Skip to content
Heavy Metal Index

NTP Technical Report on the Toxicology and Carcinogenesis Studies of Antimony Trioxide (CASRN 1309-64-4) in Wistar Han [Crl:WI(Han)] Rats and B6C3F1/N Mice (Inhalation Studies)

Technical Report 590 is the definitive two-year inhalation carcinogenesis bioassay of antimony trioxide (CASRN 1309-64-4; Sb2O3) conducted by the National Toxicology Program (NTP) in Wistar Han rats and B6C3F1/N mice. The study was nominated by the Consumer Product Safety Commission and the National Institute of Environmental Health Sciences because of potential for substantial human occupational exposure and the absence of adequate two-year carcinogenicity data at the time of nomination. The final carcinogenicity calls are some evidence in male and female Wistar Han rats and clear evidence in male and female B6C3F1/N mice.

Study design

Two-week range-finding studies were conducted first (first exposure December 3, 2007) in groups of five males and five females per species at 0, 3.75, 7.5, 15, 30, or 60 mg/m3. Separate tissue burden groups of five females per species per concentration were held for a 28-day recovery period after the two-week exposures.

The two-year carcinogenesis studies in rats began September 22, 2008, and in mice October 6, 2008. Groups of 60 male and 60 female core study animals per species were exposed at 0, 3, 10, or 30 mg/m3. An additional 25 tissue burden females per species per concentration were exposed for up to 79 weeks and sampled at 61, 124, 269 to 271, 369, and 551 days of exposure. Ten animals per sex per group per species were evaluated at a 12-month interim necropsy. Terminal necropsies were performed September 20-23, 2010 (rats) and October 4-7, 2010 (mice); animals were approximately 110 weeks old at terminal kill. Exposure was whole-body inhalation, 6 hours plus T90 (12 minutes) per day, 5 days per week, for up to 105 weeks. The study laboratory was Battelle Toxicology Northwest, Richland, Washington.

The test article (lot 3N-06159) was a finely divided white powder of antimony trioxide greater than 99.9% pure, with arsenic (approximately 0.019%) and lead (approximately 0.016%) as the only detected trace impurities above 0.01% relative to antimony trioxide; purity was confirmed by ICP/AES, X-ray diffraction, redox titration, and UV spectrophotometry (Appendix H, TR 590). Feed (irradiated NTP-2000 wafer diet) and tap water were available ad libitum except feed was withheld during exposures.

Aerosol characterization

Aerosol concentration was monitored daily from filter samples extracted in hydrochloric acid and analyzed by ICP/AES calibrated against NIST-traceable antimony trioxide standards with yttrium internal standard. Particle size was characterized monthly by cascade impactor (Mercer-style seven-stage). Across exposure concentrations, the mass median aerodynamic diameter (MMAD) ranged from 1.3 to 1.5 micrometres with a geometric standard deviation (GSD) of 1.9 in the two-week studies, and from 0.9 to 1.5 micrometres with GSD from 1.7 to 2.2 in the two-year studies (Appendix H, TR 590 p. 10). These values are within NTP specifications requiring MMAD below 3 micrometres with GSD below 3.

Key numbers

Figures below are taken from the report as printed. Units are not converted between species or studies. The survival denominators for neoplasm incidences are 50 animals per group after removal of the 10 animals evaluated at the interim kill.

Two-year study: neoplastic effects

GroupTissue / tumour typeIncidence (0, 3, 10, 30 mg/m3)Evidence call
Male ratsLung: alveolar/bronchiolar adenoma or carcinoma (combined)3/50, 4/50, 8/50, 8/50Some evidence
Male ratsAdrenal medulla: benign pheochromocytoma1/49, 0/50, 2/49, 7/50Contributes to some evidence
Female ratsLung: alveolar/bronchiolar adenoma0/50, 2/50, 6/50, 5/50Some evidence
Female ratsAdrenal medulla: benign or malignant pheochromocytoma (combined)0/49, 2/49, 2/49, 7/50Contributes to some evidence
Female ratsLung: cystic keratinizing epithelioma or squamous cell carcinoma0/50, 0/50, 0/50, 3/50May have been related
Male miceLung: alveolar/bronchiolar carcinoma4/50, 18/50, 20/50, 27/50Clear evidence
Male miceSkin: fibrous histiocytoma or fibrosarcoma (combined)0/50, 1/50, 3/50, 4/50Related to exposure
Female miceLung: alveolar/bronchiolar adenoma or carcinoma (combined)3/50, 22/50, 27/50, 18/50Clear evidence
Female miceMalignant lymphoma (all organs)7/50, 17/50, 20/50, 27/50Clear evidence
Female miceSkin: squamous cell carcinoma0/50, 0/50, 0/50, 2/50May have been related

Source: TR 590 Summary Table (pp. xxi-xxiii) and Conclusions (p. 97).

Two-year study: selected nonneoplastic effects

GroupTissue / lesionIncidence at 30 mg/m3 vs. controlReference
Male ratsLung: chronic active inflammation50/50 vs. 18/50Table 9, p. 33
Male ratsLung: alveolar proteinosis50/50 vs. 0/50Table 9
Male ratsLung: fibrosis50/50 vs. 2/50Table 9
Female ratsLung: alveolar proteinosis50/50 vs. 0/50Table 9
Female ratsAdrenal medulla: hyperplasia8/50 vs. 0/49Table 10, p. 37
Female ratsEye: retina atrophy (all exposed groups significantly increased)19/49 vs. 6/49Table 14, p. 45
Female ratsKidney: nephropathy24/50 vs. 16/50Table 14
Male miceLung: alveolar fibrosis37/50 vs. 0/50Table 22, p. 58
Male miceLung: pleura fibrosis50/50 vs. 0/50Table 22
Female miceThymus: cellular depletion29/49 vs. 9/47Table 26, p. 68
Male miceHeart: epicardial chronic active inflammation7/50 vs. 0/50Table 27, p. 70

Two-week study: selected nonneoplastic effects

In the two-week study, all rats and all mice survived. No clinical findings related to exposure were recorded. Lung weights were significantly increased at 60 mg/m3 in male rats and at 30 and 60 mg/m3 in female rats (Table 3, TR 590 p. 23). Incidences of chronic active pulmonary inflammation were significantly increased in males and females at 30 and 60 mg/m3 (Table 4, TR 590 p. 24). In mice, squamous metaplasia of the epiglottis was significantly increased at 30 and 60 mg/m3 (Table 17, TR 590 p. 48).

Survival

In the two-year rat study, survival of 10 and 30 mg/m3 females was significantly lower than controls; the reduction was attributed primarily to lung proteinosis (Table 5, TR 590 p. 26). Final survival rates for male rats were 30/50, 30/50, 28/50, 18/50 at 0, 3, 10, 30 mg/m3; for female rats 39/50, 38/50, 28/50, 20/50. In the mouse two-year study, survival of 10 and 30 mg/m3 males and females was significantly reduced; reductions were attributed primarily to alveolar/bronchiolar carcinomas and lung inflammation in males and to malignant lymphoma and lung inflammation in females. Final survival rates for male mice were 38/50, 30/50, 27/50, 17/50; for female mice 36/50, 31/50, 26/50, 15/50 (Table 18, TR 590 p. 49).

Body weights

In male rats, mean body weights of the 30 mg/m3 group fell at least 10% below controls after week 69 and reached 80% of controls by study end. In female rats, mean body weights of the 3, 10, and 30 mg/m3 groups fell at least 10% below controls after weeks 99, 81, and 65, respectively, and the 10 and 30 mg/m3 groups were 20% and 28% below controls at study end (Tables 6 and 7, TR 590 pp. 28-29). In male mice, the 30 mg/m3 group was 10 to 25% below controls after week 73. In female mice, the 30 mg/m3 group was at least 10% below controls after week 85 (Tables 19 and 20, TR 590 pp. 52-53).

Tissue burden

Antimony trioxide lung burdens increased with both exposure concentration and duration in rats and mice. In 30 mg/m3 rats following 551 days of exposure, mean lung burdens reached approximately 35 mg Sb2O3/lung; in mice at the same concentration and duration, approximately 6 mg Sb2O3/lung. Clearance half-lives in the two-week studies ranged from 73 to 122 days in female rats and from 47 to 62 days in female mice. In the two-year study, model-estimated clearance half-lives in female rats increased with exposure concentration: 136, 203, and 262 days at 3, 10, and 30 mg/m3, respectively. Pulmonary overload calculations (volumetric basis) indicated that overload was not reached at 3 mg/m3 in either species; at 10 mg/m3, overload was calculated to begin in rats by day 418 and in mice by day 369 of the two-year study; at 30 mg/m3, overload began by day 94 (rats) and day 124 (mice) (Appendix G, TR 590).

Genetic toxicology

Antimony trioxide induced small but statistically significant increases in micronucleated erythrocytes in peripheral blood of male and female B6C3F1/N mice following 12 months of inhalation exposure. Significant increases in reticulocytes were also seen in both sexes of mice. Increased DNA damage was detected by comet assay in lung tissue samples from both sexes of mice but not in peripheral blood leukocytes. No increases in micronucleated erythrocytes, reticulocytes, or DNA damage in lung or blood were observed in male or female Wistar Han rats after 12 months of inhalation exposure (Appendix E, TR 590).

Molecular pathology

In rats, Egfr mutations were found in 13/26 (50%) alveolar/bronchiolar tumors from exposed animals and in 0 of 0 spontaneous tumors from chamber controls; Kras mutations in 1/26 (4%) exposed-animal tumors. The Egfr mutations were localized within exons 18 to 21 and were all transition mutations. In mice, Kras mutations occurred in 34/80 (43%) and Egfr mutations in 37/80 (46%) of alveolar/bronchiolar tumors from exposed animals. Kras mutations were also present in 3/9 (33%) spontaneous alveolar/bronchiolar carcinomas from chamber controls, and the increase in Kras mutations in exposed mice did not achieve statistical significance. Egfr mutations occurred only in tumors from exposed mice and were located primarily within exons 18 and 20 (Appendix K, TR 590).

Methods (brief)

Whole-body inhalation, 6 hours plus T90 per day, 5 days per week. Test article purity greater than 99.9% (lot 3N-06159), confirmed by ICP/AES, XRD, redox titration, and UV spectrophotometry. Aerosol generation used a linear feed device and Trost jet mill with a cyclone separator. Aerosol concentration monitored daily by ICP/AES on hydrochloric acid extracts from filter samples. Particle size characterized monthly by cascade impactor; MMAD 0.9 to 1.5 micrometres, GSD 1.7 to 2.2 in the two-year studies. Chamber temperature 75 degrees F plus or minus 3 degrees F; relative humidity 55% plus or minus 15%; 15 plus or minus 2 air changes per hour. Animals housed individually; chambers rotated weekly; feed analyzed for contaminants (Appendix I, TR 590).

Histopathology: complete necropsies on all core study animals; full tissue list evaluated microscopically (Table 1, TR 590 pp. 16-18). Pathology reviewed by laboratory pathologist, independent QA laboratory, and NTP Pathology Working Group. For the two-year studies, the QA pathologist evaluated all tumors and all potential target organs, with PWG review of discrepancies.

Statistical methods: survival by Kaplan-Meier; pairwise and trend tests by Cox and Tarone methods. Lesion incidences by Poly-3 survival-adjusted test (Poly-k method, k=3). Continuous variables by Dunnett and Williams procedures. Neoplastic and nonneoplastic incidences at interim evaluation assessed by Fisher exact test.

Tissue burden: lung and blood samples digested in nitric acid and hydrochloric acid by microwave; lung antimony measured by Thermo Elemental IRIS Intrepid II ICP/AES; blood antimony measured by Agilent 7500ce ICP-MS. Genetic toxicology in peripheral blood and lung assessed by micronucleus assay and alkaline comet assay at the 12-month interim evaluation. Molecular pathology: DNA extracted from formalin-fixed, paraffin-embedded tumor blocks; Kras (exons 1 and 2) and Egfr (exons 18 to 21) amplified by seminested PCR and sequenced.

Limit of detection and limit of quantitation: not reported in this Technical Report. Certified reference materials: not specifically described; ICP/AES calibration used NIST-traceable antimony trioxide standards with yttrium internal standard.

Evidence fitness

This source supports the level-of-evidence carcinogenicity calls and all associated incidence data as an NTP two-year bioassay conducted under GLP and subject to retrospective QA audit. It supports the tissue burden and pulmonary overload calculations, the genetic toxicology findings, and the Kras and Egfr mutation data as reported in the respective appendices. It does not support an oral or dermal potency estimate, a concentration in food or a consumer product, or a certification threshold. The route is occupational inhalation; human risk extrapolation requires analyses beyond the scope of this report (NTP foreword, TR 590).

Limitations

Rodent inhalation study only; not a consumer-product or dietary exposure study. Pulmonary overload at the two higher concentrations confounds dose-response interpretation at those levels; the study authors discuss this in the context of clearance half-life extension and in the Discussion section of TR 590. The micronucleus and comet results in mice were noted by one peer reviewer as requiring context from historical control data. The spontaneous alveolar/bronchiolar carcinoma rate in B6C3F1/N mice introduces background; the high spontaneous Kras mutation rate in control-mouse tumors (3/9) reduces the statistical power for detecting a concentration-related increase in Kras mutations in exposed mice. LOD and LOQ values for the analytical methods are not reported in the body of the report.

Regulatory context

arsenic (As) background cited by the report (Introduction, TR 590 p. 3): the occupational exposure limit for antimony was 0.5 mg/m3 (8-hour TWA threshold limit value, ACGIH; NIOSH 10-hour TWA recommended exposure limit; OSHA 8-hour TWA permissible exposure limit). The EPA inhalation reference concentration for antimony trioxide was 0.2 micrograms/m3 at the time of the report. The EPA drinking water maximum contaminant level for antimony was 6 micrograms/L. These values are cited from the introduction of TR 590, not independently verified here.

Update history

No substantive edit history is available in this build. The full commit record is available in git.