Procurement and Characterization

The blasting sand (coarse silica sand #2340) used in the 2-week and 39-week studies was obtained from Waupaca Sand and Solutions (Division of Faulks Brothers Construction Inc.; Waupaca, WI) Midwest Research Institute (MRI; Kansas City, MO) in one lot by Midwest Research Institute (MRI; Kansas City, MO) and was assigned the lot number W100604JB. The micronized specular hematite (Barshot 50) used in the 2-week and 39-week studies was obtained from Opta Minerals, Inc. (Waterdown, ON, Canada) in one lot (0101005CJ) by MRI (Kansas City, MO). The coal slag used in the 2-week study was obtained from Reed Minerals-Harsco Corporation (LaCygne, KS) MRI in one lot (R042805KA). The crushed glass (VitroGrit^TM #30/50) used in the 2-week study was obtained from TriVitro Corporation (Kent, WA) MRI in one lot (T092205KA). The garnet used in the 2-week study was obtained from Emerald Creek Garnet Ltd. (Fernwood, ID) MRI in one lot (031605).^a Aveka, Inc. (Woodbury, MN) reduced the particle size of all five test articles using aqueous ball milling followed by aqueous bead milling.^b Identity and purity analyses were conducted by multiple analytical chemistry laboratories and the study laboratory at Battelle Toxicology Northwest (Richland, WA) (Appendix C). Reports on analyses performed in support of the abrasive blasting agent studies are on file at the National Institute of Environmental Health Sciences (NIEHS).

Aerosol Generation and Exposure Systems

For the 2-week studies of abrasive blasting agents, the aerosol generation system consisted of a linear feed dust-metering device designed and built by Battelle to meter the abrasive blasting agent from a reservoir into an air stream for aerosolization. Within the metering device, periodic blasts of compressed air suspended small volumes of blasting sand, coal slag, crushed glass, garnet, or specular hematite in the air stream for transport to the metering device exhaust tube. For blasting sand, coal slag, and garnet, a jet disperser was positioned immediately downstream from the metering device exhaust tube (Figure C-6, Figure C-7). Coal slag and garnet were moved from the jet disperser to a particle attrition chamber (PAC) to enhance the aerosolization of the test material. Crushed glass and specular hematite were processed in a Trost jet mill (Garlock, Inc., Newtown, PA), used downstream from the metering device exhaust tube, to perform initial particle size reduction; opposing compressed air gas streams drove the jet mill (Figure C-8, Figure C-9).

All generation system components were housed in a glove box in the control center room. From the jet disperser (blasting sand, coal slag, garnet) and jet mill (crushed glass, specular hematite), aerosolized blasting agents were blended with filtered, compressed air before being conveyed down the distribution line from the control center room to the exposure room. For crushed glass, as the air stream entered the exposure room, an in-line cyclone separator further decreased particle size and extracted nonrespirable aerosol. For blasting sand (2-week and 39-week studies), coal slag, crushed glass, garnet, and the 39-week study of specular hematite (Figure C-10), all chambers in the exposure room except for the control were fed aerosol from a single distribution line constructed of stainless steel, bonded and grounded to prevent electrostatic charge buildup. For the 2-week study of specular hematite, the distribution line was split into north and south branches. Aerosol was supplied to the 60 mg/m³ chamber from the south distribution line branch; all remaining chambers in the exposure room except the control chamber were fed aerosol from the north distribution line branch. During exposures to abrasive blasting agents, the airflow through the distribution line was controlled using a house vacuum regulated by a filter-protected flow meter. A second distribution line flow control system was available during off-exposure periods. This system consisted of a vacuum transducer pump (Air-Vac Engineering Company, Inc., Seymour, CT) of higher flow capacity, positioned in parallel with the flow meter control assembly, and was operational only during critical shutdown periods. At each exposure chamber, aerosol was delivered from the distribution line by a sampling tube. The flow through each sampling tube was induced by a stainless-steel air ejector pump designed and fabricated by Battelle. The flow rate and configuration of the ejector pump and sampling tube combination were chosen to optimize the efficiency of the delivery system. The aerosol then entered the chamber inlet duct where it was further diluted with humidified, Parafil-, charcoal-, and high-efficiency particulate air (HEPA)-filtered air to achieve the desired exposure concentration.

The 39-week study of blasting sand used the same aerosol generation system described for the 2-week study of this test material. For the 39-week study of specular hematite, the aerosol generation system was similar to that described for the 2-week study of blasting sand except for the additions of an in-line settling jar within the glove box and an in-line cyclone separator in the distribution line to the exposure room as described for the 2-week study of crushed glass.

The study laboratory designed the inhalation exposure chamber (Harford Systems Division of Lab Products, Inc., Aberdeen, MD) so that uniform aerosol concentrations could be maintained throughout the chamber with the catch pans in place. The total active mixing volume of each chamber was 1.7 m³.

Aerosol Concentration Monitoring

Summaries of the chamber aerosol concentrations are given in Table C-1 and Table C-2. The concentration of the abrasive blasting agent in the exposure chambers and room air was monitored using two real-time aerosol monitors (RAMs) (Model RAM-1; MIE, Inc., Bedford, MA). The monitors were connected to the chambers by a sampling system designed by Battelle incorporating a valve that multiplexed each RAM to a 0 mg/m³ chamber or the room, a HEPA-filtered room air blank, and two exposure chambers. The output (voltage) of the RAM was recorded by a program designed by Battelle (Battelle Exposure Data Acquisition and Control) to select the correct sample stream and acquire a raw voltage signal from each RAM. Equations for the calibration curves resided within the program and were used to convert the measured RAM voltages to exposure chamber concentrations.

Each RAM was calibrated by constructing a response curve using the measured RAM voltages (voltage readings were corrected by subtracting the RAM zero-offset voltage from measured RAM voltages) and chamber concentrations of the abrasive blasting agents measured gravimetrically or specific to the test article on exposure chamber filters. Developmental studies demonstrated that gravimetric and test article-specific measurements of chamber concentrations were comparable. For all abrasive blasting agents, exposure chamber atmosphere samples were collected each day on 25 mm Pallflex^® Emfab™ TX40H120WW Teflon^®-coated, glass-fiber filters and on 25 mm, 0.45 μm GH Polypro polypropylene filters (both obtained from Pall Corporation, Ann Arbor, MI). Test article-specific assays of blasting sand, coal slag, crushed glass, and garnet measured the amount of Si captured on filters extracted with 1:3 HNO₃:HF, and used an ICP/AES method. Test article-specific assays of specular hematite measured the amount of Fe captured on filters extracted with HCl, using an ICP/AES system.

The ICP/AES instrument was calibrated against serially diluted NIST-traceable spectrometric standards Si (for blasting sand, coal slag, crushed glass, and garnet) or Fe (for specular hematite) and the internal standard Co. Quality control standards and a reagent blank were analyzed after calibration, after approximately every 10th sample, and at the end of the analysis to determine accuracy and calibration drift during analysis.

Chamber Atmosphere Characterization

Particle size distribution was determined once before the 2-week and 39-week studies began, once during the 2-week studies, and once a month during the 39-week studies. Samples were taken from each exposure chamber using a Mercer-style seven-stage cascade impactor (In-Tox Products, Moriarty, NM). For the 2-week studies of blasting sand, coal slag, crushed glass, and garnet, impactor samples were collected on polypropylene filters (GH Polypro, Pall Corporation), dissolved using HNO₃, HF, and H₃BO₃, and hydroxylamine hydrochloride (NH₂OH·HCl), and assayed for Si using ICP/AES. For the 2-week study of specular hematite, impactor samples were collected on glass slides lightly coated with silicone to reduce particle bounce or on glass-fiber filters (Pallflex Emfab, Pall Corporation), dissolved using HCl and sonication, and analyzed for Fe using ICP/AES. For the 39-week studies, impactor samples of blasting sand and specular hematite were collected on stainless-steel slides or glass-fiber filters (Pallflex Emfab, Pall Corporation) and then measured gravimetrically to determine the amount of test article deposited on each stage. The relative mass of each abrasive blasting agent collected on each stage was analyzed by the NEWCAS impactor analysis program developed at Battelle and was based on probit analysis.⁵⁰ The mass median aerodynamic particle diameter and the geometric standard deviation estimates of each set of samples are given in Table C-3, Table C-4, and Table C-5. All values of mass median aerodynamic diameter were less than 3 μm as required by the protocol (Table 1).

Table 1

Particle Size Distribution in Chambers During the Inhalation Studies of Abrasive Blasting Agents.

Buildup and decay rates for chamber aerosol concentrations were determined with and without animals present in the chambers. At a chamber airflow rate of 15 air changes per hour, the theoretical value for the time to achieve 90% of the target concentration after the beginning of aerosol generation (T₉₀) and the time for the chamber concentration to decay to 10% of the target concentration after conclusion of aerosol generation (T₁₀) was approximately 9.4 minutes. For the 2-week study of blasting sand, T₉₀ and T₁₀ values ranged from 9 to 10 minutes with animals present. For the 2-week studies of coal slag, crushed glass, garnet, and specular hematite, T₉₀ values ranged from 10 to 13, 12 to 14, 13 to 15, and 9 to 22 minutes, respectively, with animals present; T₁₀ values ranged from 9 to 10, 9 to 10, 10 to 11, and 10 to 11 minutes, respectively. For the 39-week study of blasting sand, T₉₀ values ranged from 12 to 13 minutes without animals present and from 13 to 14 minutes with animals; T₁₀ values ranged from 8 to 9 minutes without animals present and from 10 to 11 minutes with animals. For the 39-week study of specular hematite, T₉₀ values ranged from 12 to 13 minutes without animals present and from 10 to 12 minutes with animals; T₁₀ values were 9 minutes without animals present and ranged from 10 to 11 minutes with animals. A T₉₀ value of 12 minutes was selected for all studies.

The uniformity of aerosol concentration in the inhalation exposure chambers without animals present was evaluated before the 39-week studies began; in addition, concentration uniformity with animals present in the chambers was measured once during the 2-week studies and three times during the 39-week studies. Chamber concentration uniformity was maintained throughout the studies.

The persistence of the abrasive blasting agents in the chambers after aerosol delivery ended was determined by monitoring the concentration overnight in the 30 mg/m³ chambers, except for the 2-week and 39-week studies of specular hematite that monitored concentrations in the 60 mg/m³ chamber, with (all studies) and without (39-week studies only) animals present in the chambers. In the 2-week studies of blasting sand, coal slag, crushed glass, garnet, and specular hematite, the concentration decreased to 1% of the starting concentration within 19, 19, 21, 20, and 21 minutes, respectively. In the 39-week study of blasting sand, the concentration decreased to 1% of the starting concentration within 20 minutes with animals present and within 19 minutes without animals. In the 39-week study of specular hematite, the concentration decreased to less than 1% of the starting concentration within 21 minutes with animals present and within 18 minutes without animals.

Stability studies of the test materials in the generation and exposure systems were performed by the analytical chemistry and study laboratories. During the 2-week studies, before the start of the 39-week studies, and twice during the 39-week studies, blasting sand, coal slag, crushed glass, garnet, or specular hematite powder samples were taken from the low and high exposure concentration chambers and the aerosol distribution lines by collection on 25 mm A/E glass-fiber or polypropylene (GH Polypro) filters (Pall Corporation). On each sample collection day, samples of the bulk test material were collected before filling the generator reservoir and from the reservoir at the end of the generation day; additional test material was added to the generator each day. Samples were analyzed by XRD to identify and quantitate crystalline phases present in each abrasive blasting agent and by ICP/AES and PIXE (2-week studies of blasting sand, coal slag, crushed glass, and garnet) to determine elemental content, and carbon content was assayed by combustion (coal slag only). Results of these stability assays showed that the composition of each abrasive blasting agent in the exposure chambers and distribution lines was stable in the presence and absence of animals, reflected the composition of the bulk test material in the generator reservoir, and was generally comparable to that found during the initial characterization assays of each test article.

Animal Source

Male F344/NTac rats were obtained from the commercial colony at Taconic Farms, Inc. (Germantown, NY) for use in the 2-week studies, and male and female Sprague Dawley (Hsd:Sprague Dawley^® SD^®) rats were obtained from Harlan Laboratories, Inc. (Livermore, CA) for use in the 39-week studies. For many years, the National Toxicology Program (NTP) used the inbred F344/N rat for its toxicity and carcinogenicity studies. Over time, the F344/N rat strain began exhibiting sporadic seizures and idiopathic chylothorax and consistently disproportionate high rates of mononuclear cell leukemia and testicular neoplasia. Because of these issues in the F344/N rat and NTP’s desire to find a more fecund rat model that could be used in both reproductive and carcinogenesis studies for comparative purposes, an alternative rat model for use in these studies was explored. Following a workshop in 2005, the F344 rat from the Taconic commercial colony (F344/NTac) was used for a few NTP studies between 2005 and 2006 to allow NTP time to evaluate different rat models.⁵¹ NTP now uses the Sprague Dawley (Hsd:Sprague Dawley^® SD^®) rat, which it obtains from Envigo (Indianapolis, IN).

Animal Welfare

Animal care and use were in accordance with the Public Health Service Policy on Humane Care and Use of Animals (Appendix E). All animal studies were conducted in an animal facility accredited by AAALAC. Studies were approved by the Battelle Toxicology Northwest Animal Care and Use Committee and conducted in accordance with all relevant NIH and NTP animal care and use policies and applicable federal, state, and local regulations and guidelines.

Two-week Studies

On receipt, rats were 3 weeks old. Animals were quarantined for 11 or 12 days and were 5 weeks old on the first day of the studies. Before the studies began, five male rats were randomly selected for parasite evaluation (pinworms: Syphacia obvalata and muris) and gross observation for evidence of disease. Serology testing was not conducted at the laboratory for the 2-week studies, but rats were obtained from a commercial colony free of the following rat pathogens: Sendai virus, pneumonia virus of mice, sialodacryoadenitis virus, Kilham rat virus, Toolan’s H1 virus, Mycoplasma pulmonis and Pneumocystis carinii.

Groups of five male F344/NTac rats were exposed by whole-body inhalation to blasting sand, coal slag, crushed glass, or garnet aerosol at concentrations of 0, 3, 15, or 30 mg/m³ or specular hematite aerosol at concentrations of 0, 3, 15, 30, or 60 mg/m³ for 6 hours plus T₉₀ (12 minutes) per day, 5 days per week for 2 weeks, plus 2 days for 12 exposures (day 16) (the term “2-week studies” specifically refers to those during which animals were exposed for 2 weeks plus 2 days). Additional groups of 35 male F344/NTac rats were exposed to the same concentrations of blasting sand, coal slag, crushed glass, garnet, or specular hematite for tissue burden analysis through day 16. These test articles were not exposed simultaneously but in separate studies. Feed was available ad libitum except during exposure periods; water was available ad libitum. Rats were housed individually. Clinical observations were recorded daily. Core-study animals were weighed initially, on days 6 and 13, and at the end of the studies. Details of the study design and animal maintenance are summarized in Table 2.

Table 2

Experimental Design and Materials and Methods in the Inhalation Studies of Abrasive Blasting Agents.

Exposure concentrations were informed by inhalation studies of silica quartz by NIOSH, which exposed male F344 rats at 15 or 20 mg/m³. In these studies, rats developed lung fibrosis after exposure to 15 mg/m³ silica over 16 calendar weeks.^1,5 The selected concentrations also accounted for the estimated lung deposited doses for a 45-year working lifetime at the Occupational Safety and Health Administration permissible exposure limit (PEL) and American Conference of Governmental Industrial Hygienists threshold limit value (TLV). Concentrations of 30 and 60 mg/m³ were expected to produce lung overload in rats, which is necessary to compare effects of overload conditions with the effects seen in nonoverloaded lungs and to obtain deposited lung doses in rats comparable with those estimated for humans over a full working lifetime.

Five pre-assigned tissue burden rats per exposure group were wiped clean, weighed, and anesthetized using 70% carbon dioxide on days 1, 5, 12, and 16 after the 6-hour exposure; days 8 and 15 before the 6-hour exposure; and on day 37 after 21 days of recovery. Paired lung and lymph nodes (bronchial and mediastinal) were removed, weighed, and stored separately in plastic containers at approximately −70°C until analysis. For determination of tissue concentrations, lung samples were acid digested using microwave sample preparation systems and analyzed using ICP/AES for either Si (blasting sand, coal slag, crushed glass, or garnet studies) or Fe (specular hematite study). Total test article burden was calculated using the percent Si (39.4%, blasting sand; 21.8%, coal slag; 31.2%, crushed glass; or 16.7%, garnet) or Fe (69.4%, specular hematite) found during preliminary bulk analysis of the test articles.

Necropsies were performed on all core study rats on day 16. Tissues for microscopic examination were harvested, fixed and preserved in 10% neutral buffered formalin, processed and trimmed, embedded in paraffin, sectioned to a thickness of 4–6 μm, and stained with hematoxylin and eosin (H&E). The lung and mediastinal and bronchial lymph nodes were weighed, and histopathological examinations were performed on selected tissues. Table 2 lists the tissues and organs examined.

Thirty-nine-week Studies

On receipt, the rats were 4 to 5 (blasting sand) or 4 (specular hematite) weeks old. Animals were quarantined for 11 or 12 days and were 5 to 7 weeks old on the first day of the studies. Before the studies began, five male and five female rats were randomly selected for parasite evaluation and gross observation for evidence of disease. The health of the animals was monitored during the studies according to the protocols of the NTP Sentinel Animal Program (Appendix E). All results were negative.

Test article and exposure concentrations selected for the 39-week studies were informed by the results of the 2-week studies. Groups of 32 male Sprague Dawley rats were exposed by whole-body inhalation to blasting sand or specular hematite aerosols at concentrations of 0, 15, 30, or 60 mg/m³ for 6 hours plus T₉₀ (12 minutes) per day, 5 days per week for up to 39 weeks (the term “39-week studies” refers specifically to those during which animals were exposed for up to 39 weeks). Additional groups of 30 male Sprague Dawley rats were exposed to the same concentrations of blasting sand or specular hematite for up to 39 weeks for tissue burden studies. Groups of 32 female Sprague Dawley rats were exposed to the same concentrations of blasting sand or specular hematite for up to 27 weeks for immunotoxicity studies. These test articles were not dosed simultaneously but in separate studies. Feed was available ad libitum except during exposure periods; water was available ad libitum. Rats were housed individually. For males, body weights were recorded initially, then body weights and clinical observations were recorded weekly beginning on day 8 for 16 weeks, monthly thereafter, and at the end of the studies. Rats were euthanized at interim time points or at the end of the study (39 weeks) by intraperitoneal injection of pentobarbital. Details of the study design and animal maintenance are summarized in Table 2. Information on feed composition and contaminants is presented in Appendix D.

For bronchoalveolar lavage (BAL) fluid studies, two BAL fluid washes were collected in succession from the right lung lobes of eight core study male rats pre-assigned from each exposure group at 4, 16, 26, and 39 weeks (end of studies) and from the whole lung of eight special study female rats from each exposure group at 4 or 26 weeks. Each wash sample was centrifuged separately. Lactate dehydrogenase activity and albumin concentration were measured in the supernatant of the first lavage by the study laboratory, and then the lavage supernatants were combined for each animal and frozen at −70°C until shipment to Virginia Commonwealth University (VCU) for analysis. The cells from both lavages were combined, washed in Hank’s balanced salt solution (HBSS), recentrifuged, and resuspended in approximately 1 mL HBSS for cell count, viability, and differential cell count determinations.

For lung and lymph node burden analysis, four or five tissue burden study male rats were pre-assigned from their cages at 1, 4, 8, 16, 26, and 39 weeks and wiped to remove excess test material. Rats were weighed, and the lung and lymph nodes (bronchial and mediastinal) were removed, weighed, processed, and analyzed for Si (blasting sand study) or Fe (specular hematite study) concentrations as described in the methods for the 2-week studies.

For the immunotoxicity studies, BAL fluid and blood were collected from eight core study males at 4, 16, 26, and 39 weeks and eight unimmunized special study females at 4 and 26 weeks. Serum was prepared at the study laboratory, and the BAL fluid and serum samples were frozen at −70°C and shipped on dry ice to VCU for analyses. In addition, spleens from the unimmunized special study females were collected and weighed, placed into tubes containing medium, and shipped on ice to VCU for next-day cell preparation. Serum and spleens were similarly collected from additional groups of eight immunized special study females at weeks 5 and 27 (females at 5 and 27 weeks had received tail vein injections of sheep red blood cells 4 days earlier) and shipped to VCU for analyses. Details of the immunotoxicity studies are presented in Appendix F. The parameters evaluated are listed in Appendix F.

Necropsies were performed on the male rats used for BAL studies at 4, 16, 26, and 39 weeks. Tissues for microscopic examination were harvested, fixed and preserved in 10% neutral buffered formalin, processed and trimmed, embedded in paraffin, sectioned to a thickness of 4 to 6 μm, and stained with hematoxylin and eosin. The lung as well as mediastinal and bronchial lymph nodes were weighed, and histopathological examinations were performed on selected tissues. Table 2 lists the tissues and organs examined.

After a review of the laboratory reports and selected histopathology slides by a quality assessment (QA) pathologist, the findings and reviewed slides were submitted to an NTP Pathology Working Group (PWG) coordinator for a second independent review. Any inconsistencies in the diagnoses made by the study laboratory and QA pathologists were resolved by the NTP pathology peer review process. Final diagnoses for reviewed lesions represent a consensus of the PWG or a consensus between the study laboratory pathologist, NTP pathologist, QA pathologist(s), and the PWG coordinator. Details of these review procedures have been described, in part, by Maronpot and Boorman⁵² and Boorman et al.⁵³

Statistical Methods

Quality Assurance Methods

The 2-week and 39-week studies were conducted in compliance with Food and Drug Administration Good Laboratory Practice Regulations.⁶⁸ In addition, the 39-week study reports were audited retrospectively by an independent QA contractor against study records submitted to the NTP archives. Separate audits covered completeness and accuracy of the pathology data, pathology specimens, final pathology tables, and a draft of this NTP Toxicity Report. Audit procedures and findings are presented in the reports and are on file at NIEHS. The audit findings were reviewed by NTP staff, and all comments were resolved or otherwise addressed during the preparation of this Toxicity Report.

Footnotes

a

ERRATUM: An error was identified in the NTP Toxicity Report on Abrasive Blasting Agents (TOX 91). The reported supplier information for each blasting agent has been corrected and trade names were added to the text in the HTML and PDF versions of this report; the new information is italicized. [September 1, 2022]

b

ERRATUM: An error was identified in the NTP Toxicity Report on Abrasive Blasting Agents (TOX 91). This sentence was added to summarize the bulk material preparation in the HTML and PDF versions of this report; the new information is italicized. [September 1, 2022]