Abrasive blasting involves forcibly projecting a stream of abrasive particles through compressed air or steam against a surface to change its quality or to remove contaminants. Blasting sand, the most commonly used abrasive blasting agent, contains high levels of crystalline silica (SiO₂) and can cause pulmonary fibrosis (silicosis) following inhalation exposure. Crystalline silica is also considered to be a lung carcinogen. Acute silicosis is characterized by alveolar proteinosis with reduced gas exchange, whereas chronic silicosis is characterized by scarring and formation of fibrotic nodules around the trapped silica particles. Alternatives to blasting sand with lower crystalline silica content exist, including specular hematite, which is mostly iron oxide (Fe₂O₃). Exposure to specular hematite via in vivo intratracheal instillation has been shown to induce less lung injury, inflammation, and fibrosis than blasting sand.¹ Other alternatives to blasting sand include coal slag, crushed glass, and garnet; however, no comprehensive acute or chronic inhalation studies have been performed to evaluate the health effects, including pulmonary toxicity, of these alternative compounds.

Studies evaluating exposure risks are needed due to the high production volume of these compounds, the number of workers exposed, and the inadequacy of available toxicity data to determine safe exposure concentrations. Acute inhalation toxicity testing to compare blasting sand to coal slag, crushed glass, garnet, and specular hematite was initially performed in male rats via whole-body exposure in separate, short-term 2-week studies. The main objectives of the 2-week studies were to determine acute toxicity, identify target organs, evaluate tissue burden, and provide a basis for selection of test article and exposure concentrations to be used in subsequent 39-week inhalation studies. Data generated from these studies might inform subsequent recommendations from the National Institute for Occupational Safety and Health for alternative abrasive compounds to blasting sand.

The lung toxicity of blasting sand, which should have served as a positive control in the 2-week studies due to the known adverse effects of silica sand, was low after the 2-week exposure window, a result likely attributable to low lung burden (steady-state lung burden was not achieved by the end of the study). In fact, all abrasive compounds (blasting sand and alternatives) tested in the 2-week studies exhibited low lung toxicity, perhaps due to low lung burdens.

The lung histopathology from the 2-week studies indicated that specular hematite and crushed glass appeared to be the least toxic in causing lung inflammation of the four alternative blasting agents tested; however, crushed glass exhibited the shortest clearance half-lives (10, 17, and 16 days for the 3, 15, and 30 mg/m³ groups, respectively), resulting in the lowest lung burdens. Thus, due to the relatively fast clearance rate of crushed glass, the 39-week study design favored comparison of blasting sand with specular hematite, which had more similar clearance half-lives (95, 35, and 33 days for blasting sand and 48, 30, and 50 days for specular hematite for the 3, 15, and 30 mg/m³ groups, respectively). Garnet, which contained approximately 6% crystalline silica and exhibited the longest clearance half-lives (90 and 89 days for the 15 and 30 mg/m³ groups, respectively)—resulting in the highest steady-state lung burdens (5,089 and 9.089 μg garnet/total lung for the 15 and 30 mg/m³ groups, respectively)—was the most toxic of the alternative blasting agents in regard to the incidence of chronic active inflammation in the lung (five of five animals at the 15 and 30 mg/m³ exposure concentrations), followed by coal slag (despite the absence of crystalline silica). This finding suggests that the presence of crystalline silica was not the only determinant of lung toxicity in the 2-week studies because the coal slag tested contained no crystalline silica, yet there was some incidence of focal inflammation (two of five animals at the 30 mg/m³ exposure concentration) and proteinosis (two of five animals at the 15 mg/m³ exposure concentration) in the lungs of coal slag-exposed rats. Garnet was the most toxic of all the alternative blasting agents tested in the 2-week studies, perhaps because it contained the most crystalline silica of all the alternatives or had the longest clearance half-lives (for 15 and 20 mg/m³ exposure concentrations) relative to the other abrasive blasting agents. However, in terms of lung inflammation, garnet was more toxic than blasting sand (which contained the highest levels of crystalline silica of all the blasting agents), suggesting that the toxicity of garnet was more likely due to longer clearance half-lives (90 and 89 days for garnet versus 35 and 33 days for blasting sand for the 15 and 30 mg/m³ groups, respectively). Crushed glass also appeared to be more reactive than the other blasting agents to the upper respiratory tract (larynx) in causing hyperplasia, squamous metaplasia, and inflammation of the epiglottis. Informed by upper airway and lung histopathology, as well as clearance data from the 2-week studies, specular hematite was chosen as the alternative test article (instead of crushed glass) to compare with blasting sand in separate 39-week studies.

The objectives of the long-term inhalation studies were to assess the pulmonary toxicity, fibrogenicity, tissue (lung and lymph node) burden, and immunotoxicity in rats after chronic exposure to blasting sand or specular hematite for up to 39 weeks. The crystalline silica content of these two abrasive blasting agents differ greatly, with blasting sand consisting of greater than 75% crystalline silica and specular hematite consisting of only 1% to 2% crystalline silica (greater than 95% iron oxide). Thus, these studies may help to further address the chronic lung effects from inhalation exposure to crystalline silica. There were no treatment-related changes in survival, body weights, or clinical observations for either test article. There was a significant exposure concentration-dependent increase in relative lung and bronchial lymph node weights for both blasting sand and specular hematite compared to the chamber control groups at week 39. When comparing the two test articles, relative lung and bronchial lymph node weights at week 39 (60 mg/m³) were increased for blasting sand compared to specular hematite, but the differences were not statistically significant.

Significant exposure concentration-dependent increases were observed in total cells counted, absolute neutrophils and lymphocytes, LDH activity, and MCP-1 protein levels in bronchoalveolar lavage (BAL) fluid at week 39 compared to chamber control groups for both blasting sand and specular hematite. Total cells counted and absolute macrophages in BAL fluid at week 39 (60 mg/m³) were significantly increased for blasting sand compared to specular hematite (Table A-3). Absolute neutrophils in BAL fluid at week 39 (60 mg/m³) were also greater for blasting sand compared to specular hematite, but the difference was not statistically significant. Neutrophils and lymphocytes, however, were significantly increased for blasting sand compared to specular hematite at earlier time points and for other exposure concentrations (Table A-3).

Furthermore, LDH activity in BAL fluid at week 26 was significantly increased for blasting sand compared to specular hematite at all exposure concentrations (Table A-3). These data suggest that markers of airway/lung injury and inflammation were increased in animals exposed to blasting sand compared to specular hematite. Except for significantly increased MCP-1 in BAL fluid for both blasting sand and specular hematite, there were minimal immunotoxic effects. MCP-1 is a potent pro-inflammatory chemokine that recruits leukocytes (including macrophages) from the circulation into the lung.⁷³ Resident and recruited macrophages are critical phagocytic cells within the lung, acting in response to particle inhalation.

Regarding histopathological changes in the airways and lung, respiratory epithelial hyaline droplet accumulation in the nose was less apparent in rats exposed to specular hematite compared to those exposed to blasting sand. The incidence of respiratory epithelial hyaline droplet accumulation in the nose at week 16 (15 and 60 mg/m³ groups) was significantly lower for specular hematite compared to blasting sand (Table 14). Lung histopathology showed that the incidences of chronic active inflammation and interstitial fibrosis were both less in rats exposed to specular hematite compared to those exposed to blasting sand for several exposure concentrations and time points. The incidence of chronic active inflammation in the lung at weeks 16 (60 mg/m³), 26 (all exposure concentrations), and 39 (15 and 30 mg/m³) was significantly lower for specular hematite compared to blasting sand (Table 14). The incidence of interstitial (alveolar) fibrosis in the lung at weeks 26 (all exposure concentrations) and 39 (15 mg/m³) was significantly lower for specular hematite compared to blasting sand. In addition, alveolar proteinosis occurred in the lungs of rats exposed to 60 mg/m³ blasting sand but not specular hematite at week 39 (Table 14). This important finding should be emphasized because acute silicosis includes alveolar lipoproteinosis, which contributes to reduced gas exchange, and specular hematite did not induce alveolar proteinosis in this 39-week study. Crystalline silica has been shown to be cytotoxic to alveolar macrophages.⁷⁴ A crystalline silica–induced decrease in resident alveolar macrophages can impair the ability of the lung to clear surfactant lipoproteins resulting in alveolar proteinosis.⁷⁵ However, alveolar epithelial hyperplasia occurred in the lung of rats exposed to 30 and 60 mg/m³ specular hematite but not blasting sand at week 16. Also, in the larynx, the incidence of squamous metaplasia of the epiglottis at weeks 4 (30 and 60 mg/m³) and 16 (60 mg/m³) and at weeks 26 and 39 (all exposure concentrations) was greater in rats exposed to specular hematite but not blasting sand. Exposure-related histopathological effects in the airways and lung were mostly observed in terms of incidence because the overall severity grades for changes in histopathological parameters were low (minimal to mild) in the 39-week studies. Blasting sand should have acted as a positive control in the 39-week studies (as in the 2-week studies), and thus we expected greater severity scores for lung histopathological parameters (in particular, interstitial fibrosis) in the 39-week blasting sand study.

Table 14

Comparisons of the Incidences of Selected Nonneoplastic Lesions in Male Sprague Dawley Rats in the 39-week Inhalation Studies of Blasting Sand and Specular Hematite^a.

Except for 15 mg/m³ specular hematite, lung overload was achieved by week 39 for both blasting sand and specular hematite at all exposure concentrations; however, blasting sand was calculated to achieve lung overload conditions in less time compared to specular hematite (55 and 30 days for blasting sand versus 118 and 55 days for specular hematite for the 30 and 60 mg/m³ groups, respectively). Marginal differences between blasting sand and specular hematite relating to lung toxicity and fibrogenicity might have been attributable to the differences in the time of onset of lung overload; however, specular hematite is denser than blasting sand, which may have affected its volume occupied in the lung tissue and/or its solubility within the lung and thus toxicity/fibrogenicity. Future studies are needed to evaluate the relative solubility of blasting sand and specular hematite within the lung.

The data from the 39-week studies suggest that specular hematite could be, to some extent, less toxic and fibrogenic to the lower respiratory tract, especially concerning the induction of alveolar proteinosis within the lung, which was absent in all rats exposed to specular hematite. These adverse lung effects might have been due to the lower crystalline silica content, lower lung burden (increased time of onset of overload), or reduced solubility of specular hematite compared with blasting sand. Specular hematite was also more reactive than blasting sand to the upper respiratory tract (larynx) in causing squamous metaplasia of the epiglottis. The 39-week studies suggest that specular hematite could be a slightly safer alternative to blasting sand for abrasive sandblasting endeavors because of blasting sand’s potential to induce lung injury, chronic inflammation, alveolar proteinosis, and interstitial fibrosis. However, both test articles exhibited some degree of lung toxicity and fibrogenicity over 39 weeks of exposure and, unlike blasting sand, specular hematite also exhibited toxic effects on the larynx.

Under the conditions of these 39-week inhalation studies, the major target tissue in male Sprague Dawley rats exposed to blasting sand or specular hematite was the lung. The incidences of chronic active inflammation and interstitial fibrosis were significantly lower in rats exposed to specular hematite (compared with blasting sand) at some time points under some exposure conditions. After 39 weeks of exposure to specular hematite, the lowest-observed-effect level was 15 mg/m³ for chronic active inflammation and interstitial fibrosis within the lung. Alveolar proteinosis was present at week 39 in the lungs of rats exposed to the highest concentration (60 mg/m³) of blasting sand but was notably absent in the lungs of rats exposed to specular hematite. Alveolar epithelial hyperplasia was present at week 16 in the lungs of rats exposed to the two highest concentrations (30 or 60 mg/m³) of specular hematite but not blasting sand. Specular hematite exhibited the potential to be an inhalation toxicant in workers exposed via abrasive blasting operations but to a lesser degree than blasting sand because the lungs of rats exposed to specular hematite showed a lower incidence of interstitial fibrosis and an absence of alveolar proteinosis.