be remembered that epidemiological studies are insensitive to recognizing small cancer risks - especially in the case of cancers which are very common, like lung cancer.

The evaluation of the OECD study (2004) concludes, on the basis of lacking positive proof and the negative findings of Takizawa et al. (1988), that there is no evidence of a carcinogenic potential arising from ingestion of synthetic amorphous silica. On the basis of an intrapleural study with sodium aluminosilicate (see OECD 2004) and negative mutagenicity data, the inhalation carcinogenicity of amorphous silica was also not considered to be a concern. No further studies on this endpoint were suggested. [Note: Our opinion is that the use of an aluminosilicate study for read across to amorphous silica is not considered appropriate in the case of lung carcinogenicity.]

International Agency for Research of Cancer evaluated silicas in 1997 (IARC 1997). It summarizes that amorphous silicas have been studied less than crystalline silica and they are generally less toxic than crystalline silica and are cleared more rapidly from the lungs. The conclusion of the IARC on amorphous silica was that there is inadequate evidence in humans and animals for the carcinogenicity of amorphous silica and amorphous silica is not classifiable as to its carcinogenicity in humans (Group 3). No data on silica fume were presented in the IARC documentation.

Thus, it is concluded that amorphous silica, including silica fume, is unlikely to be a carcinogen. However, in the assessment of the carcinogenicity of silica fume, its quartz content also has to be considered. Respirable quartz is a clear human carcinogen whose carcinogenicity is mainly related to the development of silicosis. Quartz has, however, no carcinogenicity classification under the EU harmonized classification and labelling legislation (CLP-regulation, EC 1272/2008, annex VI). Because of the silicosis, respirable quartz has been proposed by the industry to be classified to STOT cat 1 under CLP. Although bulk silica fume contain quartz, the levels of respirable quartz are well below 0.1% w/w. This means that air levels of 0.0003 mg/m³of quartz are at the proposed DNEL level of 0.3 mg/m³, which is well below the binding occupational exposure limit of respirable crystalline silica dust of 0.1 mg/m³ (EU 2017/2398). Since the respirable quartz levels are more relevant than total quartz levels when considering the cancer hazard, based on the current data the silicosis and cancer hazard due to trace respirable quartz levels of silica fume is negligible.

Silicon carbide is a possible carcinogen only in its fibrous form. No silicon carbide fibers, however, have been detected in silica fume.

Other impurities of silica fume present at levels >0.1 % in bulk material and released from silica fume in vitro include magnesium and zinc. These do not exert a carcinogenic potential.

Low-grade silica fume may contain lead at levels of up to 0.3%. Lead compounds (except some specific compounds like lead acetate and lead chromate) have not been classified as carcinogenic within the EU. However, the IARC has evaluated inorganic lead compounds as probably carcinogenic to humans (Group 2A) based on the induction of kidney tumours in rodents. The most convincing evidence comes from lead acetate (organic lead compound). Lead powder has not produced tumours in an oral or an intramuscular study or after intrarenal injection in rats. In one experiment, the inhalation of lead oxide did not produce tumours in male rats; however, in another study intratracheal instillation of combinations of lead oxide and benzo[a]-pyrene in hamsters produced lung tumours not observed with either agent alone (IARC 2006). Thus, there are some concerns about the carcinogenicity of inorganic lead compounds. However, evidence on less soluble lead oxide and lead oxide is still limited and the IARC also states that there is inadequate evidence in experimental animals for their carcinogenicity. Based on this and the current EU classification of lead compounds, no carcinogenicity classification of silica fume is needed, even in the case of low-grade silica fume containing up to 0.3% of lead.

Trace amounts of polycyclic aromatic hydrocarbons (PAH's) have also been detected in commercial silica fume. The measured levels of PAH's are, however, very minor (<100 ppm, i. e. <0.01% of silica fume, combined PAH level, see section 1.2 'Composition').