Criteria for a Recommended Standard

Home , Refractory

Background photo on cover by James Lockey, M.D.: Photomicrograph of a rat lung in inhalation studies with airborne fibers. Inset photo on cover, courtesy of Kevin H. Dunn: NIOSH industrial hygienist performing air sampling to evaluate engineering controls in simulated work activities using RCF materials. Criteria for a Recommended Standard

Occupational Exposure to Refractory Ceramic Fibers

DEPARTMENT OF HEALTH AND HUMAN SERVICES Centers for Disease Control and Prevention National Institute for Occupational Safety and Health This document is in the public domain and may be freely copied or reprinted.

Disclaimer Mention of any company or product does not constitute endorsement by the National Institute for Occupational Safety and Health (NIOSH). In addition, citations to Web sites external to NIOSH do not constitute NIOSH endorsement of the sponsoring organizations or their programs or products. Furthermore, NIOSH is not responsible for the content of these Web sites.

Ordering Information To receive documents or other information about occupational safety and health topics, contact NIOSH at NIOSH Publications Dissemination 4676 Columbia Parkway Cincinnati, OH 45226–1998 Telephone: 1–800–35–NIOSH (1–800–356–4674) Fax: 1–513–533–8573 E-mail: [email protected] or visit the NIOSH Web site at www.cdc.gov/niosh

DHHS (NIOSH) Publication No. 2006–123 May 2006 Safe • Healthier • PeopleTM

ii Refractory Ceramic Fibers Foreword

When the U.S. Congress passed the Occupational Safety and Health Act of 1970 (Public Law 91– 596), it established the National Institute for Occupational Safety and Health (NIOSH). Through the Act, Congress charged NIOSH with recommending occupational safety and health standards and describing exposure limits that are safe for various periods of employment. These limits include but are not limited to the exposures at which no worker will suffer diminished health, functional capacity, or life expectancy as a result of his or her work experience. By means of criteria documents, NIOSH communicates these recommended standards to regulatory agencies (including the Occupational Safety and Health Administration [OSHA]), health professionals in academic institutions, industry, organized labor, public interest groups, and others in the occupational safety and health community. Criteria documents contain a critical review of the scientific and technical information about the prevalence of hazards, the existence of safety and health risks, and the adequacy of control methods. This criteria document is derived from reviews of information from human and animal studies of the toxicity of refractory ceramic fibers (RCFs) and is intended to describe the potential health effects of occupational exposure to airborne fibers of this material. RCFs are amorphous synthetic fibers produced by the melting and blowing or spinning of calcined kaolin clay or a combination of alumina, silica, and other oxides. RCFs belong to the class of synthetic vitreous fibers (SVFs)—materials that also include fibers of glass wool, rock wool, slag wool, and specialty glass. RCFs are used in commercial applications requiring lightweight, high-heat insulation (e.g., furnace and kiln insulation). Commercial production of RCFs began in the 1950s in the United States, and production increased dramatically in the 1970s. Domestic production of RCFs in 1997 totaled approximately 107.7 million lb. Currently, total U.S. production has been estimated at 80 million lb per year, which constitutes 1% to 2% of SVFs produced worldwide. In the United States, approximately 31,500 workers have the potential for occupational exposure to RCFs during distribution, handling, installation, and removal. More than 800 of these workers are employed directly in the manufacturing of RCFs and RCF products. With increasing production of RCFs, concerns about exposures to airborne fibers prompted animal inhalation studies that have indicated an increased incidence of mesotheliomas in hamsters and lung cancer in rats following exposure to RCFs. Studies of workers who manufacture RCFs have shown a positive association between increased exposure to RCFs and the development of pleural plaques, skin and eye irritation, and respiratory symptoms and conditions (including dyspnea, wheezing, and chronic cough). In addition, current and former RCF production workers have shown decrements in pulmonary function. After evaluating this evidence, NIOSH proposes a recommended exposure limit (REL) for RCFs of 0.5 fiber per cubic centimeter (f/cm3) of air as a time-weighted average (TWA) concentration for up to a 10-hr work shift during a 40-hr workweek. Limiting airborne RCF exposures to this concentration will minimize the risk for lung cancer and irritation of the eyes and upper

Refractory Ceramic Fibers iii Foreword

respiratory system and is achievable based on a review of exposure monitoring data collected from RCF manufacturers and users. However, because a residual risk of cancer (lung cancer and pleural mesothelioma) may still exist at the REL, continued efforts should be made toward reducing exposures to less than 0.2 f/cm3. Engineering controls, appropriate respiratory protection programs, and other preventive measures should be implemented to minimize worker exposures to RCFs. NIOSH urges employers to disseminate this information to workers and customers. NIOSH also requests that professional and trade associations and labor organizations inform their members about the hazards of exposure to RCFs.

John Howard, M.D. Director, National Institute for Occupational Safety and Health Centers for Disease Control and Prevention

iv Refractory Ceramic Fibers Executive Summary

The National Institute for Occupational Safety lung cancer. However, studies of worker popu- and Health (NIOSH) has reviewed data char- lations with occupational exposure to airborne acterizing occupational exposure to airborne RCFs have shown an association between ex- refractory ceramic fibers (RCFs) and informa- posure and the formation of pleural plaques, tion about potential health effects obtained increased prevalence of respiratory symptoms from experimental and epidemiologic studies. and conditions (dyspnea, wheezing, chronic From this review, NIOSH determined that oc- cough), decreases in pulmonary function, and cupational exposure to RCFs is associated with skin, eye, and upper respiratory tract irritation adverse respiratory effects as well as skin and [Lemasters et al. 1994, 1998; Lockey et al. 1996]. eye irritation and may pose a carcinogenic risk Increased decrements in pulmonary function based on the results of chronic animal inhala- among workers exposed to RCFs who are cur- tion studies. rent or former cigarette smokers indicate an apparent synergistic effect between smoking In chronic animal inhalation studies, expo- and RCF exposure [Lemasters et al. 1998]. This sure to RCFs produced an increased incidence finding is consistent with studies of other dust- of mesotheliomas in hamsters [McConnell et exposed populations. The implementation of al. 1995] and lung cancer in rats [Mast et al. improved engineering controls and work prac- 1995a,b]. The potential role of nonfibrous par- tices in RCF manufacturing processes and end ticulates generated during inhalation exposures uses have led to dramatic declines in airborne in the animal studies complicates the issue of fiber exposure concentrations [Rice et al. 1996, determining the exact mechanisms and doses 1997; Maxim et al. 2000a], which in turn have associated with the toxicity of RCFs in produc- lowered the risk of symptoms and health ef- ing carcinogenic effects [Mast et al. 2000]. The fects for exposed workers. induction of mesotheliomas and sarcomas in rats and hamsters following intrapleural and In 2002, the Refractory Ceramic Fibers Coali- intraperitoneal implantation of RCFs pro- tion (RCFC) established the Product Steward- vided additional evidence for the carcinogenic ship Program (PSP), which was endorsed by potential of RCFs [Wagner et al. 1973; Davis the Occupational Safety and Health Admin- et al. 1984; Smith et al. 1987; Pott et al. 1987]. istration (OSHA). Contained in the PSP were Lung tumors have also been observed in rats recommendations for an RCF exposure guide- 3 exposed to RCFs by intratracheal instillation line of 0.5 fiber per cubic centimeter (f/cm ) of [Manville Corporation 1991]. air as a time-weighted average (TWA) based on the contention that exposures at this concen- In contrast to the carcinogenic effects of RCFs tration could be achieved in most industries observed in experimental animal studies, epide- that manufactured or used RCFs. At this time, miologic studies have found no association be- the available health data do not provide suf- tween occupational exposure to airborne RCFs ficient evidence for deriving a precise health- and an excess rate of pulmonary fibrosis or based occupational exposure limit to protect

Refractory Ceramic Fibers Executive Summary

against lung cancer. However, given what is cancer are estimated to be between 0.03/1,000 known from the animal and epidemiological and 0.47/1,000 (based on extrapolations of risk data, NIOSH supports the intent of the PSP models from Moolgavkar et al. [1999] and Yu and concurs that a recommended exposure and Oberdörster [2000]). limit (REL) of 0.5 f/cm3 as a TWA for up to a 10-hr work shift during a 40-hr workweek Maintaining airborne RCF concentrations be- will lower the risk for developing lung cancer. low the REL requires a comprehensive safety Keeping exposures below the REL should re- and health program that includes provisions duce the risk of lung cancer to estimates be- for the monitoring of worker exposures, instal- tween 0.073/1,000 and 1.2/1,000 (based on ex- lation and routine maintenance of engineering trapolations of risk models from Moolgavkar controls, and the training of workers in good et al. [1999] and Yu and Oberdörster [2000]). work practices. Industry-led efforts have like- Keeping worker exposures below the REL will wise promoted these actions by establishing also reduce the risk of irritation of the eyes and the PSP. NIOSH believes that maintaining ex- upper respiratory system. posures below the REL is achievable at most manufacturing operations and many user ap- The risk for mesothelioma at 0.5 f/cm3 is not plications, and that the incorporation of an known but cannot be discounted. Evidence action level (AL) of 0.25 f/cm3 in the exposure from epidemiologic studies showed that higher monitoring strategy will help employers deter- exposures in the past resulted in pleural plaques mine when workplace exposure concentrations in workers, indicating that RCFs do reach pleu- are approaching the REL. The AL concept has ral tissue. Both implantation studies in rats and inhalation studies in hamsters show that RCFs been an integral element of occupational stan- can cause mesothelioma. Because of limitations dards recommended in NIOSH criteria docu- in the hamster data, the risk of mesothelioma ments and in comprehensive standards pro- cannot be quantified. However, the fact that no mulgated by OSHA and the Mine Safety and mesothelioma has been found in workers and Health Administration (MSHA). that pleural plaques appear to be less likely in NIOSH also recommends that employers im- workers with lower exposures suggests a lower plement additional measures under a compre- risk for mesothelioma at the REL. hensive safety and health program, including Because residual risks of cancer (lung cancer hazard communication, respiratory protection and pleural mesothelioma) and irritation may programs, smoking cessation, and medical still exist at the REL, NIOSH further recom- monitoring. These elements, in combination mends that all reasonable efforts be made to with efforts to maintain airborne RCF concen- work toward reducing exposures to less than trations below the REL, will further protect the 0.2 f/cm3. At this concentration, the risks of lung health of workers.

vi Refractory Ceramic Fibers Contents

Foreword ...... iii Executive Summary...... v Abbreviations...... xi Glossary...... xiv Acknowledgments...... xvii 1 Recommendations for a Refractory Ceramic Fiber (RCF) Standard . . . 1 1.1 Recommended Exposure Limit (REL) ...... 1 1.2 Definitions and Characteristics...... 1 1.2.1 Naturally Occurring Mineral Fibers...... 1 1.2.2 RCFs...... 2 1.2.3 SVFs...... 2 1.3 Sampling and Analysis ...... 2 1.4 Exposure Monitoring...... 2 1.4.1 Sampling Surveys...... 3 1.4.2 Action Level...... 3 1.5 Hazard Communication...... 4 1.6 Training...... 4 1.7 Product Formulation ...... 4 1.8 Engineering Controls and Work Practices ...... 5 1.8.1 Engineering Controls...... 5 1.8.2 Work Practices...... 5 1.9 Respiratory Protection...... 6 1.9.1 Respiratory Proctection Program...... 6 1.9.2 Respirator Selection...... 7 1.10 Sanitation and Hygiene...... 7 1.11 Medical Monitoring ...... 8 1.11.1 Oversight of the Program ...... 9 1.11.2 Elements of the Medical Monitoring Program...... 9 1.12 Labeling and Posting...... 11 1.13 Smoking Cessation...... 12 2 Background and Description of RCFs...... 13 2.1 Scope...... 13

Refractory Ceramic Fibers vii Contents

2.2 Background...... 13 2.3 Chemical and Physical Properties of RCFs...... 13 2.3.1 Fiber Dimensions...... 14 2.3.2 Fiber Durability ...... 16 3 RCF Production and Potential for Worker Exposure...... 18 3.1 Production...... 18 3.2 Potential for Worker Exposure...... 18 3.3 RCF Manufacturing Process...... 18 3.4 RCF Products and Uses...... 20 3.4.1 Examples of Products...... 20 3.4.2 Examples of Applications...... 21 4 Assessing Occupational Exposure...... 22 4.1 Air Sampling and Analytical Methods...... 22 4.2 Sampling for Airborne Fibers...... 22 4.2.1 NIOSH Fiber-Counting Rules...... 22 4.2.2 European Fiber-Counting Rules...... 23 4.3 Sampling for Total or Respirable Particulates...... 23 4.4 Sampling for Airborne Silica...... 24 4.5 Industrial Hygiene Surveys and Exposure Assessments...... 24 4.5.1 University of Pittsburgh Survey of Exposures During RCF Manufacturing 25 4.5.2 University of Cincinnati Study of Exposures During RCF Manufacturing. 25 4.5.3 RCFC/EPA Consent Agreement Monitoring Data ...... 26 4.5.4 Exposures During Installation and Removal of RCF Furnace Insulation. . 31 4.5.5 International (Canadian, Swedish, and Australian) Surveys of RCF Exposure...... 33 4.5.6 Johns Hopkins University Industrial Hygiene Surveys...... 34 4.5.7 NIOSH HHEs and Additional Sources of RCF Exposure Data...... 34 4.5.8 Discussion...... 37 5 Effects of Exposure...... 38 5.1 Health Effects in Animals (In Vivo Studies)...... 38 5.1.1 Intrapleural, Intraperitoneal, and Intratracheal Studies...... 38 5.1.2 Chronic Inhalation Studies...... 42 5.1.3 Discussion of RCF Studies in Animals...... 53 5.1.4 Lung Overload Argument Regarding Inhalation Studies in Animals . . . . . 56 5.2 Cellular and Molecular Effects of RCFs (In Vitro Studies)...... 58 5.3 Health Effects in Humans...... 60 5.3.1 Morbidity and Mortality Studies...... 60 5.3.2 Radiographic Analyses...... 62

viii Refractory Ceramic Fibers Contents

5.3.3 Respiratory Conditions and Symptom Analyses...... 76 5.3.4 Pulmonary Function Testing...... 77 5.3.5 Mortality Studies ...... 78 5.3.6 NIOSH HHEs ...... 80 5.3.7 Discussion...... 81 5.4 Carcinogenicity Risk Assessment Analyses...... 83 5.4.1 DECOS [1995] ...... 83 5.4.2 Fayerweather et al. [1997]...... 85 5.4.3 Moolgavkar et al. [1999] ...... 85 5.4.4 Discussion...... 87 6 Discussion and Summary of Fiber Toxicity...... 89 6.1 Significance of Studies with RCFs...... 89 6.2 Factors Affecting Fiber Toxicity...... 90 6.2.1 Fiber Dose...... 91 6.2.2 Fiber Dimensions...... 91 6.2.3 Fiber Durability ...... 93 6.3 Summary of RCF Toxicity and Exposure Data...... 94 7 Existing Standards and Recommendations ...... 96 8 Basis for the Recommended Standard...... 99 8.1 Background...... 99 8.2 Rationale for the REL...... 100 8.2.1 Carcinogenesis in Animal Studies...... 101 8.2.2 Health Effects Studies of Workers Exposed to RCFs...... 106 8.2.3 Carcinogenic Risk in Humans...... 107 8.2.4 Controlling RCF Exposures in the Workplace...... 109 8.3 Summary...... 109 9 Guidelines for Protecting Worker Health...... 113 9.1 Informing Workers about Hazards...... 113 9.1.1 Safety and Health Training Program ...... 113 9.1.2 Labeling and Posting ...... 113 9.2 Hazard Prevention and Control...... 114 9.2.1 Engineering Controls...... 114 9.2.2 Product Reformulation ...... 117 9.2.3 Work Practices and Hygiene ...... 117 9.2.4 Personal Protective Equipment...... 118 9.2.5 Respiratory Protection...... 118 9.3 Exposure Monitoring...... 119

Refractory Ceramic Fibers ix Contents

9.3.1 Workplace Exposure Monitoring Program ...... 119 9.3.2 Action Level...... 121 9.3.3 Sampling Strategies ...... 121 9.4 Medical Monitoring ...... 122 9.4.1 Objectives of Medical Monitoring ...... 122 9.4.2 Criteria for Medical Screening...... 122 9.4.3 Worker Participation ...... 123 9.4.4 Medical Monitoring Program Director ...... 123 9.4.5 Recommended Program Elements ...... 124 9.4.6 Written Reports to the Worker ...... 126 9.4.7 Written Reports to the Employer ...... 127 9.4.8 Employer Actions...... 127 9.5 Surveillance of Health Outcomes ...... 127 9.6 Smoking Cessation...... 128 10 Research Needs...... 129 References...... 131 Appendices A. Air Sampling Methods...... 147 B. Functional Job Categories for RCF Workers...... 179 C. Cellular and Molecular Effects of RCFs (In Vitro Studies)...... 183

Refractory Ceramic Fibers Abbreviations

ACGIH American Conference of Governmental Industrial Hygienists ACS American Cancer Society Al aluminum AL action level

A12O3 alumina

Al2Si2O5[OH]4 siliceous kaolin AM arithmetic mean APF assigned protection factor ATSDR Agency for Toxic Substances Disease Registry

B2O3 boron oxide BAL bronchoalveolar lavage °C degree(s) Celsius CaO calcium oxide CFR Code of Federal Regulations CI confidence interval DECOS Dutch Expert Committee on Occupational Standards dG 2-deoxyguanosine DNA deoxyribonucelic acid EDXA energy dispersive X-ray analyzer e.g. for example EID Education and Information Division EPA U.S. Environmental Protection Agency °F degree(s) Fahrenheit f/cm3 fibers per cubic centimeter of air Fe iron

FEF25–75 forced expiratory flow (liter/second) between 25% and 75% of the forced vital capacity

Fe2O3 ferric oxide

FEV1 forced expiratory volume in one second

FEV1/FVC ratio of forced expiratory volume in one second to forced vital capacity FJC functional job category FVC forced vital capacity g/cm3 grams per cubic meter of air GM geometric mean

Refractory Ceramic Fibers xi Abbreviations

GMD geometric mean diameter

GML geometric mean length GSD geometric standard deviation GSH glutathione γ-GT γ-glutamyltransferase HEPA high-efficiency particulate air HHE Health Hazard Evaluation hr hour(s) HTE hamster tracheal epithelial IARC International Agency for Research on Cancer i.e. that is IgG immunoglobulin IJT industry job title ILO International Labor Office ILs interleukins

K2O potassium oxide lb pound(s) LDH lactose dehydrogenase LOAEL lowest observable adverse effect level LOD limit of detection MAC maximum achievable concentration MAN Manville RCF mg/m3 milligrams per cubic meter MgO magnesium oxide min minute(s) ml milliliter(s) MLE maximum likelihood estimate mm millimeter(s) MMC metal matrix composites MMMF man-made mineral fiber MMVF man-made vitreous fiber MSDS material safety data sheet MSHA Mine Safety and Health Administration MTD maximum tolerated dose MVK Moolgavkar-Venzon-Knudson n number

Na2O sodium oxide ng nanogram(s) NIEHS National Institute of Environmental Health Sciences NIOSH National Institute for Occupational Safety and Health

xii Refractory Ceramic Fibers Abbreviations

nl nanoliter(s) NOAEL no observable adverse effect level NTP National Toxicology Program ODC ornithine decarboxylase OM Osborne Mendel OR odds ratio OSHA Occupational Safety and Health Administration P probability PA posteroanterior PCM phase contrast optical microscopy PEL permissible exposure limit PFT pulmonary function test PG prostaglandin PPE personal protective equipment PSP Product Stewardship Program RCF refractory ceramic fiber RCFC Refractory Ceramic Fibers Coalition REL recommended exposure limit ROM reactive oxygen metabolite ROS reactive oxygen species RPM rodent pleural mesothelial SAED selected area electron diffraction SD standard deviation SEM scanning electron microscopy Si silicon

SiO2 silicon dioxide SMR standardized mortality ratio SVF synthetic vitreous fiber TEM transmission electron microscopy TIMA Thermal Insulation Manufacturers Association

TiO2 titanium dioxide TLV threshold limit value TNF tumor necrosis factor TWA time-weighted average UCL upper confidence limit UICC Union Internationale Contre le Cancer UJT uniform job titles WHO World Health Organization

ZrO2 zirconium dioxide µm micrometer

Refractory Ceramic Fibers xiii Glossary

Action level (AL): A statistically derived concept used to permit an employer to have confidence (e.g., 95%) that if a measured exposure concentration is below the AL, then only a small probability exists that the actual concentration is above the exposure limit. Often established as half of the exposure limit, the AL should be designated for determining when additional controls are needed or administrative actions should be taken to reduce exposures. The purpose of using this reference is to indicate when worker exposures to hazardous substances may be approaching the exposure limit. After-service refractory ceramic fiber (RCF): RCF that has been subjected to greater than 1,800 oF (~1,000 oC) and has partially converted to the silica polymorph cristobalite. In experimental studies, this fiber is also called RCF4. Aspect ratio: The length to width ratio of a fiber. Costophrenic angle: Location on a chest radiograph where the ribs and the diaphragm appear to meet. Dyspnea grade 1: Shortness of breath on exertion, classified as less severe than grade 2. Dyspnea grade 2: Shortness of breath on exertion, excluding shortness of breath associated with hurrying on the level or walking up a slight hill, and classified as more severe than dyspnea grade 1.

FEF25–75: Forced expiratory flow (liter/second) that is between 25% and 75% of the forced vital capacity.

FEV1: Forced expiratory volume in one second, or the maximum volume of air that can be forcibly expired during the first second of expiration following a maximal inspiration. Fiber counting rules: Criteria for identifying and counting fibers during air sampling and exposure assessment. The three main conventions for fiber counting are described below (and in Section 4.2.1 and Appendix A).

■ NIOSH “A” rules—any particle >5 µm long with an aspect ratio (length to width) greater than 3:1 is considered a fiber.

■ NIOSH “B” rules—any particle >5 µm long with an aspect ratio equal to or greater than 5:1 and a diameter <3 µm is considered a fiber.

■ World Health Organization (WHO) reference method for man-made mineral fiber—any particle >5 µm long with an aspect ratio equal to or greater than 3:1 and a diameter <3 µm is considered a fiber.

xiv Refractory Ceramic Fibers Glossary

FVC: Forced vital capacity or the maximum volume of air (in liters) that can be forcibly expired from the lungs following a maximal inspiration. High-efficiency particulate air (HEPA) filter: A dry-type filter used to remove airborne particles with an efficiency equal to or greater than 99.97% for 0.3-μm particles. The lowest filtering efficiency of 99.97% is associated with 0.3-μm particles, which is approximately the most penetrating particle size for particulate filters. Inspirable dust: The fraction of airborne particles that would be inspired through the mouth and nose of a worker. MAN: A refractory ceramic fiber produced by the Johns Manville Company. Occupational medical monitoring (incorporating medical screening, surveillance): The periodic medical evaluation of workers to identify potential health effects and symptoms related to occupational exposures or environmental conditions in the workplace. An occupational medical monitoring program is a secondary prevention method based on two processes, screening and surveillance. Occupational medical screening focuses on early detection of health outcomes for individual workers. Screening may involve an occupational history assessment, medical examination, and medical tests to detect the presence of toxicants or early pathologic changes before the worker would normally seek clinical care for symptomatic disease. Occupational medical surveillance involves the ongoing evaluation of the health status of a group of workers through the collection and analysis of health data for the purpose of disease prevention and for evaluating the effectiveness of intervention programs. Pleural plaques: Discrete areas of thickening that are generally on the parietal pleura and are most commonly located at the midcostal and posterior costal areas, the dome of the diaphragm, and the mediastinal pleura. Presence of plaques is an indication of exposure to a fibrous silicate, most frequently asbestos. Radiographic opacity: A shadow on a chest X-ray film generally associated with a fibrogenic response to dust retained in the lungs [Morgan 1995]. Opacities are classified by size, shape, location, and profusion according to guidelines established by the International Labor Office [ILO 2000] www.ilo.org/public/english/support/publ/books.htm). Refractory ceramic fiber (RCF): An amorphous, synthetic fiber (Chemical Abstracts Services No. 142844–00–6) produced by melting and blowing or spinning calcined kaolin clay or a combina-

tion of alumina (Al2O3) and silicon dioxide (SiO2). Oxides may be added such as zirconia, ferric oxide, titanium oxide, magnesium oxide, calcium oxide, and alkalies. The percentage (by weight) of components is as follows: alumina, 20% to 80%; silicon dioxide, 20% to 80%; and other oxides in smaller amounts.

Respirable-sized fiber: Particles >5 μm long with an aspect ratio >3:1 and diameter ≤1.3 μm. Shot: Nonfibrous particulate that is generated during the production of RCFs from the original melt batch.

Refractory Ceramic Fibers xv Glossary

Standardized mortality ratio (SMR): The ratio of the observed number of deaths (from a specified cause) to the expected number of deaths (from that same cause) that has been adjusted to account for demographic differences (e.g., age, sex, race) between the study population and the referent population. Synthetic vitreous fiber (SVF): Any of a number of manufactured fibers produced by the melting and subsequent fiberization of kaolin clay, sand, rock, slag, etc. Fibrous glass, mineral wool, ceramic fibers, and alkaline earth silicate wools are the major types of SVF, also called man-made mineral fiber (MMMF) or man-made vitreous fiber (MMVF). Thoracic-sized fiber: Particles >5 μm long with aspect ratio >3:1 and a diameter <3 to 3.5 μm. Thoracic refers to particles penetrating to the thorax (50% cut at 10-μm aerodynamic diameter). Mineral and vitreous fibers with diameters 3 to 3.5 μm have an aerodynamic diameter of approximately 10 μm.

xvi Refractory Ceramic Fibers Acknowledgments

This document was prepared by the Education and Information Division (EID), Paul Schulte, Ph.D., Director. Thomas Lentz, Ph.D.; Kathleen MacMahon, D.V.M.; Ralph Zumwalde; and Clayton Doak were the principle authors of this document. Other members of the Document Development Branch, in particular Marie Haring Sweeney, Ph.D., were extremely helpful in providing technical reviews and comments.

For contributions to the technical content and review of this document, the authors gratefully ac- knowledge the following NIOSH personnel:

Education and Information Division Division of Surveillance, Hazard Heinz Ahlers, J.D. Evaluations, and Field Studies John Bailer, Ph.D. Mitchell Singal, M.D. Jeff Bryant David Dankovic, Ph.D. Division of Respiratory Disease Studies Jerome Flesch Donald Campbell Charles Geraci, Ph.D. Robert Castellan, M.D. Stephen Gilbert Daniel Hewett G. Kent Hatfield, Ph.D. Paul Hewett, Ph.D. Eileen Kuempel, Ph.D. Mark Hoover, Ph.D. Bonita Malit, M.D. Paul Middendorf, Ph.D. Richard Niemeier, Ph.D. Ernest Moyer, Ph.D. Ronald Schuler John Parker, M.D. John Sheehy, Ph.D. Lee Petsonk, M.D. Randall Smith Leslie Stayner, Ph.D. Health Effects Laboratory Division David Votaw Joann Wess Vincent Castranova, Ph.D. John Whalen Dan Lewis Kenneth Weber, Ph.D. Division of Applied Research and Technology Pittsburgh Research Laboratory Paul Baron, Ph.D. Andrew Cecala Kevin H. Dunn Joseph Fernback Office of the Director Laurence Reed Frank Hearl, P.E. Lloyd Stettler, Ph.D. Anita Schill, Ph.D. Mark Toraason, Ph.D. Rosemary Sokas, M.D.

Refractory Ceramic Fibers xvii Acknowledgments

The authors wish to thank Susan Afanuh, Vanessa Becks, Gino Fazio, Anne Hamilton, and Jane Weber for their editorial support and contributions to the design and layout of this document. Clerical support in preparing this document was provided by Judith Curless, Pamela Graydon, Rosmarie Hagedorn, Norma Lou Helton, and Kristina Wasmund. Finally, special appreciation is expressed to the following individuals and organizations for serving as independent external reviewers and providing comments that contributed to the development of this document:

Chris Burley Eugene McConnell, D.V.M. European Ceramic Fibres ToxPath Industry Association Raleigh, NC Paris, France National Insulation Association John Cherrie, Ph.D. Alexandria, VA Liberty Safe Work Research Centre Aberdeen, United Kingdom Dennis O’Brien, Ph.D. United Automobile, Aerospace and John Dement, Ph.D. Agricultural Implement Workers of America Duke University Medical Center Detroit, MI Durham, NC Refractory Ceramic Fibers Coalition William Kojola Washington, DC AFL-CIO, Safety and Health Department Washington, DC Carol Rice, Ph.D. University of Cincinnati, College of Medicine David Lai, Ph.D. Cincinnati, OH Office of Pollution Prevention and Toxics U.S. Environmental Protection Agency Jolanda Rijnkels, Ph.D. Washington, DC Health Council of the Netherlands Grace LeMasters, Ph.D. Den Haag, The Netherlands University of Cincinnati, College of Medicine Scott Schneider Cincinnati, OH Laborers’ Health and Safety Fund Morton Lippmann, Ph.D. of North America New York University, School of Medicine Washington, DC Tuxedo, NY Jay Turim, Ph.D. James Lockey, M.D. Sciences International, Inc. University of Cincinnati, College of Medicine Alexandria, VA Cincinnati, OH Cindy Fraleigh Wheeler Peter McClure, Ph.D. Office of Pollution Prevention and Toxics Syracuse Research Corporation U.S. Environmental Protection Agency North Syracuse, NY Washington, DC

xviii Refractory Ceramic Fibers Recommendations for a Refractory 1 Ceramic Fiber (RCF) Standard The National Institute for Occupational Safety mesothelioma, and other adverse respiratory and Health (NIOSH) recommends that expo- health effects (including irritation and com- sure to airborne refractory ceramic fibers (RCFs) promised pulmonary function) associated with be controlled in the workplace by implementing excessive RCF exposure in the workplace. Lim- the recommendations presented in this docu- iting exposures will also protect workers’ eyes ment. These recommendations are designed to and skin from the mechanical irritation asso- protect the safety and health of workers for up to ciated with exposure to RCFs. In most manu- a 10-hr work shift during a 40-hr workweek over facturing operations, it is currently possible to a 40-year working lifetime. Observance of these limit airborne RCF concentrations to 0.5 f/cm3 recommendations should prevent or greatly re- or less. Exceptions may occur during RCF fin- duce the risks of eye and skin irritation and ad- ishing operations and during the installation verse respiratory health effects (including lung and removal of RCF products, when the nature cancer) in workers with exposure to airborne of job activities presents a challenge to meet- RCFs. Preventive efforts are primarily focused ing the REL. For these operations, additional on controlling and minimizing airborne fiber protective measures are recommended. Engi- concentrations to which workers are exposed. neering and administrative controls, respirator Exposure monitoring, hazard communication, use, and other preventive measures should be training, respiratory protection programs, and implemented to minimize exposures for work- medical monitoring are also important ele- ers in RCF industry sectors where airborne ments of a comprehensive program to protect RCF concentrations exceed the REL. NIOSH the health of workers exposed to RCFs. These urges employers to disseminate this informa- elements are described briefly in this chapter tion to workers and customers, and RCF man- and in greater detail in Chapter 9. ufacturers should convey this information to downstream users. NIOSH also requests that 1.1 Recommended Exposure professional and trade associations and labor Limit (REL) organizations inform their members about the hazards of exposure to RCFs. NIOSH recommends that occupational exposures to airborne RCFs be limited to 0.5 fiber per cubic centimeter (f/cm3) of air as a time- 1.2 Definitions and weighted average (TWA) concentration for up Characteristics to a 10-hr work shift during a 40-hr workweek, measured according to NIOSH Method 7400 1.2.1 Naturally Occurring Mineral Fibers (B rules) [NIOSH 1998]. Many types of mineral fibers occur naturally. This recommended exposure limit (REL) is Asbestos is the most prominent of these fi- intended to reduce the risk of lung cancer, bers because of its industrial application. The

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

asbestos minerals include both the serpentine of heat resistance, tensile strength, durability, asbestos (chrysotile) and the amphibole min- and light weight. On a continuum, however, eral fibers, including actinolite, amosite, an- RCFs are less durable (i.e., more soluble) than thophyllite, crocidolite, and tremolite [Peters the least durable asbestos fiber (chrysotile) but and Peters 1980]. Since ancient times, min- more durable than most fibrous glass and oth- eral fibers have been mined and processed for er types of SVFs. use as insulation because of their high tensile strength, resistance to heat, durability in ac- ids and other chemicals, and light weight. The 1.2.3 SVFs predominant forms of asbestos mined and SVFs include a number of manmade (not natu- used today are chrysotile (~95%), crocidolite rally occurring) fibers that are produced by the (<5%), and amosite (<1%). melting and subsequent fiberization of kaolin For the purposes of this document, naturally clay, sand, rock, slag, and other materials. The occurring mineral fibers are distinguishable major types of SVFs are fibrous glass, mineral from synthetic vitreous fibers (SVFs) based on wool (slag wool, rock wool), and ceramic fibers the crystalline structure of the mineral fibers. (including RCFs). SVFs are also frequently re- This property causes the mineral fibers to frac- ferred to as manmade mineral fibers (MMMFs) ture longitudinally when subjected to mechan- or manmade vitreous fibers (MMVFs). ical stresses, thereby producing more fibers of decreasing diameter. By contrast, SVFs are amorphous and fracture transversely, resulting 1.3 Sampling and Analysis in more fibers of decreasing length until the Employers shall perform air sampling and segments are no longer of sufficient length to analysis to determine airborne concentrations be considered fibers. Naturally occurring min- of RCFs according to NIOSH Method 7400 (B eral fibers are generally more durable and less rules) [NIOSH 1998], provided in Appendix A soluble than SVFs, a property that accounts of this document. for the biopersistence and toxicity of mineral fibers in vivo. 1.4 Exposure Monitoring 1.2.2 RCFs Employers shall perform exposure monitoring RCFs are a type of SVF; they are amorphous as follows: synthetic fibers produced from the melting ■ Establish a workplace exposure monitor- and blowing or spinning of calcined kaolin clay or a combination of alumina (Al O ) and ing program for worksites where RCFs 2 3 or RCF products are manufactured, han- silicon dioxide (SiO2). Oxides such as zirconia, ferric oxide, titanium oxide, magnesium ox- dled, used, installed, or removed. ide, calcium oxide, and alkalies may be added. ■ Include in this program routine area and The percentage of components (by weight) is personal monitoring of airborne fiber as follows: alumina, 20% to 80%; silicon diox- concentrations. ide, 20% to 80%; and other oxides in smaller amounts. Like the naturally occurring min- ■ Design a monitoring strategy that can be eral fibers, RCFs possess the desired qualities used to

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

— evaluate a worker’s exposure to RCFs, ■ When developing an exposure sampling strategy, consider variations in work and — assess the effectiveness of engineering production schedules as well as the in- controls, work practices, and other herent variability in most area sampling factors in controlling airborne fiber [NIOSH 1995a]. concentrations, and — identify work areas or job tasks in 1.4.1.1 Focused sampling which worker exposures are routinely When sampling to determine whether worker high and thus require additional ef- RCF exposures are below the REL, a focused forts to reduce them. sampling strategy may be more practical than a random sampling approach. A focused sam- 1.4.1 Sampling Surveys pling strategy targets workers perceived to be Employers shall conduct exposure monitoring exposed to the highest concentrations of a haz- surveys to ensure that worker exposures (mea- ardous substance [Leidel and Busch 1994]. This sured by full-shift samples) do not exceed the strategy is most efficient for identifying expo- REL. Because adverse respiratory health effects sures above the REL if maximum-risk work- may occur at the REL, it is desirable to achieve ers and time periods are accurately identified. lower concentrations whenever possible. When Short tasks involving high concentrations of workers are potentially exposed to airborne airborne fibers could result in elevated expo- RCFs, employers shall conduct exposure mon- sure over full work shifts. itoring surveys as follows: Sampling strategies such as those used by Corn ■ Collect representative personal samples and Esmen [1979], Rice et al. [1997], and Max- over the entire work shift [NIOSH 1997a]. im et al. [1997] have been developed and used specifically in RCF manufacturing facilities to ■ Perform periodic sampling at least an- monitor airborne fiber concentration. In these nually and whenever any major process strategies, representative workers are selected change takes place or whenever another for sampling and are grouped according to reason exists to suspect that exposure dust zones, uniform job titles, or functional concentrations may have changed. job categories. These approaches are intended to reduce the number of required samples and ■ Collect all routine personal samples in increase the confidence of identifying workers the breathing zones of the workers. at similar risk. ■ If workers are exposed to concentrations above the REL, perform more frequent 1.4.1.2 Area sampling exposure monitoring as engineering Area sampling may be useful in exposure mon- changes are implemented and until at itoring to determine sources of airborne RCFs least two consecutive samples indicate and to assess the effectiveness of engineering that exposures no longer exceed the REL controls. [NIOSH 1977a]. 1.4.2 Action Level ■ Notify all workers of monitoring results and of any actions taken to reduce their An action level (AL) at half the REL (0.25 f/cm3) exposures. shall be used to determine when additional

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

controls are needed or when administrative ■ Inform workers about adverse respira- actions should be taken to reduce exposure to tory health effects associated with RCF RCFs. The purpose of an AL is to indicate when exposures. worker exposures to hazardous substances may ■ be approaching the REL. When air samples In work involving the removal of refrac- contain concentrations at or above the AL, the tory insulation materials, make workers probability is high that worker exposures to aware of the potential for exposure to the hazardous substance exceed the REL. respirable crystalline silica, the health effects related to this exposure, and meth- The AL is a statistically derived concept permit- ods for reducing exposures. ting the employer to have confidence (e.g., 95%) ■ Make workers who smoke cigarettes that if results from personal air samples are be- or use other tobacco products aware of low the AL, the probability is small that worker their increased risk of developing RCF- exposures are above the REL. NIOSH has con- induced respiratory symptoms and concluded that the use of an AL permits the em- ditions (see Sections 1.12 and 9.6 for rec- ployer to monitor hazardous workplace expo- ommendations about smoking cessation sures without daily sampling. The AL concept programs). has served as the basis for defining the elements of an occupational standard in NIOSH criteria documents and comprehensive standards 1.6 Training promulgated by the Occupational Safety and Health Administration (OSHA) and the Mine Employers shall provide the following training Safety and Health Administration (MSHA). for workers exposed to RCFs:

■ Train workers to detect hazardous situa- 1.5 Hazard Communication tions. ■ Employers shall take the following measures to Inform workers about practices or op- inform workers about RCF hazards: erations that may generate high airborne fiber concentrations (e.g., cutting ■ Establish a safety and health training and sanding RCF boards and other RCF program for all workers who manufac- products). ture, use, handle, install, or remove RCF ■ Train workers how to protect themselves products or perform other activities that by using proper work practices, engi- bring them into contact with RCFs. neering controls, and personal protective ■ Inform employees and contract work- equipment (PPE). ers about hazardous substances in their work areas. 1.7 Product Formulation ■ Instruct workers about how to get information from material safety data sheets One factor recognized as contributing to the (MSDSs) for RCFs and other chemicals. toxicity of an inhaled fiber is its durability and resistance to degradation in the respira- ■ Provide MSDSs onsite and make them tory tract. Chemical characteristics place RCFs easily accessible. among the most durable SVFs. As a result, an

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

inhaled RCF that is deposited in the alveolar re- during the sanding of vacuum-formed gion of the lung will persist longer in the lungs RCF products [Dunn et al. 2004]. than a less durable fiber. Therefore, NIOSH recommends substituting a less durable fiber ■ Enclosed processes used during manu- for RCFs or reformulating the chemistry of facturing to keep airborne fibers con- RCFs toward this end to reduce the hazard for tained and separated from workers exposed workers. As part of product stewardship efforts, several RCF producers within the ■ Water knives, which are high-pressure Refractory Ceramic Fibers Coalition (RCFC) water jets that effectively cut and trim the have developed new and less biopersistent fi- edges of RCF blanket while suppressing bers termed alkaline earth silicate wools [Max- dust and limiting the generation of air- im et al. 1999b]. Newly developed fibers should borne fibers undergo industry-sponsored testing before their selection and commercial use to exclude possible adverse health effects from exposure. 1.8.2 Work Practices Employers shall implement appropriate work practices to help keep worker exposures at or 1.8 Engineering Controls below the REL for RCFs. The following work and Work Practices practices are recommended to help reduce concentrations of airborne fibers: 1.8.1 Engineering Controls ■ Limit the use of power tools unless they Employers shall use and maintain appropriate are equipped with local exhaust or dust engineering controls to keep airborne concen- collection systems. trations of RCFs at or below the REL during the manufacture, use, handling, installation, — Be aware that manually powered hand and removal of RCF products. Engineering tools generate less dust and fewer air- controls for controlling RCFs include the fol- borne fibers, but they often require lowing: additional physical effort and time ■ Local exhaust ventilation or dust collec- and may increase the risk of muscu- tion systems at or near dust-generating loskeletal disorders. systems — The additional physical effort required — Band saws used in RCF manufactur- by hand tools may also increase the ing and finishing operations have rate and depth of breathing and con- been fitted with such engineering sequently affect the inhalation rate controls to capture fibers and dust and deposition of fibers in the lungs. during cutting operations, thereby reducing exposures for the band saw ■ Use ergonomically correct tools and operator [Venturin 1998]. proper workstation design to reduce the risk of musculoskeletal disorders. — Disc sanders fitted with similar local exhaust ventilation systems effectively ■ Use high-efficiency particulate air-filtered reduce airborne RCF concentrations (HEPA-filtered) vacuums.

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

■ Use wet sweeping to suppress airborne control worker exposures below the REL for fiber and dust concentrations during RCFs. NIOSH recognizes that controlling ex- cleanup. posures to RCFs is a particular challenge during the finishing stages of RCF product manufac- ■ When removing after-service RCF prod- turing and during the installation and removal ucts, dampen insulation with a light wa- of refractory materials ter spray to prevent fibers and dust from becoming airborne. (However, use cau- 1.9.1 Respiratory Protection Program tion when dampening refractory linings during installation, since water can dam- When respiratory protection is needed, em- age refractory-lined equipment, causing ployers shall establish a comprehensive respi- the generation of steam and possible ex- ratory protection program as described in the plosion during heating.) OSHA respiratory protection standard [29 CFR* 1910.134]. Elements of a respiratory pro- ■ Clean work areas regularly using a HEPA- tection program must be established and de- filtered vacuum or wet sweeping to mini- scribed in a written plan that is specific to the mize accumulation of debris. workplace. The plan must include the following elements: ■ Ensure that workers wear long-sleeved clothing, gloves, and eye protection when ■ Procedures for selecting respirators performing potentially dusty activities involving RCFs or RCF products. For ■ Medical evaluations of workers required some activities, disposable clothing or to wear respirators coveralls may be preferred. ■ Fit-testing procedures

■ Routine-use procedures and emergency 1.9 Respiratory Protection respirator-use procedures

Respirators shall be used while performing ■ Procedures and schedules for cleaning, any task for which the exposure concentra- disinfecting, storing, inspecting, repairing, tion is unknown or has been documented to discarding, and maintaining respirators be higher than the NIOSH REL of 0.5 f/cm3 as a TWA. However, respirators shall not be used ■ Procedures for ensuring adequate air as the primary means of controlling worker ex- quality for supplied-air respirators posures. ■ Training in respiratory hazards

When possible, use other methods for mini- ■ Training in the proper use and mainte- mizing worker exposures to RCFs: nance of respirators

■ Product substitution ■ Program evaluation procedures

■ Engineering controls ■ Procedures for ensuring that workers who voluntarily wear respirators (excluding ■ Changes in work practices filtering facepieces known as dust masks) Use respirators when available engineering controls and work practices do not adequately *Code of Federal Regulations. See CFR in references.

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

comply with the medical evaluation and ■ Always perform a comprehensive as- cleaning, storing, and maintenance re- sessment of workplace exposures to de- quirements of the standard termine the presence of other possible contaminants (such as silica) and to en- ■ A designated program administrator who is qualified to administer the respiratory sure that proper respiratory protection is protection program used.

■ Employers shall update the written program as Use only respirators approved by NIOSH necessary to account for changes in the work- and MSHA. place that affect respirator use. In addition, For information and assistance in establishing employers are required to provide at no cost a respiratory protection program and selecting to workers all equipment, training, and medi- appropriate respirators, see the OSHA Respira- cal evaluations required under the respiratory tory Protection Advisor on the OSHA Web site protection program. at www.osha.gov. Additional information is also available from the NIOSH Respirator Se- 1.9.2 Respirator Selection lection Logic [NIOSH 2004], the NIOSH Guide When conditions of exposure to airborne RCFs to Industrial Respiratory Protection [NIOSH exceed the REL, proper respiratory protection 1987b], and the NIOSH Guide to the Selection shall be selected as follows: and Use of Particulate Respirators Certified under 42 CFR 84 [NIOSH 1996]. ■ Select, at a minimum, a half-mask, air- purifying respirator equipped with a 100 series particulate filter. This respirator 1.10 Sanitation and Hygiene has an assigned protection factor (APF) of 10. Employers shall take the following measures to protect workers potentially exposed to RCFs: ■ Provide a higher level of protection and prevent facial or eye irritation from ■ Do not permit smoking, eating, or drink- RCF exposure by using a full-facepiece, ing in areas where workers may contact air-purifying respirator equipped with RCFs. a 100-series filter; or use any powered, air-purifying respirator equipped with a ■ Provide showering and changing areas tight-fitting facepiece (full-facepiece). free from contamination where workers can store work clothes and change into ■ Consider providing a supplied-air res- street clothes before leaving the work pirator with a full facepiece for workers site. who remove after-service RCF insulation (e.g., furnace insulation) and are there- ■ Provide services for laundering work fore exposed to high and unpredictable clothes so that workers do not take con- concentrations of RCFs. These respira- taminated clothes home. tors provide a greater level of respiratory protection. Use them whenever the work ■ Protect laundry workers handling RCF- task involves potentially high concentra- contaminated clothes from airborne con- tions of airborne fibers. centrations that are above the REL.

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

Workers shall take the following protective ■ Collect baseline data for all employees measures: before they begin work with RCFs.

■ Do not smoke, eat, or drink in areas po- ■ Continue periodic medical screening tentially contaminated with RCFs. throughout their lifetime.

■ If fibers get on the skin, wash with warm ■ Use medical surveillance, which involves water and mild soap. the aggregate collection and analysis of medical screening data, to identify oc- ■ Apply skin-moisturizing cream or lotion cupations, activities, and work processes as needed to avoid irritation caused by in need of additional primary prevention frequent washing. efforts. ■ Wear long-sleeved clothing, gloves, and ■ Include all workers potentially exposed eye protection when performing poten- to RCFs (in both manufacturing and tially dusty activities involving RCFs. end-use industries) in an occupational ■ Vacuum this clothing with a HEPA-filtered medical monitoring program. vacuum before leaving the work area. ■ Provide workers with information about ■ Do not use compressed air to clean the the purposes of medical monitoring, the work area or clothing and do not shake health benefits of the program, and the clothing to remove dust. These processes procedures involved. will create a greater respiratory hazard ■ Include the following workers (who with airborne dust and fibers. could receive the greatest benefits from ■ Do not wear work clothes or protective medical screening) in the medical moni- equipment home. Change into clean toring program: clothes before leaving the work site. — Workers exposed to elevated fiber concentrations (e.g., all workers ex- 1.11 Medical Monitoring posed to airborne fiber concentrations above the AL of 0.25 F/cm3, as Medical monitoring (in combination with described in Section 9.3) resulting intervention strategies) represents — Workers in areas or in specific jobs secondary prevention and should not replace and activities (regardless of airborne primary prevention efforts to control airborne fiber concentration) in which one or fiber concentrations and worker exposures more workers have symptoms or re- to RCFs. However, compliance with the REL spiratory changes apparently related for RCFs (0.5 f/cm3) does not guarantee that to RCF exposure all workers will be free from the risk of RCF- induced respiratory irritation or respiratory — Workers who may have been previ- health effects. Therefore, medical monitoring ously exposed to asbestos or other is especially important, and employers shall recognized occupational respiratory establish a medical monitoring program as fol- hazards that place them at an in- lows: creased risk of respiratory disease

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

1.11.1 Oversight of the Program ■ A physical examination of all systems with an emphasis on the respiratory sys- Assign oversight of the medical monitor- tem and the skin ing program to a qualified physician or other qualified health care provider (as determined ■ A spirometric test (note that anyone ad- by appropriate State laws and regulations) who ministering spirometric testing as part of is informed and knowledgeable about the fol- the medical monitoring program should lowing: have completed a NIOSH-approved training course in spirometry or other equiva- ■ Administering and managing a medical monitoring program for occupational lent training) hazards ■ A chest X-ray (all chest X-ray films should be interpreted by a certified NIOSH B ■ Establishing a respiratory protection program based on an understanding of Reader using the standard International requirements of the OSHA respiratory Classification of Radiographs of Pneumo- protection standard and types of respi- conioses [ILO 2000, or the most recent ratory protection devices available at the equivalent]) workplace ■ Other medical tests as deemed appropriate by the responsible health care profes- ■ Identifying and managing work-related respiratory effects or illnesses sional

■ ■ Identifying and managing work-related A standardized respiratory symptom ques- skin diseases tionnaire, such as the American Thoracic Society respiratory questionnaire [Ferris 1978, or the most recent equivalent] 1.11.2 Elements of the Medical Monitoring Program ■ A standardized occupational history ques- Include the following elements in a medical tionnaire that gathers information about monitoring program for workers exposed to all past jobs with (1) special emphasis on RCFs: (1) an initial medical examination, (2) pe- those with potential exposure to dust and riodic medical examinations at regularly sched- mineral fibers, (2) a description of all du- uled intervals, (3) more frequent and detailed ties and potential exposures for each job, medical examinations as needed on the basis of and (3) a description of all protective the findings from these examinations, (4) work- equipment the worker has used er training, (5) written reports of medical findings, (6) quality assurance, and (7) evaluation. 1.11.2.2 Periodic examinations These elements are described in the following subsections. Administer periodic examinations (including a physical examination of the respiratory system and the skin, spirometric testing, a re- 1.11.2.1 Initial (baseline) examination spiratory symptom update questionnaire, and Perform an initial (baseline) examination as near an occupational history update questionnaire) as possible to the date of beginning employment at regular intervals determined by the medical (within 3 months) and include the following: monitoring program director. Determine the

Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

frequency of the periodic medical examina- exposures (e.g., chronic cough, difficult breath- tions according to the following guidelines: ing, wheezing, skin irritation). Instruct workers to report these symptoms to the designated ■ For workers with fewer than 10 years since medical monitoring program director or other first exposure to RCFs, conduct periodic qualified health care provider for appropriate examinations at least once every 5 years. diagnosis and treatment. ■ For workers with 10 or more years since first exposure to RCFs, conduct periodic 1.11.2.5 Written reports of medical examinations at least once every 2 years. findings A chest X-ray and spirometric testing are im- Following initial and periodic medical exami- portant on initial examination and may also nations, the physician or other qualified health be appropriate medical screening tests during care provider shall give each worker a written periodic examinations for detecting respira- report containing tory system changes, especially in workers — results of any medical tests performed with more than 10 years since first exposure to on the worker, RCFs. A qualified health care provider should consult with the worker to determine whether — a medical opinion in plain language the benefits of periodic chest X-rays warrant about any medical condition that the additional exposure to radiation. would increase the worker=s risk of impairment from exposure to airborne RCFs, 1.11.2.3 More frequent evaluations — recommendations for limiting the Workers may need to undergo more frequent worker=s exposure to RCFs (which and detailed medical evaluations if the attend- may include the use of appropriate ing physician determines that he or she has any PPE, as warranted), and of the following indications: — recommendations for further evalu- ■ New or worsening respiratory symptoms ation and treatment of any medical or findings (e.g., chronic cough, difficult conditions detected. breathing, wheezing, reduced lung function, or radiographic indications of pleu- Following initial and periodic medical exami- ral plaques or fibrosis) nations, the physician or other qualified health care provider shall also give a written report to ■ History of exposure to other respiratory the employer containing hazards (e.g., asbestos) — occupationally pertinent results of ■ Recurrent or chronic dermatitis the medical evaluation, ■ Other medically significant reason(s) for — a medical opinion about any medi- more detailed assessment cal condition that would increase the worker=s risk of impairment from ex- 1.11.2.4 Worker training posure to airborne RCFs, Provide workers with sufficient training to — recommendations for limiting the recognize symptoms associated with RCF worker=s exposure to RCFs or other

10 Refractory Ceramic Fibers 1 n Recommendations for a Refractory Ceramic Fiber Standard

agents in the workplace (which may ■ Ensure that the medical monitoring pro- include the use of appropriate PPE or gram director communicates regularly reassignment to another job), and with the employer’s safety and health personnel (e.g., industrial hygienists) to — a statement to indicate that the work- identify work areas that may require con- er has been informed about the re- trol measures to minimize exposures to sults of the medical examination and about any medical condition(s) that workplace hazards. should have further evaluation or treatment. 1.11.2.7 Evaluation Findings, test results, or diagnoses that have no Employers shall evaluate their medical moni- bearing on the worker=s ability to work with toring programs as follows: RCFs shall not be included in the report to the ■ Periodically have standardized medical employer. Safeguards to protect the confiden- tiality of the worker=s medical records shall screening data aggregated and evaluated be enforced in accordance with all applicable by an epidemiologist or other knowl- regulations and guidelines. edgeable person to identify patterns of worker health that may be linked to work 1.11.2.6 Quality assurance activities and practices requiring additional primary preventive efforts. Employers shall do the following to ensure the effective implementation of a medical moni- ■ Combine routine aggregate assessments toring program: of medical screening data with evaluations of exposure monitoring data to ■ Ensure that workers follow the qualified identify needed changes in work areas or health care provider’s recommended ex- exposure conditions. posure restrictions for RCFs and other workplace hazards.

■ Ensure that workers use appropriate PPE 1.12 Labeling and Posting if they are exposed to RCF concentra- Employers shall post warning labels and signs tions above the REL. as follows:

■ Encourage workers to participate in the ■ Post warning labels and signs describing medical monitoring program and to re- the health risks associated with RCFs at port any symptoms promptly to the pro- entrances to work areas and inside work gram director. areas where airborne concentrations of ■ Provide any medical evaluations that are RCFs may exceed the REL. part of the medical monitoring program ■ Depending on work practices and the at no cost to the workers. airborne concentrations of RCFs, state ■ When implementing job reassignments on the signs the need to wear protective recommended by the medical program clothing and the appropriate respiratory director, ensure that workers do not lose protection for RCF exposures above the wages, benefits, or seniority. REL.

Refractory Ceramic Fibers 11 1 n Recommendations for a Refractory Ceramic Fiber Standard

■ If respiratory protection is required, post Employers shall encourage smoking cessation the following statement: among RCF-exposed workers as follows:

■ Establish smoking cessation programs to RESPIRATORY PROTECTION inform workers about the increased haz- REQUIRED IN THIS AREA ards of cigarette smoking and exposure to RCFs.

■ Print all labels and warning signs in both ■ Provide assistance and encouragement English and the predominant language for workers who want to quit smoking. of workers who do not read English. ■ Prohibit smoking in the workplace. ■ Verbally inform workers about the hazards and instructions printed on the la- ■ Disseminate information about health bels and signs if they are unable to read promotion and the harmful effects of them. smoking.

■ Offer smoking cessation programs to 1.13 Smoking Cessation workers at no cost to participants. NIOSH recognizes a synergistic effect between ■ Encourage activities that promote physi- exposure to RCFs and cigarette smoking. This cal fitness and other healthy lifestyle effect increases the risk of adverse respiratory health effects induced by RCFs. In studies of practices affecting respiratory and car- workers exposed to various airborne contami- diovascular health (e.g., through training nants, combined exposures to smoking and programs, employee assistance programs, airborne dust have been shown to contribute and health education campaigns). to the increased risk of occupational respiratory diseases, including chronic bronchitis, NIOSH recommends that all workers who emphysema, and lung cancer [Morgan 1994; smoke and are potentially exposed to RCFs Barnhart 1994]. participate in smoking cessation programs.

12 Refractory Ceramic Fibers Background and Description 2 of RCFs 2.1 Scope research relating to fiber toxicity and occupational exposures. Information about RCFs was collected and reviewed for this document to assess the health hazards associated with occupational exposure 2.2 Background to this airborne fiber. Chapter 2 describes the In 1977, NIOSH reviewed health effects data background for studying the health effects of on occupational exposure to fibrous glass and workplace exposures to RCFs. Information determined the principal adverse health effects is presented about the physical and chemical to be skin, eye, and upper respiratory tract ir- properties of RCFs, including the morphol- ritation as well as the potential for nonmalig- ogy, dimensions, and durability of fibers that nant respiratory disease. At that time NIOSH make up RCF-containing products. Chapter recommended the following: 3 discusses the production and uses of RCFs as a high-temperature insulation material; the Occupational exposure to fibrous glass shall be chapter also describes the number of workers controlled so that no worker is exposed at an with potential for exposure to RCFs. Chapter 4 airborne concentration greater than 3,000,000 presents a review of the literature on potential fibers per cubic meter of air (3 fibers per cubic workplace exposures to airborne RCFs during centimeter of air); . . . airborne concentrations manufacturing and end uses of RCF products. determined as total fibrous glass shall be lim- Chapter 5 describes the effects of exposure to ited to a TWA of 5 milligrams per cubic meter RCFs—first with reviews of animal studies and of air [NIOSH 1977]. then with a description of epidemiologic stud- NIOSH also stated that until more informa- ies of RCFs, focusing on U.S. and European tion became available, this recommendation workers in the RCF manufacturing industry. should be applied to other MMMFs, also called Recent quantitative risk assessments of RCFs SVFs. Since then, additional data have become are also summarized in this chapter. Chapter available from studies in animals and humans 6 contains a discussion of fiber characteristics exposed to RCFs. The purpose of this report is and the parameters (dose, dimensions, and du- to review and evaluate these studies and other rability) that determine fiber toxicity. Chapter information about RCFs. 7 summarizes existing standards and guidelines for occupational exposure to RCFs. Chapter 8 provides the basis and rationale for the NIOSH 2.3 Chemical and Physical REL. Chapters 1 and 9 provide recommenda- Properties of RCFs tions and guidelines for minimizing exposures to airborne fibers of RCFs in the workplace. RCFs (Chemical Abstracts Service No. 142844– Finally, Chapter 10 discusses future areas for 00–6) are amorphous fibers that belong to the

Refractory Ceramic Fibers 13 2 n Background and Description of RCFs

larger class of SVFs, which also includes fibers RCFs [RCFC 1996]. The addition of stabiliz- of glass wool, mineral wool, slag wool, and ers and binders alters the properties of dura- specialty glass. SVFs vary according to chemi- bility and heat resistance for RCFs. Generally, cal and physical properties, making them three types of RCFs are manufactured, and a suitable for different uses. Like the naturally fourth after-service fiber (often recognized in occurring mineral fibers defined in Section the literature) is distinguished according to its 1.2, RCFs possess desired qualities of heat re- unique chemistry and morphology. Table 2–2 sistance, tensile strength, durability, and light presents the chemistries of the four fiber types, weight. The maximum end-use temperature numbered RCF1 through RCF4. RCF1 is a ka- for RCFs ranges from approximately 1,050 to olin fiber; RCF2 is an alumina/silica/zirconia 1,425 oC (1,920 to 2,600 oF), depending on the fiber; RCF3 is a high-purity (alumina/silica) exact chemistry of the fiber. Unlike naturally fiber; and RCF4 is an after-service fiber, charac- occurring mineral fibers, however, SVFs such terized by devitrification (i.e., formation of the as RCFs and fibrous glass are noncrystalline in silica polymorph cristobalite), which occurs structure and fracture transversely, retaining during product use over an extended period the same diameter but creating shorter fibers. of time at temperatures exceeding 1,050 to In contrast, the crystalline structure of mineral 1,100 oC (>1,900 oF). Another fiber subcategory fibers such as asbestos causes the fibers to frac- is RCF1a, prepared from commercial RCFs us- ture along the longitudinal plane under me- ing a less aggressive method than that used to chanical stresses, resulting in more fibers with prepare RCF1 for animal inhalation studies the same length but smaller diameters. These [Brown et al. 2000]. RCF1a is distinguished differences in morphology and cleavage pat- from RCF1 used in chronic animal inhalation terns suggest that work with SVFs is less likely studies, the former having a greater concentra- to generate high concentrations of airborne fi- tion of longer fibers and fewer nonfibrous par- bers than work with asbestos for comparable ticles. The lower ratio of respirable nonfibrous operations, since large-diameter fibers settle particles to fibers in RCF1a compared with out in the air faster than small-diameter fi- RCF1 has been shown to affect lung deposi- bers [Assuncao and Corn 1975; Cherrie et al. tion and clearance in animal inhalation studies 1986; Lippmann 1990]. During the manufac- [Brown et al. 2000; Bellman et al. 2001]. Chap- turing of RCFs, approximately 50% of product ter 5 presents additional discussion of animal (by weight) is generated as fiber, and 50% is a studies and test fiber characteristics. byproduct made up of nonfibrous particulate material called shot. Selected physical charac- 2.3.1 Fiber Dimensions teristics of RCFs are presented in Table 2–1. Fibers of biological importance are those that RCFs are produced by the blowing or spinning of become airborne and have dimensions within

furnace‑melted siliceous kaolin (Al2Si2O5[OH]4) inhalable, thoracic, and respirable size ranges. clay or blends of kaolin, silica, and zircon. RCFs Thoracic-sized fibers (<3 to 3.5 μm in diam- are also referred to as alumina-based or kaolin- eter) and respirable-sized fibers (<1.3 μm in based ceramic fibers because they are produced diameter) with lengths up to 200 μm [Tim- from a 50:50 mixture of alumina and silica brell 1982; Lippmann 1990; Baron 1996] are [IARC 1988]. Other oxides (including those capable of reaching the portion of the respi- of boron, titanium, and zirconium) are added ratory system below the larynx. Respirable- as stabilizers to alter the physical properties of sized fibers are of biological concern because

14 Refractory Ceramic Fibers 2 n Background and Description of RCFs

Table 2–1 . Selected physical characteristics of RCFs

Characteristic Description Softening point 1,700 to 1,800 ˚C Refractive index 1.55 to 1.57 Specific gravity (density) 2.6 to 2.7 g/cm3 Shot content (nonfibrous particulate) 20% to 50% by weight Nominal diameter (bulk) 1.2 to 3 µm Length (bulk) 2 to 254 µm Dissolution rate (at pH=7.4) 1 to 10 ng/cm2/hr Sources: RCFC [1996], TIMA [1993], and IARC [1988].

Table 2–2 . The chemistry of stock RCFs (% oxide)

Oxide component RCF1 RCF2 RCF3 RCF4

Silicon dioxide (SiO2) 47.7 50 50.8 47.7

Alumina (Al2O3) 48 35 48.5 48

Ferric oxide (Fe2O3) 0.97 <0.05 0.16 0.97

Titanium dioxide (TiO2) 2.05 0.04 0.02 2.05

Zirconium dioxide (ZrO2) 0.11 15 0.23 0.11 Calcium oxide (CaO) 0.07 0.05 0.04 0.07 Magnesium oxide (MgO) 0.98 0.01 <0.01 0.08

Sodium oxide (Na2O) 0.54 <0.3 0.19 0.54

Potassium oxide (K2O) 0.16 <0.01 <0.01 0.16 Adapted from Mast et al. [1995a].

they are capable of reaching the lower airways symptoms. Fiber dimensions are a significant and gas exchange regions of the lungs when in- factor in determining their deposition within haled. Longer or thicker airborne fibers gener- the lung, biopersistence, and toxicity. ally settle out of suspension or, if inhaled, are generally filtered out in the nasal passage or RCFs and other SVFs are manufactured to meet deposited in the upper airways. Thoracic-sized specified nominal diameters according to the fi- fibers that are inhaled and deposited in the up- ber type and intended use. RCFs are produced per respiratory tract are generally cleared more with nominal diameters of 1.2 to 3 μm [Esmen readily from the lung, but they have the poten- et al. 1979; Vu 1988; TIMA 1993]. Typical di- tial to cause irritation and produce respiratory ameters for an individual RCF (as measured in

Refractory Ceramic Fibers 15 2 n Background and Description of RCFs

RCF-containing products) range from 0.1 to ported that 99.7% of the fibers had diameters 20 μm, with lengths ranging from 5 to 200 μm <3 μm and 64% had lengths >10 μm. Measure- [IARC 1988]. In bulk samples taken from three ments of airborne fibers in the European RCF RCF blanket insulation products, more than manufacturing industry are comparable: Rood 80% of the fibers counted by phase contrast [1988] reported that all fibers observed were optical microscopy (PCM) were <3 μm in di- in the thoracic and respirable size range (i.e., diameter [Brown 1992]. This result is consistent ameter <3 to 3.5 μm), with median diameters with those from another study of bulk sam- ranging from 0.5 to 1.0 μm and median lengths ples of RCF insulation materials [Christensen from 8 to 23 μm. et al. 1993], which found the fibers to have geo- Cheng et al. [1992] analyzed an air sample metric mean diameters (GMD) ranging from 1.5 to 2.8 μm (arithmetic mean [AM] diam- for fibers during removal of after-service RCF eter range=2.3 to 3.9 μm; median diameter blanket insulation from a refinery furnace. Fi- ber diameters ranged from 0.5 to 6 μm, with a range=1.6 to 3.3 μm). median diameter of 1.6 μm. The length of fi- Studies of airborne fiber size distributions in bers ranged from 5 to 220 μm. Of 100 fibers RCF manufacturing operations indicate that randomly selected and analyzed from the air these fibers meet the criteria for thoracic- and sample, 87% were within the thoracic and re- respirable-sized fibers. One early study of three spirable size range. Another study of exposures domestic RCF production facilities found that to airborne fibers in industrial furnaces dur- approximately 90% of airborne fibers were ing installation and removal of RCF materials

<3 μm in diameter, and 95% of airborne fibers found GMD values of 0.38 and 0.57 μm, respec- were <4 μm in diameter and <50 μm long tively [Perrault et al. 1992]. [Esmen et al. 1979]. The study showed that diameter and length distributions of airborne 2.3.2 Fiber Durability fibers in the facilities were consistent, with a

GMD of 0.7 μm and a geometric mean length Fiber durability can affect the biologic activity

(GML) of 13 μm. Another study [Lentz et al. of fibers inhaled and deposited in the respira- 1999] used these data in combination with tory system. Durable fibers are more biopersis- monitoring data from two additional studies tent, thereby increasing the potential for caus- [MacKinnon et al. 2001; Maxim et al. 1997] ing a biological effect. Durability of a fiber is at RCF manufacturing plants to review char- measured by the amount of time it takes for acteristics of fibers sized from 118 air samples the fiber to fragment mechanically into short- covering 20 years (1976–1996). Fibers with di- er fibers or dissolve in biological fluids. RCFs ameters <1 μm (n=3,711) were measured by tested in vitro with a solution of neutral pH transmission electron microscopy (TEM). Of (modified Gamble’s solution) had a dissolu- these, 52% had diameters <0.4 μm, 75% had tion rate of 1 to 10 ng/cm2 per hr [Leineweber diameters<0.6 μm, and 89% had diameters 1984]. This test is biologically relevant because <0.8 μm. Fiber lengths ranged from <0.6 to >20 of the similarity of the solution to the condi- μm, with 68% of fibers measuring 2.4 to 20 μm tions of the pulmonary interstitial fluid. By long and 19% of the fibers >20 μm long. On comparison, other SVFs (glass and slag wools) the basis of the results of TEM analysis of 3,357 are more soluble, with dissolution rates in the RCFs observed on 98 air samples collected in 100s of ng/cm2 per hr [Scholze and Conradt RCF manufacturing sites, Allshouse [1995] re- 1987]. Along a continuum of fiber durability

16 Refractory Ceramic Fibers 2 n Background and Description of RCFs

determined in tests using simulated lung flu- Chrysotile, which is considered the most sol- ids at pH 7.4, the asbestos fiber crocidolite has uble form of asbestos, has a dissolution rate of a dissolution rate of <1 ng/cm2 per hr, RCF1 <1 to 2 ng/cm2 per hr. and MMVF32 (E glass) have dissolution rates of 1 to 10 ng/cm2 per hr, MMVF21 has a dis- RCFs dissolve more rapidly than chrysotile, even solution rate of 15 to 25 ng/cm2 per hr, other though RCFs have a thicker diameter (by an or- fibrous glass and slag wools have dissolution der of magnitude) than chrysotile. The rate of rates in the range of 50 to 400 ng/cm2 per hr, dissolution is an important fiber characteristic and the alkaline earth silicate wools have dis- that affects the clearance time and biopersis- solution rates ranging from approximately tence of the fiber in the lung. The significance of 60 to 1,000 ng/cm2 per hr [Christensen et al. fiber dimension, clearance, and dissolution (i.e., 1994; Maxim et al. 1999b; Moore et al. 2001]. breakage, solubility) is discussed in Chapter 6.

Refractory Ceramic Fibers 17 RCF Production and Potential for 3 Worker Exposure 3.1 Production and Vesuvius (King of Prussia, PA, formerly Premier Refractories and Chemicals). The lat- RCF production in the United States began ter three producers account for an estimated in 1942 on an experimental basis, but RCFs 90% of domestic production and are members were not commercially available until 1953. of the RCFC, which has been active in moni- Sales of RCFs were modest initially, but they toring exposures, developing product steward- began to expand when the material gained ac- ship programs, and funding research to study ceptance as an economical alternative insula- RCF hazards and safe work practices for RCF tion for high-temperature kilns and furnaces. manufacturing and use. Commercial production of RCFs first reached significant levels in the 1970s as oil shortages necessitated reductions in energy consump- 3.2 Potential for Worker tion. The growing demand for RCFs has also Exposure been strongly influenced by the recognition of health effects associated with exposure to Approximately 31,500 workers in the United asbestos-containing materials and the increas- States are potentially exposed to RCFs during ingly stringent regulation of these products in manufacturing, processing, or end use. A simi- the United States and many other countries. lar number of workers are potentially exposed to RCFs in Europe. Of these workers, about 800 Annual domestic production of RCFs was an (3%) are employed in the actual manufactur- estimated 85.7 million lb in 1990; in 1997, pro- ing of RCFs and RCF products [Maxim et al. duction of RCFs in the United States totaled 1997; RCFC 2004]. 107.7 million lb annually [RCFC 1998]. Cur- rently, total U.S. production is estimated to be 80 million lb per year, representing about 1% 3.3 RCF Manufacturing to 2% of the worldwide production of SVFs Process [RCFC 2004]. RCFs are also produced in Mex- ico, Canada, Brazil, Venezuela, South Africa, The manufacture of RCFs (Figure 3–1) begins Australia, Japan, China, Korea, Malaysia, Tai- by blending raw materials, which may include wan, and several countries in Europe [RCFC kaolin clay, alumina, silica, and zirconia in 1996]. In the United States and Puerto Rico, a batch house. The batch mix is then trans- the primary producers of RCFs include A.P. ferred either manually or automatically to a Green Industries (Pryor, OK), Unifrax Cor- furnace to be melted at temperatures exceed- poration (Niagara Falls, NY, formerly Carbo- ing 1,600 oC. On reaching a specified temper- rundum), Thermal Ceramics (Augusta, GA), ature and viscosity in the furnace, the molten

18 Refractory Ceramic Fibers 3 n RCF Production and Potential for Worker Exposure

Raw materials (AI2O3, SiO2, TiO2, MgO, CaO, Na2O, K2O, etc.) are Bulk fiber is packaged. added to batch mix. Product is shipped, stored, or fabricated into specialty product.

Batch is melted in Fibers settle into furnace at >1600˚C. collection chamber. Blanket is cut to specifications and packaged. Lubricants are added and fibers are processed Fibers are produced by needle-felting machine. by blowing or spinning molten materials drained from furnace. Felt or blanket is cured in tempering oven.

Figure 3–1. Process flow chart for RCF production.

Refractory Ceramic Fibers 19 3 n RCF Production and Potential for Worker Exposure

batch mixture drains from the furnace and is woodstoves), and in space shuttle tiles. RCFs fiberized, either through exposure to pressur- have been formed into noise-control blankets ized air or by flowing through a series of spin- [Thornton et al. 1984] and used as a replace- ning wheels [Hill 1983]. Fans are used to create ment for refractory bricks in industrial kilns a partial vacuum that pulls the fibers into a col- and furnaces [RCFC 1996]. RCFs have found lection or settling chamber. RCFs may then be increasing application as reinforcements in conveyed pneumatically to a bagging area for specialized metal matrix composites (MMC), packaging as bulk fiber. Some bulk fiber may especially in the automotive and aerospace be used directly in this form, or it may be pro- industries [Stacey 1988]. A summary of RCF cessed to form textiles, felts, boards, cements, products and applications are provided here. and other specialty items. Other RCFs are formed into blankets as bulk fiber in the collection chamber settles onto a conveyor belt. The 3.4.1 Examples of Products blanket passes through a needle felting ma- ■ Blankets—high-temperature insulation chine that interlocks the fibers and compresses produced from spun RCFs in the form of the blanket to a specified thickness. From the a mat or blanket needler, the blanket is conveyed to a tempering oven to remove lubricants that were add- ■ Boards—high-temperature insulation pro- ed in the settling chamber. The lubricants are duced from bulk fibers in the form of a burned off, and the blanket is cut to desired size compressed rigid board (boards have and packaged. As with the bulk fiber, the RCF a higher density than blankets and are blanket may undergo additional fabrication to used as core material or in sandwich as- create other specialty products. Many of the semblies) processes are automated and are monitored by ■ Bulk RCFs—fibers with qualities of high- machine operators. Postproduction processes temperature resistance to be used as feed- such as cutting, sanding, packaging, handling, stock in manufacturing processes or other and shipping are more labor intensive, but the potential exists for exposure to airborne fibers applications for which product consisten- throughout production. cy is critical—typically in the manufacture of other ceramic-fiber-based products

■ Ropes and braids—high-temperature in- 3.4 RCF Products and Uses sulation produced by textile operations RCFs may be used in bulk fiber form or as one and used for packing, seals, and wicking of the RCF specialty products in the form of applications mats, paper, textiles, felts, and boards [RCFC ■ Woven textiles—high-temperature in- 1996]. Because of its ability to withstand tem- sulation produced by textile processes in peratures exceeding 1,000 °C, RCFs are used the form of cloth, tape, or sleeves predominantly in industrial applications, including insulation, reinforcement, and ther- ■ Papers and felts—flexible high-temper- mal protection for furnaces and kilns. RCFs ature insulation produced by papermak- can also be found in automobile catalytic ing processes and used for seals, gaskets, converters, in consumer products that oper- and other automotive and aerospace ap- ate at high temperatures (e.g., toasters, ovens, plications

20 Refractory Ceramic Fibers 3 n RCF Production and Potential for Worker Exposure

■ Vacuum cast shapes—high-temperature ■ Hot spot repair of industrial furnace insulation produced by forming special- linings ized shapes on prefabricated molds with wet fibers and then drying them by vacu- ■ Industrial furnace curtains, gaskets, and um and heat, thereby transforming bulk seals fiber into rigid, shaped products ■ Insulation of pipes, ducts, and cables as- ■ Specialties—forms (i.e., mixes, cements, sociated with high-temperature indus- and caulking compounds) that contain trial furnaces wet, inorganic binder and are used as ■ Fire protection for industrial process protective coating putties as well as ad- equipment hesives and heat and fire barriers in high- temperature applications ■ Aircraft and aerospace heat shields

■ Modules—packaged functional assembly ■ Commercial and consumer appliances con- of blanket insulation with hardware for sisting of prefabricated chimneys, pizza ov- attaching to the surfaces of furnaces and ens, self-cleaning ovens, and wood-burning kilns stoves

■ Automobile applications consisting of 3.4.2 Examples of Applications brake pads, clutch facings, catalytic con- ■ Insulation linings of high-temperature in- verters, air bags, shoulder belt controls, dustrial furnaces and related equipment and passenger compartment heat shields

Refractory Ceramic Fibers 21 Assessing Occupational 4 Exposure 4.1 Air Sampling and Both methods (listed in the NIOSH Manual of Analytical Methods [NIOSH 1998] and pro- Analytical Methods vided in Appendix A) involve using an air sam- The conventional method used to assess the pling pump connected to a cassette. The cas- characteristics and concentrations of expo- sette consists of a conductive cowl equipped sures to airborne fibers is to collect personal with a 25-mm cellulose ester membrane filter (0.45- to 1.2-µm pore size). The pump is used and environmental (area) air samples for labo- to draw air through the sampling cassette at a ratory analysis. constant flow rate between 0.5 and 16 L/min. Personal samples are the preferred method for Airborne fibers and other particulates are estimating the exposure characteristics of a trapped on the filter for analysis using microscopic methods. Methods 7400 and 7402 can worker performing specific tasks. For personal be used to count the number of fibers (and sampling, a worker is equipped with the air therefore calculate concentration based on the sampling equipment, and the collection me - volume of air sampled) and measure the fiber dium is positioned within the worker’s breath- dimensions. Fiber concentration is reported as ing zone. Area sampling is performed to evalu- the number of fibers per cubic centimeter of ate exposure characteristics associated with an air (f/cm3). Although the two methods differ in area or process. Sampling equipment for area preparation of the sampling media for analy- sampling is stationary, in contrast to personal sis, the major distinction between them is the sampling, which allows for mobility by accom- resolving capabilities of the microscope. With panying the worker throughout the sampling PCM, 0.25 µm is approximately the diameter period. of the thinnest fibers that can be observed [De- ment and Wallingford 1990]. TEM has a lower resolution limit well below the diameter of the 4.2 Sampling for Airborne smallest RCF (~0.02 to 0.05 µm) [Middleton Fibers 1982]. TEM also allows for qualitative analysis of fibers using an energy-dispersive X-ray The two NIOSH methods for the sampling and analyzer (EDXA) to determine elemental com- analysis of airborne fibers of asbestos and oth- position and selected area electron diffraction er fibrous materials are as follows: (SAED) for comparing diffraction patterns with reference patterns for identification. ■ Method 7400 describes air sampling and analysis by PCM 4.2.1 NIOSH Fiber-Counting Rules

■ Method 7402 describes air sampling and The appendix to NIOSH Method 7400 speci- analysis by TEM fies two sets of fiber-counting rules that vary

22 Refractory Ceramic Fibers 4 n Assessing Occupational Exposure

according to the parameters used to define a fi- 95% of the counts determined using the WHO ber. Under the A rules, any particle >5 µm long reference method. In studies with other SVFs, with an aspect ratio (length to width) >3:1 is Lees et al. [1993] also found that fiber expo- considered a fiber. No upper limit exists on sure estimates were slightly higher using the the fiber diameter in the A counting rules. In A rules but were comparable to the values ob- the B rules, a fiber is defined as being >5 µm tained using B rules. Breysse et al. [1999] re- long with an aspect ratio ≥5:1 and a diameter ported a similar finding when comparing RCF <3 µm. The upper-diameter limit in the B rules fiber counts determined by both A and B rules: restricts the measurement to thoracic and re- the ratio of A to B counts was 1.33. These respirable fibers. It is important to note which set sults suggest that for airborne RCF exposures, of fiber-counting criteria is used when report- most fibers with a >3:1 aspect ratio also meet ing analytical results. NIOSH recommends us- the ≥5:1 aspect ratio criterion and are <3 µm ing Method 7400 with the B rules for evalu- in diameter. ating exposures to airborne RCFs. NIOSH Method 7402 specifies use of the A rules, with a lower-diameter limit of 0.25 µm to allow com- 4.3 Sampling for Total or parison with results obtained from NIOSH Respirable Airborne Method 7400. Method 7402 can also be used to Particulates compare fiber counts obtained from Method 7400 (B rules). TEM permits the identification Airborne exposures generated during work and counting of fibers <0.25 µm in diameter; with RCFs may also be estimated by sampling 0.25 µm is the approximate resolution limit for for general dust concentrations. Sampling for PCM. particulates not otherwise regulated is described in NIOSH Method 0500 for total dust 4.2.2 European Fiber-Counting Rules concentrations and in NIOSH Method 0600 for the respirable fraction [NIOSH 1998]. Both In Europe, a slightly different fiber-counting methods (also included in Appendix A) use a convention is used. The World Health Organi- sampling pump to pull air through a filter that zation (WHO) reference method for MMMFs traps suspended particulates. NIOSH Method [WHO/EURO Technical Committee for Mon- 0600 uses a size-selective sampling apparatus itoring and Evaluating Airborne MMMF 1985] (cyclone) to separate the respirable fraction of recognizes a fiber as >5 µm long with a diameter airborne material from the nonrespirable frac- <3 µm and an aspect ratio ≥3:1. Several studies tion. The mass of airborne particulates on the comparing fiber counts determined with dif- filter is measured using gravimetric analysis, ferent counting conventions have found good and airborne concentration is determined as agreement in air sampling for RCF exposures. the ratio of the particulate mass to the volume Buchta et al. [1998] compared fiber counts of of air sampled, reported as mg/m3 (or µg/m3). air samples for RCF exposures as analyzed us- This method does not distinguish fibers from ing the NIOSH A and B rules; both methods nonfibrous airborne particles. No NIOSH REL produced similar results, with no statistically exists for either total or respirable particulates significant difference in fiber density measure- not otherwise regulated. The OSHA permis- ments on sample filters. Maxim et al. [1997] sible exposure limit (PEL) for particulates not found that fiber counts made using NIOSH otherwise regulated is 15 mg/m3 for total par- Method 7400 B rules are equal to approximately ticulates and 5 mg/m3 for respirable particulates

Refractory Ceramic Fibers 23 4 n Assessing Occupational Exposure

as 8-hr TWA concentrations. The American fiber concentrations associated with manufac- Conference of Governmental Industrial Hy- turing, handling, and using RCFs, have been gienists (ACGIH) threshold limit value (TLV) performed using industrial hygiene surveys for particles (insoluble or poorly soluble) not and air sampling techniques at multiple work- otherwise specified is 10 mg/m3 for inhalable sites. Sources of monitoring data that charac- particles and 3 mg/m3 for respirable particles terize occupational exposures to RCFs include as 8-hr TWA concentrations [ACGIH 2005]. the following:

■ University of Pittsburgh studies of expo- 4.4 Sampling for Airborne sures at RCF manufacturing sites in the Silica 1970s [Corn and Esmen 1979; Esmen et al. 1979] Because silica is a major constituent of RCFs, the potential exists for exposure to silica dur- ■ An ongoing University of Cincinnati ing work with RCFs (e.g., in manufacturing or epidemiologic study with exposure as- during removal of after-service RCF furnace sessments that use historical monitoring insulation). As with sampling for respirable data and current monitoring strategies particulates, sampling for respirable silica in- [Rice et al. 1994, 1996, 1997] volves using a pump to draw air through a cyclone before collecting respirable airborne ■ A 5-year consent agreement between the particles on a filter. Qualitative and quantita- RCFC and the U.S. Environmental Pro- tive analysis of the sample for silica content tection Agency (EPA) to monitor worker can be performed using the following analyti- exposures in RCF manufacturing plants cal methods: and in secondary users of RCFs and RCF products [RCFC 1993; Everest 1998; ■ X-ray powder diffraction (NIOSH Meth- Maxim et al. 1994, 1997, 2000a] od 7500) ■ Studies of exposure to airborne fibers ■ Visible absorption spectrophotometry during the installation and removal of (NIOSH Method 7601) RCF insulation in industrial furnaces ■ Infrared absorption spectrophotometry [Gantner 1986; Cheng et al. 1992; van (NIOSH Method 7602) den Bergen et al. 1994; Sweeney and Gil- The NIOSH REL for respirable crystalline sili- grist 1998; Maxim et al. 1999b] 3 ca is 0.05 mg/m as a TWA for up to 10 hr/day ■ International (Canadian, Swedish, Aus- during a 40-hr workweek [NIOSH 1974]. The tralian) industrial hygiene surveys of oc- 3 ACGIH TLV for crystalline silica is 0.05 mg/m cupational exposures to RCFs [Perrault as an 8-hr TWA [ACGIH 2005]. et al. 1992; Krantz et al. 1994; Rogers et al. 1997] 4.5 Industrial Hygiene ■ A study of end-user exposures to RCF in- Surveys and Exposure sulation products by researchers at Johns Assessments Hopkins University [Corn et al. 1992]

Assessments of occupational exposures, in- ■ NIOSH Health Hazard Evaluations (HHEs) cluding quantitative measurement of airborne of occupational exposures to RCFs

24 Refractory Ceramic Fibers 4 n Assessing Occupational Exposure

4.5.1 University of Pittsburgh Survey of the airborne fibers measured were <4.0 µm

of Exposures During RCF in diameter and <50 µm long with a GMD of

Manufacturing 0.7 µm and a GML of 13 µm. In the mid 1970s, researchers from the Univer- sity of Pittsburgh conducted environmental 4.5.2 University of Cincinnati Study monitoring to assess worker exposures to air- of Exposures During RCF borne fibers at domestic RCF manufacturing Manufacturing facilities. This research effort was one of the In 1987, researchers from the University of Cin- pioneering studies in the use of workplace ex- cinnati initiated an industry-wide epidemio- posure groupings or dust zones for establishing logic study of workers who manufacture RCFs. a sampling strategy [Corn and Esmen 1979]. One aim of the study was to characterize current In a series of industrial hygiene surveys, Esmen and former exposures to RCFs and silica in U.S. et al. [1979] collected 215 full-shift air samples RCF manufacturing facilities. Data from initial at three RCF manufacturing plants. Table 4–1 surveys conducted at five RCF manufacturing summarizes the sampling data for the three plants indicated airborne RCFs with a GMD facilities (A, B, and C) by fiber concentra- ranging from 0.25 to 0.6 µm and a GML ranging tion of total airborne dust. Although a wide from 3.8 to 11.0 µm [Lockey et al. 1990]. The range of values for individual samples existed airborne TWA fiber concentrations for these (<0.01 to 16 f/cm3), average (AM) concentra- five plants ranged from <0.01 to 1.57 f/cm3. Af- tions ranged from 0.05 to 2.6 f/cm3. The high- ter the first two rounds of quarterly sampling, est exposure concentrations were measured in Rice et al. [1994] had collected data from 484 manufacturing and finishing operations dur- fiber count samples (382 samples with values ing which sanding, cutting, sawing, and drill- greater than the analytic limit of detection ing operations were performed and ventilation [LOD], 39 overloaded samples, 36 samples was lacking. A large number of these opera- with values below the LOD, and 27 samples tions were noted in plant A, which is reflected voided because of tampering or pump failure). by the elevated fiber and dust concentrations They also collected 35 samples from persons for this plant. When data were compared for working with raw materials that were analyzed similar operations and dust zones, exposure quantitatively and qualitatively for respirable concentrations were consistent across plants. mass and for silica polymorphs (quartz, tridy- Analyses of air samples also included measure- mite, and cristobalite). A sampling strategy was ment of fiber dimensions. Approximately 95% developed by identifying more than 100 job

Table 4–1 . Industrial hygiene survey data for three RCF* manufacturing plants†

AM total airborne dust AM fiber concentration Plant No . samples mg/m3 Range f/cm3 Range A 76 6.05 0.37–100.00 2.6 0.02–16.0 B 67 1.6 0.19–9.73 0.63 0.04–6.7 C 72 0.85 0.05–2.34 0.05 <0.01–0.29 Source: Esmen et al. [1997]. *Abbreviations: AM=arithmetic mean; RCF=refractory ceramic fiber. †Fibers were defined as having an aspect ratio >3:1. Transmission electron microscopy was used to measure fibers ≤1 µm in diameter.

Refractory Ceramic Fibers 25 4 n Assessing Occupational Exposure

functions across 5 facilities. These job func- for 10 years of follow-up sampling (1991−2001) tions were consolidated into industry job titles at 5 of 7 facilities (2 facilities had closed before based on similarities of function, proximity to 1991). The researchers found the following es- certain processes, and exposure characteristics timates for 122 job titles still active in 2001: within designated dust zones. Table 4–2 presents median TWA exposures to airborne con- Number and % Exposure estimate 3 centrations of RCFs by job title at plants sam- of job titles (f/cm ) pled in 1987. TWA fiber concentrations ranged 97 (79%) ...... ≤0.25 from below the analytical LOD to 1.04 f/cm3 for 17 (14%) ...... >0.25 to 0.5 workers in 20 different industry job titles. Fiber 8 ( 7%)...... >0.5 concentrations obtained by rinsing the walls of the sampling cowl, where a significant number The study shows that exposures decreased for of fibers accumulated during sampling [Cor- 25% of job titles, remained stable for 53%, nett et al. 1989; Breysse et al. 1990], ranged from and increased for 22%. Of the job titles with below the analytical LOD to 1.54 f/cm3. Of the increased exposure estimates, 9 estimates were 35 samples analyzed for the silica polymorphs, >0.1 f/cm3 (range = 0.1 to 0.21 f/cm3), and 19 quantifiable silica was found in 5 samples: 4 of estimates were <0.1 f/cm3. The exposure es- the samples contained cristobalite in concen- timates for this study do not include adjust- 3 trations ranging from 20 to 78 µg/m , and 1 of ments for respirator use. the samples contained 70 µg/m3 quartz. The measurable silica exposures occurred among workers employed as raw material handlers and 4.5.3 RCFC/EPA Consent Agreement furnace operators. Monitoring Data As the study progressed, approximately 1,820 In 1993, the RCFC and the EPA entered into a ne- work history interviews were conducted and gotiated 5-year consent agreement to determine evaluated to refine uniform job titles and to the magnitude of RCF exposures in the primary identify dust zones according to the meth- RCF manufacturing industry and in secondary od of Corn and Esmen [1979]. Four years of RCF-use industries [RCFC 1993; Maxim et al. sampling data (1987–1991) were merged with 1994, 1997; Everest 1998]. Another purpose of historic sampling data to construct exposure this consent agreement was to document changes estimates for 81 job titles in 7 facilities for spec- in RCF exposures during the 5 years of the agree- ified time periods [Rice et al. 1997]. Overall ment (1993–1998). The Quality Assurance Proj- exposures decreased. The maximum exposure ect Plan in the consent agreement contains the estimated was 10 f/cm3 in the 1950s for carding analytical protocols, statistical design, descrip- in a textile operation; subsequent changes in tion of the program objectives, and timetables for engineering, process, and ventilation reduced meeting the objectives [RCFC 1993]. exposure estimates for all 20 job titles to near or below 1 f/cm3 [Rice et al. 1996, 1997]. The During each year of the consent agreement, a study reported that at more recent operations minimum of 720 personal air samples (mea- (1987–1991), exposure estimates ranged from sured as 8-hr TWAs) were collected according below the analytic LOD to 0.66 f/cm3. to a stratified random sampling plan. Of these, 320 samples were collected in RCF manufac- Subsequently, Rice et al. [2005] published the turing and processing (primary) facilities. The results from an analysis of exposure estimates remaining 400 samples were collected in RCF

26 Refractory Ceramic Fibers 4 n Assessing Occupational Exposure — — — — — — — — — — — — SD 0 0.1 0.28 0.07 0.47 0.12 0.41 0.25 3 — — — — — — — — — — — — — — 1.04 0.61 0.31 0.41 0.62 0.56 f/cm Median TWA Median Plant 5

. 9 1 3 4 2 1 2 † defined were fibers 16 — — — — — — — — — — — — No samples — — — — — — — — — — — — — SD 0.01 0.18 0.01 0.01 0.02 0.01 0.01 3 — — — — — — — — — — — — — — — 0.02 0.03 0.02 0.02 0.03 f/cm Median TWA Median Plant 4

. 2 3 3 3 1 1 2 — — — — — — — — — — — — — No samples microscopy, scanning electron For — — — — — — — SD 0.1 0.26 0.02 0.04 0.05 0.03 0.05 0.01 0.01 0.14 0.06 0.01 3 — — — — — — — — — — 0.02 0.51 0.01 0.13 0.04 0.08 0.04 0.16 0.25 0.01 Median TWA Median f/cm by industry by title job at plants sampled in 1987 * Plant 3

. 1 4 8 3 1 3 5 3 2 2 1 1 2 — — — — — — — No samples — — — — — — — — — — SD 0.05 0.94 0.03 0.03 0.04 0.03 0.02 0.04 0.04 0.06 3 — — — — — — — — — — — — — — 0.14 0.11 0.13 0.06 0.16 0.15 using transmission sized micoscopy. electron f/cm Median TWA Median Plant 2 average. TWA=time-weighted µm if

. 3 1 6 1 1 4 3 1 3 21 — — — — — — — — — — No samples 0 — — — — — — — — — — — SD 0.02 0.01 0.06 0.01 0.19 0.03 0.01 0.01 3 ≥5:1 and length >0.5 length ≥5:1 and

— — — — — — — — — — — — deviation; SD=standard 0.05 0.01 0.01 0.03 0.02 0.02 0.02 0.03 f/cm Median TWA Median Plant 1

. No Median TWA exposures to airborne to exposures of concentrations RCFs TWA Median samples — — — 12 5 3 5 7 — — — 1 — — — 6 — — 5 6 Table 4–2 . Table the furnace. the plant before [1994]. fiber; ceramic RCF=refractory et al. Rice Industry title job (furnace) ‡ with the same aspect ratio >5 µm. and length Abbreviations: of is the area Fore an aspect as having defined ratio of were Fibers Source: * † I Fore Furnace Maintenance Needler Office Ship Supervisor Textiles Utility Engineer (nonproduction) (wet) Fabrication (nonfurnace) Fore line Blanket (dry)Fabrication (wet/dry)Fabrication plant Office Plant cleanup Quality control and development Research materials Raw

Refractory Ceramic Fibers 27 4 n Assessing Occupational Exposure

customer facilities referred to as end-use (sec- A comparison of values from Tables 4–4, 4–5, ondary) facilities. The researchers collected a to- and 4–6 with those in Table 4–3 indicates that tal of 4,576 samples. A number of end-use facili- average airborne concentrations for 1993–1998 ties were randomly selected from a list of known were lower than those for the preceding base- purchasers of RCF products. The remainder con- line sampling period (1989–1993). However, sisted of facilities that volunteered for sampling a comparison of values in Tables 4–5 and 4–6 once they learned of the consent agreement. The shows that average concentrations for the entire strata from which the 720 samples were collected 5-year consent agreement monitoring period consist of eight functional job categories derived (1993–1998) are equal to those of year 5 (i.e., so that results could be aggregated for compari- no change). son across industries, facilities, and similar job After the first 3 years (1993–1996) of the con- functions [RCFC 1993]. This categorization was sent agreement monitoring period, Maxim et al. based on the approach instituted by Corn and [1997] performed interim analyses of these data Esmen [1979]. Appendix B lists definitions and combined with historical data from the baseline major work tasks for each functional job cat- monitoring period (1989–1993). The following egory. TWA and task-length average air sampling conclusions about RCF exposures were based data were gathered according to NIOSH Meth- on these analyses of data from 1,600 baseline od 7400 (B rules) and analyzed using PCM and samples and 3,200 consent agreement samples: TEM. Data on respirator use (by type) were also ■ Airborne concentrations of RCFs are collected [Maxim et al. 1998]. generally decreasing in the workplace. As background for the consent agreement ■ Ninety percent of airborne concentra- monitoring plan, baseline (now referred to as tions of RCFs in the workplace are below historical) information about airborne fiber 1 f/cm3. concentrations was obtained through personal sampling of workers at RCF manufacturing ■ RCF concentrations have an approxi- facilities from January 1989 to May 1993. Ex- mately log-normal distribution. posure monitoring strategies used during the ■ Significant differences exist in workplace baseline period (1989–1993) provided the concentrations by facility. framework for the consent agreement (1993– ■ 1998) monitoring protocol. Table 4–3 pres- Workplace concentrations vary with func- ents AM and geometric mean (GM) concen- tional job category. trations of RCF exposures determined from ■ Respirator usage varies with the worker’s historical data (1989–1993) by functional job functional job category and the associ- category. Table 4–4 contains these summary ated average fiber concentration. statistics for all 5 years of RCF consent agree- ■ Workplace samples have a lower ratio of ment monitoring data. Table 4–5 summarizes respirable nonfibrous particles to fibers data from samples collected during the 5th than samples used in initial animal inha- year of the consent agreement only (June 1997 lation studies [Mast et al. 1995a,b; Mc- to May 1998). Table 4–6 presents the average Connell et al. 1995]. airborne fiber concentrations for the baseline (1989–1993) and consent agreement monitor- Functional job categories with the highest aver- ing (1993–1998) periods by manufacturing and age TWA fiber concentrations include removal end-use sectors. (AM=1.2 f/cm3), finishing (AM=0.8 f/cm3), and

28 Refractory Ceramic Fibers 4 n Assessing Occupational Exposure

Table 4–3 . TWA* concentrations of airborne RCFs in personal samples collected during the baseline sampling period (1989–1993),† by functional job category

Manufacturing (primary production) End use (secondary processing) AM GM AM GM No . No . Functional job category samples f/cm3 SD f/cm3 GSD samples f/cm3 SD f/cm3 SD Assembly 120 0.5 0.92 0.22 3.94 130 0.29 0.36 0.13 4.08 Auxiliary 119 0.15 0.18 0.07 4.01 26 1.1 2.26 0.2 6.33 Fiber 438 0.52 0.79 0.22 4.17 — — — — — Finishing 127 0.76 0.63 0.49 3.11 84 1.57 5.72 0.47 4.18 Installation — — — — — 201 0.69 1.09 0.3 4.31 Mixing forming 89 0.27 0.34 0.15 3.23 47 0.41 0.55 0.17 4.71 Other 129 0.33 0.86 0.09 4.25 57 0.38 0.69 0.14 4.88 Removal — — — — — 49 1.36 2.97 0.28 6.48 Total 1,022 0.46 0.74 0.19 4.37 594 0.75 2.49 0.23 4.85

*Abbreviations: AM=arithmetic mean; GM=geometric mean; GSD=geometric standard deviation; RCF=refractory ceramic fibers; SD=standard deviation; TWA=time-weighted average. †Data collected from August 1989 to May 1993 [RCFC 1993].

Table 4–4 . TWA* concentrations of airborne RCFs in personal samples collected during the 5-year consent agreement monitoring period, 1993–1998,† by functional job category

Manufacturing (primary production) End use (secondary processing) AM GM AM GM No . No . Functional job category samples f/cm3 SD f/cm3 GSD samples f/cm3 SD f/cm3 GSD Assembly 362 0.28 0.27 0.18 2.76 369 0.31 0.4 0.14 4.1 Auxiliary 237 0.12 0.19 0.05 3.87 311 0.19 0.37 0.07 4.68 Fiber 421 0.26 0.47 0.14 3.27 — — — — — Finishing 359 0.65 0.56 0.47 2.44 622 0.99 2.09 0.35 4.5 Installation — — — — — 456 0.42 0.51 0.2 3.83 Mixing forming 379 0.28 0.27 0.17 2.96 332 0.31 0.47 0.17 3.07 Other 167 0.14 0.21 0.07 3.22 385 0.17 0.46 0.04 4.66 Removal — — — — — 176 1.92 2.85 0.82 4.22 Total 1,925 0.31 0.42 0.16 3.65 2,651 0.56 1.39 0.16 5.22 Source: Maxim et al. [1999a]. *Abbreviations: AM=arithmetic mean; GM=geometric mean; GSD=geometric standard deviation; RCF=refractory ceramic fibers; SD=standard deviation; TWA=time-weighted average. †Data collected from June 1993 to May 1998.

Refractory Ceramic Fibers 29 4 n Assessing Occupational Exposure

Table 4–5 . TWA* concentrations of airborne RCFs in personal samples collected during year 5 of the consent agreement monitoring period, June 1997 to May 1998

Manufacturing (primary production) End use (secondary processing)

No . AM GM No . AM GM Functional job category samples f/cm3 SD f/cm3 GSD samples f/cm3 SD f/cm3 GSD Assembly 78 0.28 0.25 0.19 2.48 92 0.28 0.39 0.1 5.32 Auxiliary 44 0.16 0.21 0.08 4.05 89 0.18 0.36 0.06 4.98 Fiber 85 0.29 0.29 0.18 2.85 — — — — — Finishing 77 0.6 0.57 0.44 2.11 126 0.93 1.49 0.37 4.43 Installation — — — — — 81 0.34 0.49 0.17 3.54 Mixing forming 75 0.23 0.24 0.14 2.78 69 0.28 0.31 0.18 2.65 Other 39 0.22 0.34 0.12 3 70 0.05 0.12 0.02 3.07 Removal — — — — — 39 2.3 3.9 0.58 6.15 Total 398 0.31 0.37 0.18 3.12 566 0.54 0.14 0.13 5.83

Source: Maxim et al. [1999a]. *Abbreviations: AM=arithmetic mean; GM=geometric mean; GSD=geometric standard deviation; RCF=refractory ceramic fibers; SD=standard deviation; TWA=time-weighted average.

installation (AM=0.4 f/cm3). The remainder of Regarding particle-to-fiber ratio, Maxim et al. the functional job categories had average TWA [1997] found average workplace values to be concentrations near or below 0.3 f/cm3. Al- much lower (0.53; n=10; range not reported) though different jobs and activities are associ- than the average ratio (9.1; n=7) in the samples ated with the three higher exposure functional used in a series of animal inhalation toxicity job categories, similarities exist that contribute studies with RCFs [Mast et al. 1995a,b, 2000; McConnell et al. 1995]. to exposure concentrations. First, removal and installation activities are performed at remote Monitoring performed during the baseline pe- jobsites where implementing fixed engineering riod (August 1989–May 1993) and the 5‑year controls may be difficult or impractical for re- consent agreement period (June 1993–May ducing airborne fiber concentrations. Removal 1998) provided data from nearly 6,200 air sam- requires more mechanical energy and may ples in the domestic RCF industry. Table 4–6 involve fracturing the structure of the RCF presents the summary statistics of workplace RCF exposure concentrations for the baseline product, resulting in fiber release and higher (historical) and consent agreement monitor- concentrations of airborne fibers. Finish- ing data. The data suggest that (1) the AMs and ing activities are performed at fixed locations GMs of RCF concentrations were higher for where it is possible to implement engineering workers during the baseline period than dur- controls, but they also involve mechanical ening the more recent (consent monitoring data) ergy to shape RCF products by drilling, sand- period, and (2) AM and GM exposure concen- ing, and sawing. These processes also result in trations were lower for workers in manufactur- the dispersal of airborne fibers. ing facilities than at end-use sites.

30 Refractory Ceramic Fibers 4 n Assessing Occupational Exposure

Table 4–6 . TWA* concentrations RCFs in personal samples collected at manufacturing facilities and end-use site during the baseline (1989–1993) monitoring periods Consent monitoring data Baseline data (1989–1993)† (1993–1998)‡ AM GM AM GM No . No . Type of site samples f/cm3 SD f/cm3 GSD samples f/cm3 SD f/cm3 GSD Manufacturing (primary 1,022 0.46 0.74 0.19 4.37 1,527 0.31 0.42 0.16 3.65 production) End-use (secondary 594 0.75 2.49 0.23 4.85 2,085 0.56 1.39 0.16 5.22 processing) Total 1,616 0.56 1.63 0.2 4.56 4,576 0.46 1.1 0.16 4.53 Sources: RCFC [1993] and Maxim et al. [1999a]. *Abbreviations: AM=arithmetic mean; GM=geometric mean; GSD=geometric standard deviation; RCFs=refractory ceramic fibers; SD=standard deviation; TWA=time-weighted average. †Data collected from August 1989 to May 1993 [RCFC 1993]. §Data collected from June 1993 to May 1998 [Maxim et al. 1999a].

4.5.4 Exposures During Installation samples (3 for respirable dust and 7 for to- and Removal of RCF Furnace tal dust concentrations). Bulk samples of the Insulation insulation materials were analyzed for cristobalite content, which ranged between 0% and To evaluate exposures to airborne dust as- 21%. In area air samples, cristobalite content sociated with removing RCF furnace insula- ranged from 4% to 15%. Personal respirable tion, Gantner [1986] conducted surveys with dust concentrations averaged 4.99 mg/m3 air sampling at five sites. The surveys were (range=0.12 to 16.9 mg/m3), and personal total performed at sites where workers removed dust samples averaged 13.95 mg/m3 (range=0.31 modules or blanket-type insulation manu- to 35.8 mg/m3). Concentrations in area samples ally using knives or trowels. During removal were lower, averaging 1.61 mg/m3 (range=0.1 activities, workers wore disposable, single-use to 3.4 mg/m3) for respirable dust and 8.98 mg/m3 respirators, disposable protective clothing or (range=0.96 to 36.2 mg/m3) for total dust. As their own personal clothing, and (in some expected, the highest cristobalite concentra- cases) goggles or other protective eyewear. tions in bulk samples were found on the face Personal sampling was performed for total of insulation materials closest to high tem- dust concentration as well as respirable dust peratures in furnaces (threshold temperature concentration using a cyclone. Area samples near 1,700 oF). Results of the surveys indicated were collected in the center of work zones that concentrations of respirable cristobalite (industrial furnaces) at 9 ft above the floor, exceeded the ACGIH TLV (then [10 mg/m3]/ which was at the breathing zone level of the [% SiO2 + 2]/2) in 75% of the samples, al- workers, who were on scaffolding. A total of though all sampling times were short because 24 air samples were collected, including 14 the removal task lasts only 26 to 183 min. The personal samples (9 for respirable dust and TLV for cristobalite has since been lowered to 5 for total dust concentrations) and 10 area 0.05 mg/m3 as an 8-hr TWA [ACGIH 2005].

Refractory Ceramic Fibers 31 4 n Assessing Occupational Exposure

Cheng et al. [1992] studied exposures to RCFs RCF insulation revealed exposures of 0.15 and during the installation and removal of RCF 0.16 f/cm3. Exposures to total particulate (18.3 insulation in 13 furnaces at 6 refineries and 2 and 22.4 mg/m3 as 8-hr TWAs) were above the chemical plants. Air samples were collected OSHA PEL of 15 mg/m3. Exposure concentra- and analyzed according to NIOSH Method tions for respirable dust containing crystalline 7400 (A rules); sampling times ranged from silica (2.4% and 4.3%) were also above the 15 to 300 min. Samples collected during mi- OSHA PEL. The elevated concentrations of nor maintenance and inspection tasks (n=27) respirable and total dust were associated with showed GM concentrations of 0.08 to 0.39 f/cm3 removal of conventional refractory lining us- (range=0.02 to 17 f/cm3). Sampling performed ing jackhammers, crowbars, and hammers. A during installation of RCF insulation (n=59) worker performing routing to install new RCF revealed GM concentrations of 0.14 to 0.62 insulation was exposed at 1.29 f/cm3 (8-hr f/cm3 (range=0.02 to 2.6 f/cm3). The highest TWA). Personal samples from another worker exposures were observed in samples collect- using a bandsaw to cut new RCF insulation reed during removal of RCF insulation (n=32), vealed concentrations of 1.02 f/cm3 as an 8-hr with GM concentrations of 0.02 to 1.3 f/cm3 TWA. (range=<0.01 to 17 f/cm3). Workers working outside of enclosed spaces (furnaces) were In the RCF industry, worker exposures to re- rarely exposed to more than 0.2 f/cm3. One spirable crystalline silica (including quartz, sample of after-service RCF insulation was cristobalite, and tridymite) may occur during analyzed for fiber diameter and length: median the use of silica in manufacturing, removal of diameter was reported as 1.6 µm (range=0.5 to after-service insulation, and waste disposal. 6 µm), and length ranged from 5 to 220 µm. Focusing on exposures of workers who in- Of 100 fibers randomly selected and analyzed stall, use, or remove RCF insulation, Maxim et from the air sample, 87% were within the re- al. [1999a] collected 158 personal air samples spirable size range. Four personal samples were analyzed for respirable quartz, cristobalite, and collected during removal of after-service RCF tridymite over the RCFC/EPA 5-year consent modules and fire bricks and were analyzed for agreement monitoring period (1993–1998). A respirable crystalline silica (cristobalite). Sam- total of 42 removal projects were sampled. For ples revealed concentrations ranging from small jobs, all workers engaged in insulation 0.03 mg/m3 to 0.2 mg/m3 (GM=0.06 mg/m3). removal were sampled; for larger jobs, workers were selected at random for sampling. Air At a Dutch oil refinery, van den Bergen et al. sampling and analysis were performed accord- [1994] performed personal air monitoring for ing to NIOSH Method 7500 for crystalline sili- airborne fibers to assess worker exposures dur- ca by X-ray diffraction; sampling times ranged ing the removal of RCF insulation from expan- from 37 to 588 min (AM=260 min, standard sion seams in a heat-treating furnace. The 8-hr deviation [SD]=129 min). Short sampling times TWA exposures for 5 workers sampled ranged reflected the short duration of RCF insulation from 9 to 50 f/cm3 (GM=16 f/cm3). Sweeney removal tasks (a benefit over time-intensive re- and Gilgrist [1998] also monitored worker ex- moval of conventional refractories). Removal posures to airborne RCFs and respirable silica of RCF blankets and modules is performed by during the removal of RCF materials from using knives, pitchforks, rakes, and water lanc- furnaces. Personal samples from two work- es, or by hand-peeling. The study noted that ers taken during the removal of after-service most (>90%) workers wear respirators (with

32 Refractory Ceramic Fibers 4 n Assessing Occupational Exposure

protection factors from 10 to 50 or more) when µm). For airborne fibers, rock wool (sprayed removing insulation. Analysis of 158 samples on) had a GMD of 2.0 µm, followed by RCFs found the following: (1.1 µm), composite RCFs and glass wool (0.71 µm), glass wool (0.5 µm) and blown rock wool ■ Fourteen samples had task-time respi- (0.5 µm). Elemental analysis and comparison rable quartz concentrations ranging from of bulk samples with air samples revealed a 3 0.01 to 0.44 mg/m (equivalent 8-hr TWA greater concentration of fibers with oxides of 3 range=0.004 to 0.148 mg/m ); the remain- silicon and aluminum in air samples. For sites der of samples were below the LOD. with either glass wool or rock wool insulation, airborne samples contained fewer fibers with ■ Three samples had detectable concentrations of cristobalite that were below the silicon oxide as the sole constituent than bulk samples. The authors concluded that airborne NIOSH REL of 0.05 mg/m3. fiber concentrations were affected by the type ■ One sample contained tridymite (0.2 mg/m3) of fiber material used and the confinement at a concentration exceeding the NIOSH of worksites. The authors also concluded that REL of 0.05 mg/m3. characterization of fibers in bulk samples is not a good representation of physical and chemical parameters of the airborne fibers. 4.5.5 International (Canadian, Swedish, A report by the Swedish National Institute and Australian) Surveys of RCF for Occupational Health [Krantz et al. 1994] Exposure describes exposure to RCFs in smelters and Perrault et al. [1992] reported on the charac- foundries based on industrial hygiene surveys teristics of fiber exposures that occurred during and sampling at 4 facilities: a specialty steel the use of synthetic fiber insulation materials on foundry (2,500 workers), a metal smelting construction sites in Canada. Fiber dimensions plant (1,500 workers), an aluminum foundry were measured from bulk samples of insulation (450 workers), and an iron foundry (450 work- materials used at five construction sites. Area air ers). RCF products were used in these plants in samples were also collected during the installa- ladles, tapping spouts, holding furnaces, heat tion of composite RCF and glass wool insula- treatment furnaces, and spill protection mats. tion, glass wool alone, rock wool (both blown Workers and contractors were placed into three and sprayed on), and RCFs alone. exposure categories, depending on their potential for exposure (as determined by distance Respirable fiber concentrations were highest from a fiber source). The highest exposures to during removal and installation of RCFs (0.39 airborne ceramic fibers (category 1) had medi- to 3.51 f/cm3) compared with concentrations an concentrations of 0.26 to 1.2 f/cm3 and in- measured during installation of rock wool volved about 3% (n=160) of the workers at the (0.15 to 0.32 f/cm3), composite RCF and glass plants surveyed. Secondary exposures (catego- wool (0.04 f/cm3), and glass wool alone (0.01 ries 2 and 3) involved another 33% (n=1,650) of f/cm3). Diameters of fibers in bulk samples dif- the workers and had median concentrations fered significantly from diameters in airborne of 0.03 to 0.24 f/cm3. During certain opera-

fibers. RCFs had the smallest GMD of fibers in tions such as removal or demolition of RCF bulk samples (0.38 to 0.55 µm) compared with materials in enclosed spaces, concentrations glass wool (0.93 µm) and rock wool (1.1 to 3.9 of up to 210 f/cm3 were measured. Total dust

Refractory Ceramic Fibers 33 4 n Assessing Occupational Exposure

concentrations increased with fiber concen- 4.5.6 Johns Hopkins University tration and were as high as 600 mg/m3 during Industrial Hygiene Surveys demolition and 60 mg/m3 during reinsulation. A report of RCF end-user exposure data pre- Median fiber diameters from bulk samples an- pared for the Thermal Insulation Manufac- alyzed by electron microscopy ranged from 0.6 to turers Association (TIMA) showed that us- 1.5 µm, which was comparable to the diameters ing blanket, bulk, and vacuum-formed RCFs of airborne fibers. On the basis of air sampling during certain operations resulted in high fi- data, fiber dose (assuming a working lifetime of ber concentrations [Corn et al. 1992]. For ex- 40 years [fiber concentration × exposure time ample, 25 personal air samples collected from per year × 40 years]) was estimated for 8 oc- workers installing RCF blanket modules had cupations with category 1 exposures. Dose es- 3 timates ranged from 0.05 fiber-years/cm3 for a an AM, 8‑hr TWA concentration of 1.36 f/cm cleaner to 85 fiber-years/cm3 for a bricklayer or (SD=1.15). The fibers were collected and ana- contractor. Dose estimates for the 6 other oc- lyzed using NIOSH Method 7400 (B rules). cupations ranged from 0.6 fiber-years/cm3 to Seventeen vacuum formers had AM expo- 3 3.1 fiber-years/cm3. sure concentrations of 0.71 f/cm (SD=0.83) while using bulk RCF products. Twenty-eight Researchers at the Australian National Occupa- workers with the job title vacuum-formed RCF tional Health and Safety Commission established cast finisher had AM exposures of 1.55 f/cm3 a technical working group to investigate typi- (SD=1.51). Table 4–7 summarizes exposure cal exposures in SVF manufacturing and user data collected for the 17 occupations sampled industries [Rogers et al. 1997]. The RCF man- during the study. Scanning electron micros- ufacturing industry is relatively small in Aus- copy (SEM) was used to measure dimensions tralia: 2 plants employing roughly 40 workers of approximately 3,500 fibers from selected air have been manufacturing RCFs since 1976 and samples of the 17 occupations. GM fiber diam- 1977. Since the plants began manufacturing eters ranged from 0.9 to 1.5 µm, and GM fiber RCFs, 152 persons have been involved with pro- lengths ranged from 20.4 to 36.1 µm. Fiber as- duction. Airborne fiber concentrations in both pect ratios based on these data ranged between plants decreased over time as a result of (1) the 16:1 and 30:1. introduction of a national exposure standard 3 of 0.5 f/cm for synthetic fibers and a second- 4.5.7 NIOSH HHEs and Additional 3 ary standard of 2 mg/m for inspirable dust, (2) Sources of RCF Exposure Data the use of various controls and handling tech- nologies, and (3) increased awareness of dust NIOSH has conducted HHEs involving poten- suppression by the workforce. GM concentra- tial exposures to RCFs at the following work tions of airborne fibers before implementa- places: an RCF manufacturing facility [Lyman tion of the synthetic fiber exposure standard 1992], a steel foundry [O'Brien et al. 1990], (1983–1990) measured 0.52 f/cm3 (geometric a power plant [Cantor and Gorman 1987], a standard deviation [GSD]=3.9) and 0.29 f/cm3 foundry [Gorman 1987], and a railroad car (GSD=2.5) for plants 1 and 2, respectively. wheel and axle production facility [Hewett GM concentrations for the subsequent period 1996]. Table 4–8 summarizes data on airborne (1991–1996) dropped to 0.11 f/cm3 (GSD=4.1) fiber concentrations and dimensions from at plant 1 and 0.27 f/cm3 (GSD=3.3) at plant 2. these studies.

34 Refractory Ceramic Fibers 4 n Assessing Occupational Exposure

Table 4–7 . Summary of 8‑hr TWA* RCF exposures for workers using RCF insulation products

Gravimetric PCM (f/cm3) SEM (f/cm3) (mg/m3) RCF product Occupation n AM SD n AM SD n AM SD

RCF blanket Module fabricator 5 0.44 0.4 7 0.54 0.8 4 6.26 6.5 Module installer 25 1.36 1.15 23 1.19 0.8 11 14.2 18.7 Blanket installer 8 0.37 0.29 9 0.33 0.24 4 1.42 1.2 Investment caster 20 0.73 0.88 18 0.65 0.57 6 3.59 3.75 General fabricator 20 0.55 0.55 19 0.46 0.55 7 0.86 0.49 Fabrication maintenance — — — — — — — — — RCF bulk Vacuum former 17 0.71 0.83 13 0.6 0.57 7 1.1 0.7 Vacuum maintenance — — — — — — — — — Vacuum warehouse — — — — — — — — — Sprayer 1 1.53 — 1 1.15 — — — — Spray feeder 1 0.24 — 1 0.21 — — — — Vacuum-formed RCFs General fabricator 2 0.52 0.58 2 0.2 0.05 2 0.57 0.35 Paper fabricator — — — — — — — — — Paper finisher — — — — — — — — — Cast finisher 28 1.55 1.51 32 1.17 1.17 8 4.05 5.42 Finishing maintenance 1 0.12 — 2 0.07 0.01 1 0.75 — Board installer 9 0.78 0.84 9 0.66 0.67 1 6.09 — Source: Corn et al. [1992]. *Abbreviations: AM=arithmetic mean; PCM=phase contrast microscopy; RCF-refractory ceramic fiber; SD=standard deviation; SEM=scanning electron microscopy; TWA=time-weighted average.

Refractory Ceramic Fibers 35 4 n Assessing Occupational Exposure

Table 4–8 . NIOSH Health Hazard Evaluations involving investigation of exposures to RCFs* Samples Concentration Reference Worksite No . Type f/cm3 SD Fiber dimension Lyman [1992] RCF manufacturing 286 Breathing zone 0.69 — — 4 Breathing zone 4.02 1.82 — 126 Breathing zone 0.81 — AMD=0.6 μm — — AML=13.8 μm 24 Breathing zone 1.65 — — O’Brien et al. [1990] Steel foundry 48 Fibers in an insu- — — D=<1.5 μm (81% of fibers) lating blanket — — L= 4-64 μm (77% of fibers) 54 Fibers in settled — — D=<0.5 μm (73% of fibers) dust — — L=4-64 μm (62% of fibers) Cantor and Gorman Power plant 4 Breathing zone 0.26 0.08 D=0.5-2.0 μm (73% of [1987] fibers) 2 Area 0.08 0.01 L=>20 μm (60% of fibers) Gorman [1987] Foundry 7 Breathing zone 0.1 0.06 D=<2 μm (96% of fibers) 5 Area 0.4 0.26 L=<20 μm (80% of fibers) Hewett [1996] Railroad car wheel 6 Breathing zone 0.024 0.012 and axle manufac- near heat — turer treatment plant 14 Breathing zone 1.44 0.84 during RCF — removal 1 Breathing zone 3.04† — — 1.7‡ — Mean D=0.71 (SD=0.44) — — Mean L=11.9 (SD=11.3)

*Abbreviations: AMD=arithmetic mean diameter; AML=arithmetic mean length; D=diameter; L=length; RCFs=refractory ceramic fibers; SD=standard deviation. †Measured by phrase control optical microscopy (PCM). ‡Measured by transmission electron microscopy (TEM).

36 Refractory Ceramic Fibers 4 n Assessing Occupational Exposure

4.5.8 Discussion manufacturing, removal operations, and installation performed by end users. Activities Recent and historical environmental monitoring data [Esmen et al. 1979; Cantor and Gor- in these three categories require additional man 1987; Gorman 1987; O=Brien et al. 1990; mechanical energy in handling RCF prod- Cheng et al. 1992; Brown 1992; Corn et al. 1992; ucts (e.g., sawing, drilling, cutting, sanding), Lyman 1992; Allshouse 1995; Hewett 1996] in- which increases the generation of airborne dicate that airborne concentrations of RCFs fibers. Removal and installation activities are include fibers in the thoracic and respirable performed at remote sites where conventional size range (<3.5 µm in diameter and <200 µm engineering strategies and fixed controls are long [Timbrell 1982; Lippmann 1990; Baron more difficult to implement. For certain oper- 1996]). Workers are exposed to these concen- ations in which airborne fiber concentrations trations during primary RCF manufacturing, are greater (such as removal of RCF products secondary manufacturing or processing, and from furnaces), jobs are performed for short end-use activities such as RCF installation periods and almost universally with the use of and removal. Sampling data from studies of respiratory protection [Maxim et al. 1998]. domestic primary RCF manufacturing sites indicate that average airborne fiber concentra- One additional consideration during work tions have steadily declined by nearly 2 orders involving RCF exposure is the potential for of magnitude over the past 2 decades. Rice et exposure to respirable silica in the forms of al. [1997] report an estimated maximum air- quartz, tridymite, and cristobalite. Although 3 borne concentration of 10 f/cm associated the potential for such exposure exists in pri- with an RCF manufacturing process in the mary manufacturing (because silica is a major 1950s. Esmen et al. [1979] recognized average component of RCFs), monitoring data indicate exposure concentrations ranging from 0.05 to that these exposures are generally low [Rice et 2.6 f/cm3 in RCF manufacturing facilities in al. 1994]. Maxim et al. [1999a] reported that the mid- to late-1970s. During the late 1980s, Rice et al. [1994, 1996, 1997] calculated aver- many airborne silica samples collected to as- age airborne concentrations in manufacturing sess exposures during installation and removal facilities that ranged from

Refractory Ceramic Fibers 37

5 Effects of Exposure 5.1 Health Effects in Animal studies report the concentration(s) to Animals (In Vivo Studies) which the animals were exposed. The distinction between administered exposure concen- The health effects of RCF exposures have been tration and received dose is important when evaluated in animal studies using intrapleural, analyzing these studies. The dose affecting the intraperitoneal, intratracheal, and inhalation target tissues is known only when the amount routes of exposure. All of these routes have dem- of fiber present in the lung is measured and onstrated the carcinogenic potential of RCFs. reported. To analyze the results of RCF stud- Chronic inhalation studies provide information ies, the number of fibers per exposure, their di- that is most relevant to the occupational route mensions, durabilities, and the delivered dose of exposure and human risk assessment. Mech- should be considered for making comparisons anistic information about fiber toxicity may also and conclusions regarding potential and rela- be derived from other types of studies. Studies tive toxicity. investigating the cellular effects of RCFs in vitro are reviewed in Section 5.2 and Appendix C. 5.1.1 Intrapleural, Intraperitoneal, and Intratracheal Studies When comparing the effects of a fiber dose in animal studies, it is possible to compare fibers Instillation and implantation studies deliver on a gravimetric basis (effect per unit weight) fibers directly to the trachea, pleural cavity, or or a fiber basis (effect per number of fibers). peritoneal cavity, bypassing some of the defense The same gravimetric dose of different fiber and clearance mechanisms that act on inhaled types may contain vastly different numbers of fibers. Implantation of fibers into either the fibers because of differences in their dimen- pleural or abdominal cavities delivers fibers di- sions. RCF1 is a relatively thick fiber compared rectly to the pleural or abdominal mesothelium, with many types of asbestos, such as chrysotile, bypassing some or all of the normal defense and a fiber commonly used as a positive control in clearance mechanisms of the respiratory tract. pulmonary carcinogenesis experiments in ani- Intratracheal instillation delivers fibers directly mals (see Table 2–2 for descriptions of RCF1, to the trachea, bypassing the upper respiratory RCF2, RCF3, and RCF4). A gravimetric dose of tract. These exposure methods do not mimic RCF1 usually contains far fewer fibers than the an occupational inhalation exposure of several same gravimetric dose of chrysotile asbestos fi- hours per day for several days per week over bers, making a direct comparison of their effects an extended period. However, one advantage difficult when the number of fibers per unit of these studies is that they allow the admin- weight is not reported. Comparison on a per- istration of a precise dose of fibers that can be fiber basis rather than a weight basis provides replicated between animals. They also permit information most applicable to occupational the administration of higher doses than may be risk assessment. obtainable by inhalation exposure.

38 Refractory Ceramic Fibers 5 n Effects of Exposure

Although the results of implantation and in- (19/23) in OM rats and 13% (2/15) and 24% stillation studies may not be directly applicable (5/21) in two groups of male hamsters. Cro- to occupational exposure and human health cidolite asbestos at 25 mg induced abdominal effects, they provide important information mesotheliomas in 80% (20/25) of OM rats and about the potential toxicity of RCFs. Experi- 32% (8/25) of hamsters. The difference in tu- ments that control fiber dimensions and other mor incidence reported by Davis et al. [1984] variables provide information about the physi- and Smith et al. [1987] may be explained in part ological characteristics relevant to fiber tox- by differences in fiber length. Eighty-three per- icity. They provide a less expensive, quicker cent of RCF fibers used by Smith et al. [1987] means to screen the potential toxicity of a fiber had a length >10 µm; 86% had a diameter than inhalation studies. <2.0 µm. Ninety percent of the ceramic aluminum silicate material used by Davis et al. [1984] Many of the implantation and instillation stud- had a length <3 µm and a diameter <0.3 µm. ies reviewed here report the administered fiber dose on a gravimetric basis rather than on a per- Pott et al. [1987] dosed female Wistar rats by fiber basis. Some studies assess the toxicity of intraperitoneal injection with 9 or 15 mg/week both RCFs and asbestos independently, which for 5 weeks with 2 ceramic (aluminum sili- allows for the comparison of these fibers on a cate) wool fibers, Fibrefrax (RCFs), and MAN gravimetric basis but not on a per-fiber basis. (Manville RCFs); total doses of 45 and 75 mg were administered, respectively. Fifty percent of Fibrefrax fibers had a length <8.3 µm and 5.1.1.1 Intraperitoneal Implantation Studies diameter <0.91 µm. Exposure to Fibrefrax fi- In intraperitoneal studies, fibers are implanted bers induced abdominal tumors (sarcomas, directly into the abdominal cavity, bypassing mesotheliomas, or carcinomas) in 68% of the the respiratory system defense and clearance rats. Fifty percent of MAN fibers had a length mechanisms that act on inhaled fibers. Al- <6.9 µm and diameter <1.1 µm. The number though the implanted fibers act on some of the of fibers in different length categories was not same target cell types as the fibers of an inhala- reported. Exposure to MAN fibers induced ab- tion exposure (such as the mesothelium), the dominal tumors in 22% of the rats. Chrysotile effects elicited in the abdominal mesothelium (UICC/B) injected intraperitoneally at a single cannot be assumed to be identical to the re- dose of 0.05, 0.25, or 1.00 mg induced abdomi- sponse of the pleural mesothelium. Table 5–1 nal tumors in 19%, 62%, or 86% of rats, re- summarizes the results of three RCF intraperi- spectively. Fifty percent of chrysotile fibers had toneal implantation studies [Davis et al. 1984; a length <0.9 µm and diameter <0.11 µm. The Smith et al. 1987; Pott et al. 1987]. A brief de- number of fibers per dose was not reported for scription of these studies follows. the ceramic fibers and asbestos. Saline induced tumors in 2% of rats. Davis et al. [1987] dosed Wistar rats with 25 mg ceramic aluminum silicate dust by intraperi- 5.1.1.2 Intrapleural Implantation Studies toneal injection. Tumors were induced in 3 of 32 rats: 2 fibrosarcomas and 1 mesothelioma. Intrapleural implantation studies permit the Smith et al. [1987] dosed Osborne Mendel investigation of the effect of RCFs directly (OM) rats and Syrian hamsters with 25 mg on the pleural mesothelium while controlling RCFs by intraperitoneal injection. Abdomi- variables such as inhalation kinetics and trans- nal mesothelioma induction rates were 83% location.

Refractory Ceramic Fibers 39 5 n Effects of Exposure

or or or or sarcomas, sarcomas, sarcomas, sarcomas,

the abdominal the abdominal the abdominal the abdominal Tumor incidence Tumor

850 days occurred tumor First postinjection. of adenomas cavity of adenomas cavity of adenomas of adenomas cavity 1 mesothelioma 32 mesotheliomas, mesotheliomas 20 abdominal (Continued) 2 fibrosarcomas 12 mesotheliomas, 31 mesotheliomas, cavity 2 mesotheliomas,

=0.9 =25.0 in animals L D * (µm) NA L=90% <3 D=90% <0.3 L=50% <8.3 D=50% <0.91 L=50% <6.9 D=50% <1.1 L=50% <0.9 D=50% <0.11 GM L=83% >10 GM D=80% <2 Fiber dimensions Fiber

Fiber dose Fiber glass) silicate (aluminum (Fibrefrax) (MAN) RCFs fibers 25 mg ceramic 9 mg (×5)=45 RCFs 15 mg (×5)=75 Manville 1 mg UICC/B chrysotile 2 ml saline (×5)=10 (Fibrefrax) 25 mg RCFs

Intraperitoneal implantation studies of RCFs

per group unspecified sex female female female female female Number and sex Number 32, 47, 54, 36, 23,

102, Table 5–1 . Table

table. Species rats Mendel rats Wistar rats Wistar Osborne

[1984] [1987] [1987] etDavis al. et al. Pott et al. Smith Reference of See at end footnotes

40 Refractory Ceramic Fibers 5 n Effects of Exposure

Tumor incidence Tumor

mesotheliomas 20 abdominal mesotheliomas 0 abdominal mesotheliomas 2 abdominal mesotheliomas 5 abdominal mesotheliomas 8 abdominal mesotheliomas 0 abdominal mesotheliomas 0 abdominal mesotheliomas 0 abdominal in animals * fibers; ceramic RCFs=refractory

10.2) 10.2)

5 5 (µm) ≤ ≤ applicable; NA=not = 0.9 =0.9 =25.0 =25.0 L D L D Fiber dimensions dimensions Fiber L=3.1 (SD, Mean L=95% NA NA L=83% >10 GM GM GM GM L=3.1 (SD, Mean L=95% NA NA D=80% <2 D=80% <2 L=83% >10 L=length; B. le Cancer/Type Contre Internationale UICC/B=Union Fiber dose Fiber =geometric mean length; L Intraperitoneal implantation studies of RCFs GM 25 mg UICC crocidolite saline 0.5 ml physiological controls Cage (Fibrefrax) 25 mg RCF 25 mg RCFs 25 mg UICC crocidolite saline 0.5 ml physiological controls Cage

male female female female male male male

25, 25, 15, 21,male 25, 25, 112, 125, per group Number and sex Number . 5–1 (Continued) Table =geometric mean diameter; D

GM le Cancer; Contre Internationale UICC=Union Species golden hamsters Syrian D=diameter;

Reference [1987] (continued)

et al. Smith Abbreviations: * deviation; SD=standard

Refractory Ceramic Fibers 41 5 n Effects of Exposure

Table 5–2 summarizes the results of the in- by intratracheal instillation once a week for trapleural study of Wagner et al. [1973]. In- 5 weeks (10 mg total). The animals were main- trapleural injection of 20 mg of ceramic fiber tained for the rest of their lives. Approximately (unspecified type) or 20 mg for each of two 50% of the RCFs were <20 µm long with a mean samples of chrysotile produced mesotheliomas fiber diameter of 1.8 µm. No primary lung tu- in 10% (3/31), 64% (23/36), and 66% (21/32) mors developed in RCF-exposed animals. These of Wistar rats, respectively. The mean ceramic animals did not have an increased incidence of fiber diameter was 0.5 to 1.0 µm. The lengths pulmonary fibrosis or tumor production com- of the chrysotile fibers were mostly <6 µm. The pared with controls; however, the rats had a sta- chrysotile fiber diameter, RCF fiber length, and tistically significant increase in bronchoalveolar number of fibers per dose were not reported, metaplasia. The median lifespan was 479 days making a direct comparison of the samples for hamsters and 736 days for rats. Hamsters difficult. (median lifespan 657 days) and rats (median lifespan 663 days) exposed to the same dosing schedule with 2 mg crocidolite asbestos had a 5.1.1.3 Intratracheal Instillation Studies statistically significant increase in bronchoalve- The technique of intratracheal instillation has olar lung tumors in 20 of 27 (74%) and 2 of 25 the advantage of affecting the same target tis- (8%) animals, respectively. The fiber numbers sues (other than the upper respiratory tract) per dose were not reported. as an inhalation exposure. Other advantages, compared with inhalation exposure, include Manville [1991] reported a statistically signifi- a simpler technique, lower cost, accurate dos- cant increase in lung tumors in Fischer rats ex- ing, and the ability to deliver materials (such posed intratracheally to 2 mg of RCF1, RCF2, as long fibers) that may not be respirable to RCF3, and RCF4 in saline [Manville 1991]. An- rodents [Driscoll et al. 2000]. The faster dose imals were terminally sacrificed at 128 weeks with interim sacrifices at 13, 26, 52, 78, and 104 rate and bolus delivery of tracheal instillation weeks. RCF1, RCF2, RCF3, and RCF4 exposure may affect the response of the lung defense resulted in adenomas or adenocarcinomas mechanisms, resulting in differences in clear- in 6 of 109 (5.5%), 4 of 107 (3.7%), 4 of 109 ance and biopersistence relative to an inhala- (3.7%), and 7 of 108 (6.5%) rats, respectively. tion exposure. Intratracheal instillation may One mesothelioma was identified in a rat ex- also produce a clumping of fibers with a result- posed to RCF2. Exposure to 0.66 mg chrysotile ing effect on fiber distribution and clearance asbestos resulted in 8 primary lung tumors in [Davis et al. 1996; Driscoll et al. 2000]. Intra- 8 of 55 rats (14.5%). The fiber dimensions and tracheal instillation results in a heavier, more numbers per dose were not reported. centralized distribution pattern; inhalation exposure results in a more evenly and widely dis- tributed pattern [Brain et al. 1976]. Table 5–3 5.1.2 Chronic Inhalation Studies summarizes the results of two RCF intratracheal instillation studies [Smith et al. 1987; In animal bioassays, administering RCFs by Manville 1991]. A brief description of these chronic inhalation most closely mimics the studies follows. occupational route of exposure. Exposure to RCFs over a time period that approximates the In the study by Smith et al. [1987], Syrian lifespan of the animal provides the most accu- golden hamsters and OM rats were dosed with rate prediction of the potential pathogenicity 2 mg of RCFs suspended in saline (Fibrefrax) and carcinogenicity of these fibers in animals.

42 Refractory Ceramic Fibers 5 n Effects of Exposure

Table 5–2 . Intrapleural study of RCFs* in animals

Number Fiber dimensions Reference Species per group† Fiber dose (µm) Tumor incidence

Wagner et al. Wistar rats 31 20 mg ceramic fibers D=0.5–1.0 3 mesotheliomas [1973] (aluminum silicate)

35 20 mg aluminum oxide Area D=<10 1 mesothelioma

35 20 mg fiberglass L=60%>20 0 mesotheliomas D=55% 2.5–7

35 20 mg glass powder Area D=<8 1 mesothelioma

36 20 mg Canadian chrysotile L=92% <6 23 mesotheliomas

32 20 mg Canadian chrysotile L=92% <6 21 mesotheliomas

*Abbreviations: D=diameter; L=length; RCFs=refractory ceramic fibers. †The sex ratio for all groups was approximately 2 male rats to 1 female rat.

Refractory Ceramic Fibers 43 5 n Effects of Exposure

(Continued) 3 adenomas 1 carcinoma 1 mesothelioma 2 adenomas 2 carcinomas 4 adenomas 4 carcinomas Tumor incidence incidence Tumor adenomas 6 lung tumors: 4 lung tumors: 4 lung adenomas 7 lung tumors: 8 lung tumors 0 lung

NR NR NR NR NR NR (µm) Fiber dimensions Fiber in animals *

Fiber dose Fiber a 10-mg/ml suspension) a 10-mg/ml suspension) a 10-mg/ml suspension) a 3.3-g/ml suspension) a 10-mg/ml suspension)

(0.2 ml of (0.2 ml of (0.2 ml of (0.2 ml of not specified) (vehicle (0.2 ml of 2 mg RCF2 2 mg RCF3 2 mg RCF4 chrysotile0.66 mg Canadian 0.2 ml 2 mg RCF1 Intratracheal studies of RCFs

male male male male male male Table 5–3 . Table 55, 109, 109, 108, 118, 107, Number and Number sex per group Species 344 rats Fischer table.

[1991] Manville

Reference of See at end footnotes

44 Refractory Ceramic Fibers 5 n Effects of Exposure tumors

Tumor incidence Tumor

tumors 2 bronchoalveolar tumors 0 lung tumors 0 lung 20 bronchoalveolar tumors 0 lung tumors 0 lung tumors 0 lung tumors 0 lung RCFs=refractory

10.2) 10.2)

5 5 ≤ ≤ NR=not reported; =0.9 =0.9 =25.0 =25.0 L D L D in animals

* Fiber dimensions (µm) dimensions Fiber GM GM NA NA L=95% NA NA GM GM mean L=3.1 (SD, L=3% >10 mean L=3.1 (SD, D=80% <2 L=83% >10 D=80% < 2 L=95%

applicable; NA=not crocidolite * L=length; Fiber dose Fiber

× 5=10 mg) (2 mg/week × 5=10 mg) (2 mg/week × 5=10 mg) (2 mg/week × 5=10 mg) (2 mg/week Saline controls controls Cage (Fibrefrax) 10 mg RCFs 10 mg UICC (Fibrefrax) 10 mg RCFs 10 mg UICC crocidolite Saline controls controls Cage

Intratracheal studies of RCFs

=geometric mean length; L female male female female female male male male GM 22, 25, 25, 25, 27, 24, per group 125, 112, . 3 (Continued) Number and sex Number –

Table 5 Table le Cancer. Contre Internationale UICC=Union =geometric mean diameter; D Species GM hamsters Syrian golden

rats Osborne Mendel deviation; SD=standard D=diameter;

[1987] et al. Smith Reference

Abbreviations: * fibers; ceramic

Refractory Ceramic Fibers 45 5 n Effects of Exposure

The effects seen in animals may be used to including 1 mesothelioma and 2 bronchoal- predict the effects of these fibers in humans, veolar tumors. No pulmonary tumors were although interspecies differences exist in re- observed in crocidolite-exposed hamsters. Ex- spiratory anatomy, physiology, and tissue sen- posure to slag wool at 10 mg/m3 (200 f/cm3) sitivity. Chronic inhalation studies provide the and several fibrous glasses at similar gravimet- best means to predict the critical disease end- ric concentrations did not result in pulmonary points of cancer induction and nonmalignant neoplasms (not shown in Table 5–4). respiratory disease that may occur in humans because of fiber exposure [McConnell 1995; Mast et al. [1995a] exposed Fischer 344 rats 3 Vu et al. 1996]. by nose-only inhalation to 30 mg/m (187±53 WHO f/cm3 RCF1, 220±52 WHO f/cm3 RCF2, Five chronic RCF inhalation studies have been 182±66 WHO f/cm3 RCF3, 153±49 WHO conducted on rats or hamsters [Davis et al. f/cm3 RCF4) of one of four types of RCFs for 1984; Smith et al. 1987; Mast et al. 1995a,b; Mc- 6 hr/day, 5 days/week for 24 months and held Connell et al. 1995]. These studies are summa- until sacrifice at 30 months. Groups of 3 to 6 rized in Tables 5–4 and 5–5 and are described animals were sacrificed at 3, 6, 9, 12, 15, 18, below. and 24 months to examine lesions and determine fiber lung burdens. Other animals were Davis et al. [1984] exposed Wistar rats by removed from exposure at the same time whole-body inhalation to 10 mg/m3 (95 f/cm3) points and held until sacrifice at 24 months. ceramic (aluminum silicate glass) dust for Positive control rats were exposed to 10 mg/m3 7 hr/day, 5 days/week for 12 months. Ninety (1.06±1.14×104 WHO f/cm3) chrysotile under percent of the exposure fibers were short (<3 similar exposure conditions. RCF fibers with a µm) and thin (<0.3 µm). The particle ratio of mean diameter of 1 µm and mean length of 20 nonfibrous particulate to fibers was 4:1. Eight to 30 µm were selected. A particle ratio of non- of 48 exposed rats (17%) developed pulmo- fibrous particulate to fiber of 1.02–1.88:1 was nary neoplasms: 1 adenoma, 3 bronchial carci- reported. Interstitial fibrosis was first observed nomas, and 4 histiocytomas. Interstitial fibro- at 6 months with RCF1, RCF2, and RCF3 and sis was observed. No pulmonary tumors were at 12 months with RCF4 exposure. Pleural observed in control animals. fibrosis was first observed at 9 months with Smith et al. [1987] exposed OM rats and Syr- RCF1, RCF2, and RCF3 and at 12 months with ian golden hamsters by nose-only inhala- RCF4 exposure. A progression in the severity tion to 10.8±3.4 mg/m3 (200 f/cm3) ceramic of pleural fibrosis was seen in animals exposed fiber (Fibrefrax) for 6 hr/day, 5 days/week for to 30 mg/m3 for 24 months and examined at 6 24 months. The ratio of nonfibrous particulate months post exposure. The incidence of total to fibers was 33:1. Exposure to RCFs did not lung tumors was significantly increased from induce pulmonary tumors in rats. One RCF- controls after exposure to RCF1, RCF2, and exposed rat and one chamber control rat de- RCF3 but not RCF4. Neoplastic disease, includ- veloped primary lung tumors. Rats exposed to ing adenomas and carcinomas, was observed in RCFs had more severe pulmonary lesions than all treatment groups: with RCF1, in 16 of 123 hamsters, and a greater percentage of rats had rats (13%); RCF2, 9 of 121 (7.4%); RCF3, 19 of fibrosis than hamsters (22% versus 1%, respec- 121(15.7%); RCF4, 4 of 118 (3.4%); and chrys- tively). Under similar conditions, exposure to otile, 13 of 69 (18.5%). Mesotheliomas were 7 mg/cm3 (3,000 f/cm3) crocidolite asbestos induced in some rats in all treatment groups: 2 produced pulmonary tumors in 3 of 57 rats, with RCF1; 3 with RCF2; 2 with RCF3; 1 with

46 Refractory Ceramic Fibers 5 n Effects of Exposure (Continued)

carcinomas mesothelioma Tumor incidence incidence Tumor 1 adenoma 3 bronchoalveolar 4 histiocytomas 8 benign 8 malignant 1 peritoneal 11 benign 9 malignant 8 adenomas 8 carcinomas 2 pleural mesotheliomas 4 adenomas 5 carcinomas 3 pleural mesotheliomas 8 pulmonary tumors: pulmonary tumors No tumors: Nonpulmonary tumors: 16 lung tumors: 9 lung 16 nonpulmonary tumors: 0.61) 0.69) 17.0) 15.5) Fiber dimensions (µm) dimensions Fiber in animals L~90%<3 D~90%<0.3 NA mean L=22.3 (SD, mean L=18.7 (SD, mean D=0.98 (SD, mean D=1.07 (SD, * 3 3

3 3 RCF1 RCF2 3 3 f/cm † WHO f/cm WHO 3 Fiber dose Fiber fibers ceramic 5.2) mg/m 4.5) mg/m 52) f/cm 35) total f/cm 45) total 3 53) glass) silicate (aluminum 10 mg/m 95 fibers/cm Control 187 (SD, 220 (SD, 29.1 (SD, 234 (SD, 268 (SD, 28.9 (SD,

Chronic inhalation studies of Chronic RCFs per group male

male unspecified sex unspecified sex Table 5–4 . Table Number and sex Number 48, 123, 40, 121, table. Species rats Wistar 344 rats Fischer

[1984] [1995a] Reference of See at end footnotes etDavis al. et al. Mast

Refractory Ceramic Fibers 47 5 n Effects of Exposure (Continued)

Tumor incidence incidence Tumor 10 adenomas 9 carcinomas 2 pleural mesotheliomas 2 adenomas 2 carcinomas 1 pleural mesothelioma 4 adenoma 1 carcinoma 1 mesothelioma 1 adenoma 1 carcinoma tumors: 19 lung tumors: 4 lung adenomas 2 lung adenomas 2 lung tumors: 2 lung adenoma 1 lung tumors: 5 lung

17.93) 17.08) 16.87) 0.7) 0.7) 0.73) 0.72) 0.72) 17.9) 9.9) in animals * Fiber dimensions (µm) dimensions Fiber mean L=24.2 (SD, mean D=1.05 (SD, mean L=12.7 (SD, mean D=1.38 (SD, NA mean L=19.88 (SD, mean L=20.54 (SD, mean L=20.11 (SD, NA mean D=1.03 (SD, mean D=1.04 (SD, mean D=1.06 (SD, 3 3 3† 3 3 3 3 RCF3 RCF4 RCF 3 3 3 RCF1 3 3 3 3 3 WHO f/cm WHO f/cm WHO f/cm WHO f/cm WHO f/cm 7.0) mg/m 7.8) mg/m 1.1) mg/m 66) 49) 35) f/cm 44) total f/cm 48) total 37) f/cm Fiber dose Fiber 0.4) mg/m 0.7) mg/m 12) 34) f/cm 35) 17) f/cm Chronic inhalation studies of Chronic RCFs 182 (SD, 153 (SD, 26 (SD, 91 (SD, 75 (SD, 120 (SD, 213 (SD, 206 (SD, 36 (SD, 162 (SD, 29.2 (SD, 30.1 (SD, only Air 3.0 (SD, 8.8 (SD, 16.5 (SD, only Air

male male male male male male male

121, 118, 130, 131, 134, 132, 132, per group Number and sex Number . 5–4 (Continued) Table table. Species 344 rats Fischer

Reference [1995b] et al. Mast of See at end footnotes

48 Refractory Ceramic Fibers 5 n Effects of Exposure

Tumor incidence incidence Tumor None None None None None None None tumor 1 bronchoalveolar 42 pleural mesotheliomas 1 mesothelioma 1 mesothelioma tumors 2 bronchoalveolar 6.7) 0.63) 0.06) 2.71) 5 5 fibers; ceramic RCFs=refractory ≤ in animals ≤ * =25 =0.9 =0.9 =25 L D L D Fiber dimensions (µm) dimensions Fiber mean L=22.12 (SD, mean D=0.94 (SD, mean L=1.68 (SD, NA GM GM L=95% NA NA GM GM L=91% NA NA mean D=0.09 (SD, 3 applicable, NA=not 3 3 Canadian 3 WHO f/cm f/cm 4 3 L=length;

3 3 WHO f/cm (Fibrefrax), RCFs (Fibrefrax), RCFs 3 3 3 3 Fiber dose Fiber RCF1 chrysotile,

3 3 UICC crocidolite, UICC crocidolite, 56) 58) f/cm 1.4)×10 9.0)×10 3 3 200 f/cm 3,000 f/cm 200 f/cm 3,000 f/cm controls Negative controls Chamber controls Room controls Room 215 (SD, 3.0 (SD, 10 mg/m 10.8 mg/m 7 mg/m controls Chamber 7 mg/m 30 mg/m 10.8 mg/m 256 (SD, 8.4 (SD, Chronic inhalation studies of Chronic RCFs

=geometric mean length; L GM per group male female male male male female female female

male male male Number and sex Number and aspect ratios >3:1 [WHO 1985]. Organization. Health WHO=World

106, 58, 125, 49, 55, 57, 112, 59, 58, 102, 70,

. 5–4 (Continued) Table >5µm, lengths =geometric mean diameter; D Species GM hamsters rats hamsters Syrian golden Syrian golden Osborne-Mendel

D=diameter;

Reference [1995] [1987] et al. McConnell et al. Smith

<3µm, diameters have WHO fibers Abbreviations: * † le Cancer; Contre Internationale UICC=Union

Refractory Ceramic Fibers 49 5 n Effects of Exposure (Continued) dry fibers/mg RCF1 lung dry fibers/mg RCF2 lung dry fibers/mg RCF3 lung dry fibers/mg RCF4 lung 5 5 5 5 n=4 rats µg Range=2,800–6,800 Lung fiber burden fiber Lung 3.70×10 9.58×10 2.57×10 5.95×10 µg Mean=4,130

RCF3) by inhalation by * collagen alveolar and alveolar (RCF1) 6 months (RCF2, 9 months ( RCF4) 12 months proteinosis macrophages microgranulomas, deposition, deposition macrophage at seen were of areas Large bronchiolization, Alveolar Lung changes Lung and aggregated Foamy

RCF3) RCF3) RCF2, area lung fibrosis sacrifice (n=6 rats)=5.0% Range=0.2%–14.5% (RCFs) 3 months (RCF2, 6 months (RCF4) 12 months (RCF1, 9 months (RCF4) 12 months at seen fibrosis Interstitial Fibrosis at final fibrosis Interstitial % of Mean at seen Pleural fibrosis peribronchiolar Minimal Nontumor lung effects in animals exposed RCFs effects to lung Nontumor Table 5–5 . Table Species rats Wistar 344 rats Fischer table.

[1995a] [1984] Reference of See at end footnote et al. Mast etDavis al.

50 Refractory Ceramic Fibers 5 n Effects of Exposure

between fibers fibers 4 4 animals with long Lung fiber burden fiber Lung and in lung occurred increases weights. lung/body lungs of periods. recovery 9 and 12 months. with decreasing burdens RCF and widths.lengths from quickly cleared were RCFs decreased animals showed Recovery 2.18±0.99×10 0.86±0.45×10 off leveled burdens RCF Dose- and time-dependent

by inhalation by

and at group dose group 3 3 Lung changes Lung 1/55=2% 16-mg/m 3 months groups dose other at 6 months 16-mg/m microgranulomas, deposition; collagen macrophage alveolar aggregation 2/69=3% in less severe but Similar in all 3 groups Progression in bronchiolization Alveolar

in Microgranulomas bronchiolization, Alveolar metaplasia, Bronchoalveolar metaplasia, Bronchoalveolar

group such group 3 3 Fibrosis Nontumor lung effects in animals exposed RCFs effects to lung Nontumor at 12 months seen fibrosis in 16-mg/m in 16-mg/m that it was irreversible at irreversible were fibrosis final sacrifice. as dif- 12 months after raised to fused thickening nodules. showed 12 and 18 months severity. increased interstitial Reversible plateaued Pleural fibrosis Pleural and interstitial between seen Fibrosis 12/55=22% 1/70=1% was minimal Pleural fibrosis progressed fibrosis Interstitial

. 5 (Continued) – Species rats hamsters hamsters 344 rats Fischer Syrian golden Osborne-Mendel Syrian golden Table 5 Table

Reference [1995b] [1995] [1987] et al. McConnell et al. Smith

et al. Mast Abbreviation: RCFs=refractory ceramic fibers. Abbreviation: *

Refractory Ceramic Fibers 51 5 n Effects of Exposure

RCF4; and 1 in the chrysotile exposure group. by 21 months of recovery, after 12 months of All mesotheliomas were detected at or after 24 exposure, and after 24 months of exposure to months of exposure. Most RCF fibers recovered all doses of RCFs. Animals exposed for 3 or 6 in the lung were 5 to 10 µm long regardless of months and then allowed to recover until sac- exposure time and recovery time. An 80% re- rifice at 24 months had lung burdens reduced duction in fiber lung burden was seen in rats by 96% to 97% compared with animals not al- allowed to recover for 21 months following 3 lowed recovery time. months of RCF exposure. McConnell et al. [1995] exposed Syrian golden Mast et al. [1995b] exposed Fischer 344 rats by hamsters by nose-only inhalation to 30 mg/m3 nose-only inhalation to 0 (air), 3, 9, or 16 mg/m3 RCF1 (256±58 WHO f/cm3) for 6 hr/day, 5 (0, 26±12, 75±35, or 120±35 WHO f/cm3) RCF1 days/week for 18 months and held them until for 6 hr/day, 5 days/week for 24 months and sacrifice at 20 months. Positive control animals held them until sacrifice at 30 months. Fibers were exposed to 10 mg/m3 (8.4±9.0×104 WHO were selected by size as in Mast et al. [1995a]. f/cm3) chrysotile asbestos. Groups of 3 to 6 A particle ratio of nonfibrous particulate to fi- animals were sacrificed at 3, 6, 9, 12, 15, and bers of 0.9–1.5:1 was reported. Groups of 3 to 6 18 months to examine lesions and determine animals were sacrificed at 3, 6, 9, 12, 18, and 24 fiber lung burdens. Other animals were re- months to examine lesions and determine fi- moved from exposure at the same time points ber lung burdens. Other animals were removed and held until sacrifice at 20 months. Intersti- from exposure at the same time points and held tial and pleural fibrosis were first observed after until sacrifice at 24 months. Interstitial fibrosis 6 months of exposure in RCF-exposed ham- was observed after 12 months of exposure in the sters. No pulmonary neoplasms developed. 9- and 16-mg/m3 exposure groups. Pulmonary Forty-two of 102 (41.2%) RCF-exposed ani- fibrosis was first observed after 12 months with mals developed pleural mesotheliomas. Most 16 mg/m3 exposure and after 18 months with mesotheliomas developed after 18 months of 9 mg/m3 exposure. The mean Wagner grades of exposure. Animals exposed to chrysotile de- pulmonary cellular change and fibrosis in rats veloped a more severe interstitial fibrosis and exposed to 0, 3, 9, 16, and 30 mg/m3 of RCFs pleural fibrosis than those exposed to RCFs. for 24 months were 1.0, 3.2, 4.0, 4.2, and 4.0, No neoplasms were observed in the lungs or respectively. Rats exposed at the same range of pleura of the chrysotile-exposed or air control doses for 24 months and allowed to recover for animals. The greatest percentage of retained 6 months had mean Wagner grades of 1.0, 2.9, fibers had lengths of 5 to 10 µm and diam- 3.8, 4.0, and 4.3. The severity of interstitial and eters <5 µm in the lungs after 6 months of ex- pleural fibrosis was similar between those ani- posure followed by 12 months of recovery. mals sacrificed at 24 months and those allowed 6 months of recovery following the 24 months McConnell et al. [1999] conducted a multidose of exposure. The incidence of pulmonary neo- chronic study of the effects of amosite inhala- plasms was not statistically different from the tion in hamsters. The data can be compared controls in all exposure groups. One pleural with the effects of RCF1. Syrian golden ham- mesothelioma was observed in the 9-mg/m3 sters were exposed to 0.8 (36±23 WHO f/cm3), exposure group. A dose-related increase oc- 3.7 (165±61 WHO f/cm3), or 7 mg/m3 (263±90 curred in fiber lung burden. Fiber lengths of 5 WHO f/cm3) amosite asbestos. Pleural meso- to 10 µm were most prevalent in the lung fibers thelioma incidences of 3.6%, 25.9%, and 19.5%, recovered after 3 months of exposure followed respectively, were reported. The aerosol mean

52 Refractory Ceramic Fibers 5 n Effects of Exposure

diameter of the amosite asbestos was 0.60 µm the plausible carcinogenic potency estimates ±0.25; its aerosol mean length was 13.4 µm for RCFs relative to amosite, based on hamster ±16.7. The dimensions of this asbestos fiber mesotheliomas, range from about half to near- were more similar to those of the RCFs used in ly twice the carcinogenicity of amosite. the chronic inhalation studies of McConnell et al. [1995] than the chrysotile asbestos used as the positive control in that same study. 5.1.3 Discussion of RCF Studies in Animals NIOSH [Dankovic 2001] analyzed the hamster data from the RCF [McConnell et al. 1995] and The intrapleural, intraperitoneal, and intratra- amosite studies [McConnell et al. 1999]. A dose- cheal RCF studies have demonstrated the car- response model was developed for amosite and cinogenicity of RCFs. Because of the nonphysi- was used to predict the amosite response at the ologic delivery of fibers by these methods, it is one and only dose at which RCFs were tested in difficult to compare their results with those hamsters. The modeled amosite response was of an inhalation exposure. Although tracheal compared with the observed RCF response. instillation may result in different distribu- These results are presented in Figures 5–1 and tion patterns than an inhalation exposure, this 5–2. Log-probit, log-logistic, multistage, and route of exposure is useful as a screening test unrestricted Weibull models were analyzed. for relative toxicity and to compare the toxicity The transformation for the log-probit and of new materials with the toxicity of materi- log-logistic models was log (fibers/cm3 +1). als for which data already exist [Driscoll et al. The dose metric of the multistage and Weibull 2000]. Tracheal instillation also is useful when models was fibers/cm3, as they did not require testing fibers respirable by humans but not ro- a log-transformation. Results of the log-probit dents. Chronic inhalation studies provide the model analysis of these data indicated RCF/ data most relevant to occupational exposure to amosite relative potency estimates of 1.85 and RCFs. 1.19, using WHO fibers and fibers >20 µm as The RCF chronic animal inhalation studies the dose metric, respectively. The model fits described above allow for the comparison of were poor when the amosite high-dose group health effects of exposure to different doses of and 20 µm-fiber dose were included. Sensitivity RCF1, different types of RCFs, and the inter- analyses in which the high-dose amosite group species susceptibility of the rat and hamster to was dropped suggest that the relative potency RCF exposure. of RCFs to amosite could be as low as 0.66 based on the log-probit model. Results using Results of the multidose chronic inhalation the log-logistic, multistage, and Weibull mod- testing of RCF1 in rats indicate the pathogen- els were similar to those using the log-probit ic potential of RCFs at high doses [Mast et al. model, with an overall range of RCF/amosite 1995a,b]. The incidence of total lung tumors relative potency estimates from these models was significantly increased from controls after using all four amosite dose groups of 1.03 to exposure to 30 mg/m3 RCF1, RCF2, and RCF3 1.89. Although no clear toxicologic basis exists but not RCF4. A dose-response relationship was for disregarding the high-dose amosite data, demonstrated for nonneoplastic pulmonary sensitivity analyses based on excluding these changes in rats exposed to 3, 9, and 16 mg/m3 data suggest that the potency of RCFs relative RCFs. The severity of interstitial and pleural fi- to amosite could be as low as 0.47, based on the brosis was similar between those animals sacri- multistage model. These models indicate that ficed at 24 months and those allowed 6 months

Refractory Ceramic Fibers 53 5 n Effects of Exposure

Exposure concentration (f/cm3)

Figure 5–1 . Proportion of hamsters with mesotheliomas following exposure to amosite or RCFs. Con- centrations are based on fibers >20µm long. The 95% confidence limits are based on assuming a binomial distribution. Dashed lines represent the log-probit model fitted to the amosite data [Dankovic 2001]. (Source: McConnell et al. [1995, 1999].)

Exposure concentration (f/cm3) Figure 5–2 . Proportion of hamsters with mesotheliomas following exposure to amosite or RCFs. Con- centrations are based on WHO fiber dimension criteria. The 95% confidence limits are based on assuming a binomial distribution. Dashed lines represent the log-probit model fitted to the amosite data [Dankovic 2001]. (Source: McConnell et al. [1995, 1999].)

54 Refractory Ceramic Fibers 5 n Effects of Exposure

of recovery following the 24-month exposure. but high incidences of mesothelioma occurred Spontaneous primary pulmonary mesothelio- at doses of 125 and 250 f/cm3 [McConnell et al. mas are rare in rats [Analytical Sciences Incor- 1999]. Many of the mesotheliomas in the more porated 1999]. Therefore, the presence of any recent hamster studies were identified only on mesothelioma in treated animals is biologically microscopic examination [Mast et al. 1995a; significant and warrants caution. McConnell et al. 1995, 1999]. Previous studies reporting mesotheliomas only by macroscopic Comparing the chronic effects of RCF1 with identification may have underestimated the its positive control, chrysotile asbestos, in the mesothelioma incidence. Recent, short-term hamster is difficult because of the differences inhalation studies indicate that hamster me- in dose, dimensions, and durability of the two sothelial cells may have a more pronounced fibers tested [McConnell et al. 1995]. More re- inflammatory and proliferative response to cent dose-response data on amosite asbestos RCF1 exposure than those of rats [Everitt 1997; provide a comparison because these amosite Gelzleichter et al. 1996a,b, 1999]. The reasons fiber dimensions more closely resemble those for this species difference in response to RCFs of RCF1 [McConnell et al. 1999]. The mean have not been explained. The results of these lengths of the RCFs and amosite asbestos fi- animal studies indicate the need for the inclu- bers were 22.1 (±16.7) and 13.4 (±16.7) µm, sion of the hamster as a sensitive test species in respectively. Forty-three percent of RCF fi- those studies in which pleural mesothelioma is bers and ~26% of amosite asbestos fibers an endpoint of concern. were longer than 20 µm. The mean diameters of the RCFs and amosite asbestos fibers were Results from Mast et al. [1995a] indicate that 0.94 (±0.63) and 0.60 (±0.25) µm, respectively. under the conditions studied, exposure to Interstitial and pleural fibrosis were seen much RCF4 may have a less pronounced effect on earlier with amosite exposure than with RCF pulmonary pathology than exposure to RCF1, exposure. RCF exposure at 215 (±56) WHO RCF2, and RCF3. Rats exposed to RCF4 did f/cm3 resulted in mesotheliomas in 42 of 102 not have a significant increase in total lung tu- (41%) hamsters. Amosite asbestos exposure at mors compared with controls; those exposed 263 (±90) WHO f/cm3 resulted in mesothelio- to RCF1, RCF2, and RCF3 did. Exposure to mas in 17 of 87 (19.5%) hamsters. Modeling RCF4 produced a less severe fibrosis than was of these data indicates that the plausible car- seen in the other RCF exposure groups. Differ- cinogenic potency estimates for RCFs relative ences in the dimensions or physical properties to amosite, based on hamster mesotheliomas, of RCF4 may explain its different respiratory range from about half to nearly twice the car- effects from RCF1, RCF2, and RCF3. RCF4 cinogenicity of amosite [Dankovic 2001]. Dif- was produced by heating RCF1 in a furnace ferences in the physical characteristics and at 2,400 ºF for 24 hr. This Aafter-service@ fiber biopersistence of RCF1 and amosite asbestos contained approximately 27% free crystal- must be considered before extrapolating these line silica. Silicotic nodules were observed in animal data to human risk. the RCF4-exposed animals. RCF4 fibers were shorter (~34% between 5 and 10 µm ) and Hamsters showed a greater susceptibility to thicker (~35% <0.5 µm) than those of RCF1, mesothelioma induction after RCF1 exposure RCF2, and RCF3. than did rats under similar exposure conditions [Mast et al. 1995a; McConnell et al. 1995]. The particle content of the RCF test material Chronic inhalation studies of amosite asbestos may have been responsible for some of the re- in hamsters showed no pulmonary neoplasms, spiratory pathology observed in these studies.

Refractory Ceramic Fibers 55 5 n Effects of Exposure

However, an analysis of the ratio of nonfibrous animals. Building on the concept of lung over- to fibrous particulates in the reviewed studies load (first advanced by Bolton et al. [1983]), does not indicate a correlation between the Mast et al. [2000] considered particulate coex- particulate content and observed effects. Smith posure (i.e., nonfibrous particulate or shot) to et al. [1987] performed testing with the high- be a confounding factor that may have had a est particulate to fiber ratio at 33:1 and did major effect on the observed chronic adverse not report a high tumor incidence. Compar- effects. The authors propose that the MTD was ing studies based on the ratio of nonfibrous exceeded at the highest exposure concentration particulates to fibers is complicated by differ- of 30 mg/m3 for RCF1 in the rat bioassay. ences among the studies in fiber preparation, The concept of pulmonary overload in the doses tested, fiber dimensions, and methods Fischer 344 rats is based on the recognition that of fiber analysis. The techniques used to de- excessive particulate exposures (>1,500 µg/rat, tect and measure nonfibrous particulates have according to Bolton et al. [1983]) eventually improved over time so that the comparison of reduce the clearance effectiveness of the lungs, recent and older studies may reflect these in- causing the normal linear clearance kinetics to consistencies. follow a nonlinear pattern. On a cellular level, These chronic RCF inhalation studies indicate the overload conditions may result in alveolar the ability of RCFs to induce cancer in two lab- macrophages becoming engorged with par- oratory species—mesotheliomas in hamsters ticulate, pulmonary and alveolar inflamma- and pulmonary tumors in rats. The late onset tion, increased translocation of particles to the of tumors indicates the importance of chronic interstitium and lymph, granuloma formation, studies on the effects of RCF exposure. Short- pulmonary fibrosis, and lung tumors, depend- term intraperitoneal, intrapleural, intratrache- ing on the time and severity of the overload al, and inhalation studies provide important [Mast et al. 2000]. Ambiguity about the defini- information about the action of fibers, the fi- tion of MTD for chronic inhalation studies with ber characteristics associated with toxicity, and animals was also a concern expressed by the au- potential toxicity. Currently it is only through thors. One reference [Morrow 1986] recognizes lifespan toxicologic testing of animals that the the MTD as that which causes “a significant respiratory and other chronic health effects of functional impairment of lung clearance.” At a RCFs can be accurately assessed. National Toxicology Program (NTP) workshop on establishing exposure concentrations for inhalation studies in animals, it was concluded 5.1.4 Lung Overload that the highest exposure concentration should Argument Regarding produce only minimal changes in lung defense Inhalation Studies in Animals mechanisms as measured by clearance [Lewis et al. 1989]. At a similar workshop convened by Mast et al. [2000] published a review interpret- the EPA, it was proposed that the MTD for fi- ing the results of chronic inhalation studies of ber inhalation studies is equivalent to the lung RCF1 in rats and hamsters [Mast et al. 1995a,b; dose produced at the maximum achievable con- McConnell et al. 1995]. In the review, the au- centration (MAC) [Vu et al. 1996]. The MAC thors suggest the possibility that the maximum is calculated as the highest fiber concentration tolerated dose (MTD) may have been exceeded based on a 90-day study that results in signifi- and that lung overload may have compromised cant changes in alveolar macrophage clear- the pulmonary clearance mechanisms of test ance rates, lung burden normalized to exposure

56 Refractory Ceramic Fibers 5 n Effects of Exposure

concentration, cell proliferation, inflammation, concentrations were 130 fibers/ml >20 µm for lung weight, and other measures. RCF1 and 125 fibers/ml >20 µm for RCF1a. The nonfibrous content of RCF1 was approxi- The methodology described for the RCF chron- mately 25%, whereas the nonfibrous content ic inhalation studies involved procedures (i.e., of RCF1a was 2%. The mean diameter of the wet cyclone separation technology) for remov- nonfibrous particles was 2 to 3 µm. The aero- ing the nonfibrous particulate fraction from sol exposure to RCF1 contained twice as many the commercial fiber (RCF1) used for the in- short fibers (<20 µm) as RCF1a and twice the halation exposures [Mast et al. 1995a,b 2000; amount of dust (fibers and nonfibrous dust/ McConnell et al. 1995]. This process resulted mg · m3) as RCF1a (51 versus 25.8 mg/m3). At the in an aerosol with a 9.1:1 particle-to-fiber ratio end of the inhalation period, animals exposed [Maxim et al. 1997; Mast et al. 2000], compared to RCF1a had a higher pulmonary concentra- with a study by Smith et al. [1987], which re- tion of long fibers but lower concentrations of ported 33 nonfibrous particles per fiber in short fibers and nonfibrous particles. The dif- airborne exposures. Results from Esmen et al. ference in particle content was enhanced in the [1979] indicate that despite a poor correlation lungs—15 times more particles were found in between mass of total airborne dust and fiber the lungs of the RCF1-exposed animals than concentration in RCFs measured in manufac- in those exposed to RCF1a. In the aerosol ex- turing, fibers generally constitute only a small posure, only an eightfold difference was found portion of the total dust. This finding is con- in the number of particles between RCF1 and sistent with other reported measures of occu- RCF1a. The RCF1a-exposed animals had a pational exposures to airborne RCFs [Krantz half-time alveolar clearance of 80 days (71–91) et al. 1994; Trethowan et al. 1995]. However, compared with 60 days (49–77) for the con- Maxim et al. [1997] reported an average trols; clearance half-time for exposed RCF1 particle-to-fiber ratio of 0.53:1 (n=10, range animals was 1,200 days (573-infinity) com- not reported), or roughly 1 particle to 2 fibers pared with 66 (58–88) for the corresponding in RCF manufacturing facilities. controls. To evaluate respiratory inflamma- Muhle and Bellmann [1996] conducted a 5-day tion, bronchoalveolar lavage (BAL) measure- inhalation study with Fischer 344 rats to mea- ments (lactose dehyrdogenase [LDH], γ-gluta- sure the biopersistence of RCF1 (with the 9:1 myltransferase [γ –GT], total protein, reduced particulate-to-fiber ratio) and RCF1a (RCF1 glutathione [GSH]) were taken at the end of that is further processed to reduce particu- the 3-week study period and at subsequent in- late mass). The study showed a 1.5-fold longer tervals over the next 12 months. Immediately time-weighted half-life for RCF1 (t =78 days) following the 3-week inhalation study, all BAL 1/2 measurements were statistically elevated in compared with RCF1a (t1/2=54 days). That study also involved a 3-week inhalation experi- both RCF1 and RCF1a animals. However, after ment with Fischer 344 rats, in which the clear- 91 days of recovery, the BAL measurements for RCF1a animals returned to normal. Indications ance of RCF1 (t1/2=103 days) was almost twice as long as that of RCF1a (t =54 days). of inflammation continued for RCF1 through 1/2 the entire observation period. The greater and In a follow-up study by Brown et al. [2000], more persistent inflammation seen with RCF1 female Wistar rats were exposed to RCF1 and was attributed to the greater mass of material RCF1a by inhalation for 3 weeks and followed or to increased activity of the nonfibrous par- for 12 months to evaluate alveolar macrophage ticles, although the high concentration of short clearance and inflammation. The exposure fibers in RCF1 (twice that of RCF1a) could

Refractory Ceramic Fibers 57 5 n Effects of Exposure

have contributed to the observed impedance duction of RCFs and RCF products, however, in alveolar macrophage clearance and inflam- the nonfibrous particulate fraction is associ- mation. ated with the fiber, as shown in Table 2–1 (i.e., 20% to 50% of RCFs by weight is nonfibrous Tran et al. [1997] examined how overloading particulate). This suggests that occupational the alveolar macrophage defense system affects exposures to airborne RCFs necessarily involve the clearance of fibers versus that of nonfibrous coexposures to a fraction of nonfibrous par- particles. Modeling was performed based on ticulate, a suggestion that has been supported data for rats exposed by inhalation to titanium by exposure assessment studies [Esmen et al. dioxide (TiO ) at 1, 10, and 50 mg/m3 or to 2 1979; Krantz et al. 1994; van den Bergen et al. glass wool (MMVF10) at 3, 16, and 30 mg/m3. Lung burdens and clearance kinetics during 1994; Trethowan et al. 1995; Maxim et al. 1997; exposure (0 to 100 weeks) were compared with Mast et al. 2000]. those at 3, 10, and 38 days post-exposure. The models showed that overloading of the lung by fibers or nonfibrous particles are similar when 5.2 Cellular and Molecular fibers are short (<15 µm). This observation Effects of RCFs (In Vitro is plausible, as nonfibrous particles and short Studies) fibers smaller than the diameter of the alveolar macrophage are most readily engulfed and The cellular and molecular effects of RCF ex- cleared via the macrophages. When this de- posures have been studied with two different fense is overwhelmed (lung burden ≥10 mg), objectives. One purpose of these in vitro stud- these particles are cleared less effectively. For ies is to provide a quicker, less expensive, and fibers longer than 15 µm, phagocytosis by alve- more controlled alternative to animal toxicity olar macrophage is reduced. As fiber length in- testing. These experiments are best interpret- creases, fibers tend to be cleared by dissolution ed by comparing their results with those of in and disintegration of long fibers into shorter vivo experiments. The second objective of in fibers or fragments. Therefore, clearance of vitro studies is to provide data that may help long fibers is not affected by the overloading of to explain the pathogenesis and mechanisms macrophage-mediated defenses with shorter of action of RCFs at the cellular and molecu- fibers or nonfibrous particles. lar levels. These cytotoxicity and genotoxicity studies are best interpreted by comparing the The exposure concentrations for the RCF effects of RCFs with those of other SVFs and chronic inhalation bioassays were measured asbestos fibers. In vitro studies serve as screen- and reported as mass in mg/m3. Monitoring of ing tools and provide insights into the molecu- exposures as performed by gravimetric analy- lar mechanisms of fibers. They are an impor- sis does not distinguish fibers from nonfibrous tant complement to animal studies. Currently particulate, although fiber concentration and dimensions were also checked by phase contrast it is not possible to use these data to derive the and electron microscopy [Mast et al. 1995a,b]. NIOSH REL for RCFs. For this reason, a dis- Consequently, the particulate fraction was in- cussion of in vitro studies is included here, but cluded in the dose measurements. This fact the more comprehensive summaries of studies does complicate efforts to compare the relative are included in Appendix C. toxicity of fibers, nonfibrous particulate, and total combined particulate, especially regard- The toxicity of fibers has been attributed to ing the lung overload hypothesis. During pro- their dose, dimensions, and durability. Any test

58 Refractory Ceramic Fibers 5 n Effects of Exposure

system that is designed to assess the potential cytotoxicity. Hart et al. [1992] reported the toxicity of fibers must address these factors. shortest fibers to be the least cytotoxic. Brown Durability is difficult to assess using in vitro et al. [1986] reported an association between studies because of their acute time course. length, but not diameter, and cytotoxic activity. However, in vitro studies provide an opportu- Wright et al. [1986] reported that cytotoxicity nity to study the effects of varying doses and was correlated with fibers >8 µm long. Yegles et dimensions of fibers in a quicker, more effi- al. [1995] reported that the longest and thickest cient method than animal testing. They do not fibers were the most cytotoxic. The four most currently provide data that can be extrapolated cytotoxic fibers had GM lengths ≥13 µm and to occupational risk assessment. GM diameters >0.5 µm. The production of abnormal anaphases and telophases was associ- The association between fiber dimension and ated with Stanton fibers with a length >8 µm toxicity has been documented and reviewed and diameter <0.25 µm. Hart et al. [1994] re- [Stanton et al. 1977, 1981; Pott et al. 1987; ported that cytotoxicity increased with increas- Warheit 1994]. RCFs may have different toxici- ing average fiber lengths from 1.4 to 22 µm, ties, depending on the fiber length relative to but did not increase with average lengths from macrophage size. Longer fibers are more toxic. 22 to 31 µm. Fiber length has been correlated with the cytotoxicity of glass fibers [Blake et al.1998]. Man- Additional studies assessing the cytotoxicity ville code 100 (JM–100) fiber samples with of specific RCF fiber lengths are needed. Such average lengths of 3, 4, 7, 17, and 33 µm were studies will help to describe the association be- assessed for their effects on LDH activity and tween fiber length and toxicity for RCFs and rat alveolar macrophage function. The greatest may allow determination of a threshold length cytotoxicity was reported in the 17- and 33-µm above which toxicity increases significantly. samples, indicating that length is an impor- In addition to providing data on the correla- tant factor in the toxicity of this fiber. Multiple tion between fiber length and toxicity, in vitro macrophages were observed attached along studies have provided data on the relative tox- the length of long fibers. Relatively short fibers icity of RCFs compared with other fibers, al- (<20 µm) were usually phagocytized by one though some uncertainties remain in the inter- rat alveolar macrophage [Luoto et al. 1994]. pretation of these studies because of differences Longer fibers were phagocytized by two or in fiber doses, dimensions, and durabilities. more macrophages. Incomplete or frustrated RCFs have direct and indirect effects on cells phagocytosis may play a role in the increased and alter gene function in similar ways. They toxicity of longer fibers. Long fibers (17 µm are capable of inducing enzyme release and cell average length) were a more potent inducer of hemolysis. They may decrease cell viability and tumor necrosis factor (TNF) production and inhibit proliferation. RCFs affect the produc- transcription factor activation than shorter tion of TNF and reactive oxygen species (ROS) fibers (7 µm average length) [Ye et al. 1999]. and affect cell viability and proliferation. They These studies demonstrate the important role induce necrosis in rat pleural mesothelial cells. of length in fiber toxicity and suggest that the They may also induce free radicals, micronu- capacity for macrophage phagocytosis may be clei, polynuclei, chromosomal breakage, and a critical factor in determining fiber toxicity. hyperdiploid cells in vitro. Several of the in vitro RCF studies (summa- In vitro studies provide an excellent opportu- rized in Appendix C) reported a direct associa- nity for investigating the pathogenesis of RCFs. tion between a longer fiber length and greater However, comparisons are difficult to make

Refractory Ceramic Fibers 59 5 n Effects of Exposure

between in vitro studies based on differences [Rossiter et al. 1994; Trethowan et al. 1995; in fiber doses, dimensions, preparations, and Burge et al. 1995]. A followup cross-sectional compositions. Important information such study conducted in 1996 evaluated the same as fiber length distribution is not always de- medical endpoints in workers from six of these termined. Even when comparable fibers are seven European manufacturing plants (one studied, the cell line or conditions under which plant had ceased operation) [Cowie et al. 1999, they are tested may vary. Much of the research 2001]. Current and former workers were in- to date has been done in rodent cell lines and cluded as study subjects in the followup study. in cells that are not related to the primary tar- The studies of U.S. plants began in 1987 and get organ. In vitro studies using human pul- involved evaluations of current workers at five monary cell lines should provide pathogenesis RCF manufacturing plants and former work- data most relevant to human health risk assess- ers at two RCF manufacturing plants [Lemas- ment. ters et al. 1994, 1998; Lockey et al. 1993, 1996, 1998, 2002]. Short-term in vitro studies cannot take into account the influence of fiber dissolution and In the United States, the earliest commercial fiber compositional changes that may occur production of RCFs and RCF products began over time. In an in vivo exposure, fibers are in 1953; in Europe, RCF production began in continually modified physically, chemically, 1968. The demographics of the U.S. and Eu- and structurally by components of the lung en- ropean populations were similar at the time vironment. This complex set of conditions is they were studied, although the average age difficult to recreate in vitro. Just as it is unlikely of U.S. workers was slightly higher than that that only one factor is an accurate predictor of of the workforce in the 1986 European stud- fiber toxicity, it is unlikely that any one in vitro ies because of the earlier development of this test is able to predict fiber toxicity. industry in the United States. The mean age for the European RCF workers was 37.7 in the 1986 study [Trethowan et al. 1995] and 42.0 5.3 Health Effects in Humans for males and 39.4 for female workers in the 1996 study [Cowie et al. 1999]. In the U.S. 5.3.1 Morbidity and Mortality Studies RCF manufacturing industry, the average age is 40 for current workers and 45 for former Two major research efforts evaluated the mor- workers [Lemasters et al. 1994]. The mean du- bidity of RCF-exposed workers—one conduct- ration of employment in the European cohort ed in U.S. plants and one in European plants. was 10.2 years (range 7.2 to 13.8 years) in 1986 Table 5–6 describes the populations analyzed [Trethowan et al. 1995] and 13.0 years in 1996 for both studies. The objective of these research [Cowie et al. 1999]. The U.S. study reports the efforts was to evaluate the relationship between mean duration of employment for 23 workers occupational exposure to RCFs and potential with pleural plaques as 13.6 years (±9.8); the adverse health effects. These studies included median is 11.2 years (range 1.4 to 32.7) [Le- standardized respiratory and occupational his- masters et al. 1994]. tory questionnaires, chest radiographs, and pulmonary function tests (PFTs) of workers, The following text and Table 5–7 summarize as well as air sampling to estimate worker ex- findings from the U.S. and European research posures. The studies of European plants began efforts, organized according to results from in 1986. Study subjects included only current radiographic examinations, respiratory symp- workers at seven RCF manufacturing plants toms, and PFTs. Discussion of two related

60 Refractory Ceramic Fibers 5 n Effects of Exposure

Table 5–6 . Cited studies of populations with occupational exposures to RCFs*

Population analyzed Outcome measures Employment % male % female Study Design status Number workers workers Radiography PFT Symptoms

European: Burge et al. 1995‡ Cross-sectional Current§ 532 100 0 N Y Y Rossiter et al. 1994‡ Cohort morbidity Current** 543 100 0 Y N N Trethowanet al. 1995‡ Cross-sectional Current 628 91 9 Y Y Y Cowie et al. 1999†† Cross-sectional Current 695 90 10 Y Y Y Former 79 85 15

United States:‡‡ Lemasters et al. 1994 Cross-sectional Current 627 83 17 Y N N Lemasters et al. 1994 Cross-sectional Former§§ 220 91 9 Lockey et al. 1993: Cohort mortality Current 684 (including 100 0 N N N and former 46 deceased (Cause of and 5 lost to death) followup)*** Cohort morbidity Current 801 (par- 85 15 Y Y Y and former ticipants; 99% provided respiratory history, 94% provided PFTs, and 90% provided chest X-rays [radiography])

Lockey et al. 1996 Cohort morbidity Current 370 NA NA Y N N Former 282††† NA NA NA NA NA Nested Both (17 cases NA NA Y Y N N case-control with 3 controls each matched on current versus former status)

Lockey et al. 1998 Cross-sectional Current 361‡‡‡ 100 0 N Y N and longitudinal

See footnotes on next page.

Refractory Ceramic Fibers 61 5 n Effects of Exposure

*Abbreviations: N=number; NA=not available from published citation; PFT=pulmonary function test; RCFs=refractory ceramic fibers; Y=yes. †Current versus former (and leaver) worker status at an RCF manufacturing plant as determined at time of survey. ‡Study included current workers at seven ceramic fiber manufacturing plants in three European countries. §From a possible 708 current workers, 628 eligible participants were identified and 596 had chest X-ray examinations; 51 female workers and 13 unexplained others were excluded from analysis. **From a possible 708 current workers, 628 eligible participants were identified and 596 had chest X-ray examinations; 2 unreadable films and those of 51 female workers were excluded from the analysis. ††Study included current workers at six ceramic fiber manufacturing plants in three European countries as well as leavers from the first three European studies [Burge et al. 1995; Rossiter et al. 1994; Trethowan et al. 1995] (one of the seven plants included earlier had ceased operation). ‡‡Studies included current and former workers at five RCF manufacturing plants in the United States. §§From a possible 1,030 eligible current and former workers, 183 were either deceased, not located, or did not agree to chest X- ray examinations. ***From a possible 729 eligible current and former workers at 2 plant sites for whom individual work histories were available, 45 were excluded on the basis of insufficient exposures to fibers or insufficient data regarding fiber exposures. †††From a possible 868 eligible current and former workers at 2 plant sites, 148 were eliminated for lack of exposure characterization data and loss to followup. Of the remaining 720 workers, 68 did not agree to chest X-ray examinations. ‡‡‡From a possible 963 eligible current workers at five plant sites, 209 female workers were excluded as well as 393 male workers with fewer than 5 PFT sessions.

mortality studies is also presented in Sec- 5.3.2.1 Pleural abnormalities tion 5.3.5 [Lockey et al. 1993; Lemasters et In the 1986 study of European RCF workers, re- al. 2003]. Two HHEs of workplaces involving sults of the chest radiography indicated a preva- workers exposed to RCFs are also described in lence of 2.8% (15/543) for pleural abnormali- Section 5.3.6 [Kominsky 1978; Lyman 1992]. ties among male workers [Rossiter et al. 1994]. Of the 15 cases with pleural abnormalities, 4 5.3.2 Radiographic Analyses had bilateral diffuse thickening (1 with calcification), 1 showed bilateral pleural calcification In both the European and U.S. studies cited only, 7 presented with unilateral diffuse thick- in Table 5–6, the study populations includ- ening, and 3 showed costophrenic angle blunt- ed workers at multiple plants involved in the manufacture of RCFs or RCF products. As part ing only. The possibility for confounding effects of the investigation of potential effects of expo- was recognized because of other exposures: sure to airborne RCFs, chest radiography was 52% of workers reported previous employ- performed. In all studies, chest radiographs ment in dusty jobs, including 4.5% with prior were read independently by three readers us- asbestos exposures and 7% with prior MMMF ing the International Labour Office (ILO) 1980 exposures. When female workers were included International Classification of the Radiographs in the same population, Trethowan et al. [1995] of Pneumoconioses [ILO 1980]. Identifiers on reported a prevalence of 2.7% (16/592) for films were masked to ensure a blind review by pleural abnormalities. Two cases were known to readers, and quality control measures and tests have previous exposure to asbestos, and the pos- of agreement were used to check consistency sibility for exposure to other respiratory hazards among the readers. For each type of abnormal- was acknowledged for other persons with pleural ity analyzed, the median of the three readings abnormalities. Cowie et al. [1999, 2001] reported for each film was used. pleural abnormalities in 10% (78/774) and

62 Refractory Ceramic Fibers 5 n Effects of Exposure - - - - regres

RCF (Continued) logistic Comments a statis found sion exposure. known for ment exposure asbestos and a statistically significant associa tically significant association between and pleural plaques RCF first time since (>20 job production adjust after years) pleuraltion between and durationplaques of (>20 years) vital FVC=forced Multiple as

9.7 29.1 7.0 30.1 — — workers

95% CI 0.8, 2.0, 0.9, 2.6, observed.

classified

were

‡,§ ‡,§ **,†† **,†† production *

were of

1.0 2.9 7.7 1.0 2.5 8.8

OR nonproduction

opacities 3.4% of

in changes

Results 1.3 3.6 1.9 4.3 in (%)

seen 11.4 20.7

in 1 second; expiratory volume =forced irregular 1

pleural

and Prevelance

Pleural changes † No were of

FEV

91.3% (23/686)

changes plaques.

RCF latency:

the FVC;

Pleural workers (0/161); pleural >0–10 >10–20 >20 >0–10 >10–20 >20 of Years Item of Years Radiographic analyses: changes Pleural exposure: - - - used.

using

work

function

Standard examina

were

847

) respiratory for

X-ray obtained 25–75

and European morbidity and European studies with RCFs and

pulmonary FEF

. S were chest

U questionnaire.

and

performed

Evaluation methods Evaluation FVC, histories

, were 1

measures

Occupational

(FEV Posteroanterior standardized health ers. tions ized Table 5 –7 . Table 25% and 75% of expiratorybetween flow =forced

25–75 exposure. asbestos

FEF

Female and

Male latency. exposure. RCF at least 1 year at least 1 year fibers. ceramic RCFs=refractory in per week division

production RCF Study designStudy hr and population

in employment in employment 4 of and years latency and 1,030 current at five workers former U.S. facilities employed 10/87–8/89. workers: of fiber division. workers: of fiber ≥ production. Cross-sectional study: Cross-sectional interval; CI=confidence OR=odds ratio; exposure. RCF first time since 1994 study) Reference Lemasters et al. (U.S. Abbreviations: with of >0–10 years Compared of years for Adjusted Latency: variables not described). (confounding Adjusted with of >0–10 years Compared * † ‡ § capacity; ** ††

Refractory Ceramic Fibers 63 5 n Effects of Exposure in

for

- than

(Continued) workers. (except

respira

approximately

higher workers

was

Comments 5-fold expiratory = forced

symptoms 1

nonproduction production asthma) 2- tory Prevalence FEV ‡ ‡ * the FVC; 8.5 4.7 3.0 4.6 38.8 46.9 33.5 6.2 3.2 5.3 30.5 — =0.03 =0.31 =0.001 =0.21 =0.21 95% CI P P P P P

1.4, 1.1, 1.7, † ‡ ‡ ‡ OR —

7.3

—

— — — — tion workers Produc- n=86 — Results n=80 n=517

Prevalence (%) Prevalence duction workers Nonpro- Symptom analysis Symptom

= refractory fibers. ceramic RCFs 1 Dyspnea 2.5 15.7 Wheezing Asthma cough Chronic 3.8 phlegm Chronic 5.0 3.8 Pleuritic pain 10.3 2.5 7.4 0.0 2.5 5.8 1.0 0.8, 2.3 1.2 0.3, 0.4, 1.6 1.0 0.2, Wheezing Asthma 1.7 0.0 7.0 3.9 3.5 0.4, more One or cough Chronic phlegm Chronic 1.7 1.7 Pleuritic pain more One or 9.3 0.0 9.3 5.4 3.8 0.6, 0.4, 3.5 1 Dyspnea 2 Dyspnea 18.6 0.0 25.6 1.3 10.5 0.6, 2 Dyspnea 0.0 4.8 symptoms 11.3 29.6 2.9 symptoms 20.3 40.7 2.4 25% and 75% of expiratorybetween flow = Forced workers Male n=59 workers Female

Symptom

= probability; P

- 25%−75% - and European morbidity and European studies with RCFs FEF 4 hr/ . S ≥ and : U : Measures FVC, , time) in work 1 in a logistic model. regression exposure possible asbestos OR = odds ratio; . American Thoracic 25%–75% Evaluation methods Evaluation interview. ATS of and years

n = number; FEF view Defined as spending Spirometric evaluation (pul questionnaire. Society (ATS) Modified historyOccupational inter (10% of week areas. production monary function) at time of FEV included interval; CI = confidence Medical Evaluation Medical assessment Exposure work: Production

. 5–7 (Continued) Table -

occupational 753 active in <0.05). Study designStudy P vital capacity; = forced FVC and population American Thoracic Society; at five 145 female) manufacturing RCF who particisites 742 of (597 male; workers 1989. 1987 and between pated history interviews Cross-sectional study: Cross-sectional smoking (category years), and pack = ATS

1998 study) Reference Lemasters et al. (U.S. Statistically significant ( age, for Adjusted in 1 second; volume *Abbreviations: † ‡

64 Refractory Ceramic Fibers 5 n Effects of Exposure

Comments (Continued) -

- in male

1 RCF Results and (3) FVC of Years production employment signifi were smokers. to related cantly decline volume for in (1) FVC and male current past smokers, (2) FEV ers, never female for smok current fibers. ceramic RCFs=refractory * 8.7) 40.7) 92.9) n=68 n=173 (–692.0, ‡ Never smokers Never O=odds ratio; and pack-years for interaction term n=number; –9.1) –40.6 (–175.1,193.8) 211.1) –350.3 97.4) –223.5 (–487.7, 50.9) –20.4 (–133.7, n=20 (–301.9, n=174 ‡ smokers Past Δ volume in (ml) Δ volume vital capacity; FVC=forced –51.0) –155.5 –38.5) –72.5 (–195.9, 190.0) –330.5 (–872.1, 246.5) –321.4 (–740.2, and smoking category, production RCF Evaluation methods Evaluation Pulmunary function n=56 and European morbidity and European studies with RCFs n=245 (–279.8, (–231.3, ‡ ‡ . S U 13.9 (–218.7, smokers Current –110.8 (–411.5, –134.9 –165.4

RCF RCF RCF RCF duration of for interaction term Item of ] years of ] years in 1 second; expiratory volume =forced 1 1 1 [FEV [FEV employment employment employment of years [FVC] employment of years [FVC] FEV . 5–7 (Continued) Table workers Male workers Female

smoking category,

of and plant location. † 0.05). ≤

P height- weight, Study designStudy Multiple Multiple regression analysis in change (ml) volume of adjusted spirometric measures (95% CI) associated with RCF employment duration. Controlled smoking for status. and population interval; CI=confidence

1998, study) Reference continued (U.S. Lemasters et al. Statistically significant ( in model: included Variables Abbreviations: smoking category, * † ‡

Refractory Ceramic Fibers 65 5 n Effects of Exposure

pleural one 18 produc One was and cumula . duration of 3 20 cases were Comments (Continued) <0.001) with P workers. models, regression pleural plaques associated were ( first since years production RCF job, production RCF jobs, cm and tion workers 2 nonproduction fiber-months/ tive reinterviewed; in included were the case-control study. deceased, interview, refused withand one exposure asbestos was excluded. 20 cases of were plaque in the identified cohort: logistic three In 18 of the cohort study, In 6.1 29.7 137.0 125.4 176.2 224.9 1.5 10.3 11.2 48.2 5.8 9.9 95% CI *

† OR

— 2.8 1.3, 0.9 1.4 1.0 1.1 0.2, — 0.3 5.3 1.0 6.4 15.4 7.8 21.3 1.9, 24.2 — 2.6, 2.6, 0.9 1.7 1.0 1.4 0.2, — 2.8 2.2 0.4, 7.4 6.1 1.2, (%) 12.5 9.5 1.9, 26.3 22.3 3.6, Plaque Plaque Results as indicated. (log):

: 3 3 RCF

years asbestos

Radiographic analyses job — 1.2 1.0, asbestos

first for for RCF

index — 3.8 1.5,

first

RCF since

‡ 0 Adjusted Adjusted fiber-months/cm since exposure rating production prevalence employment: 0 >0–10 >10–20 >20 Cumulative fiber-months/cm >0–15 >15–45 >45-135 >135 Case-controls: Years Cumulative >0–10 >10–20 >20 of Years

RCF first since Years job: production Item

1996]. using

of regard- and inter- expo-

and European morbidity and European studies with RCFs X-rays. distance index design

job

(applica- . S 1994, interview,

from and

worker reinterviewed work/home al. chest

RFCs=refractory fibers. ceramic

each

controls, (rating et Asbestos

questions

process, Last entry rating asbestos index, for is adjusted

low)

for

plus

history

cases

about exposure

[Rice concentrations

plant data and

exposure;

Evaluation methods Evaluation medium, exposure). exposure manipulation, data. additional

categorized

engineering

asbestos OR=odds ratio;

from with Health: Exposure: Estimated Exposure: Controls Occupational sure interviews historical fiber view asbestos tion, sampling and high, information

ing

or =probability;

P . 5–7 (Continued) Table

two

and to RCF

10/87

of em-

posterior- X-rays. em- history sex

status two who

the

former pleural

one

and or completed at of were

production chest

year

matched between

1 sites,

Study designStudy 12/91 provided ≥ workers

who

who and population

cases

controls

study: anterior and Cohort study: Cohort 652 (a) (b) case-control Nested 17 plaque nonproduction) employment ployed and 3 ployment (current plant oblique had division occupational worker; and interval; CI=confidence

1996 study) Reference Lockey et al. (U.S. Abbreviations: workers. Nonproduction exposure. asbestos first since last entry years All but for adjusted are * † ‡

66 Refractory Ceramic Fibers 5 n Effects of Exposure

- - - pre On (Continued) nonpar and had lower Comments values. function lung dicted from excluded were because the analysis they did not partici older, ticipants were and weighed smoked more, height-adjusted and of percentages in at least five pate PFT sessions. average, 193 male workers

†,‡ 1 * FEV smoking category,

†,‡ fibers. ceramic RCFs=refractory Regression coefficient coefficient Regression Results FVC

test]), [first >7 years –219.4 (<0.01) –205.2 (<0.01) –65.6 (0.44) –80.6 (0.25) § § 7 ≤ 7 years, ≤

Pulmonary analysis cross-sectional of function: initial pulmonary 522 workers for function test production RCF

years PFT=pulmonary function test; - - - ex (fiber- for

and European morbidity and European studies with RCFs

plants

from

cumula

estimated only.

. S of

: two

U : fiber exposure

exposure

plants

from

from ) estimated RCF RCF 3

job-group-specific

manufactur RCF five

these calculation

. data concentrations of

(0, years production RCF categorical at vital capacity; FVC=forced

and PFT (1987) through first

Evaluation methods Evaluation date fiber

months/cm and FEV FVC included Measures spirometric Yearly ing facilities and assigned each to history. job based on worker Cumulative Period 1987 and 1994. evaluation (pulmonary function) between concentrations; posure of each for tive first workers of date final PFT date. Pre-1987 concentrations permitted Medical evaluation Medical assessment Exposure

. 5–7 (Continued) Table

the

a pos- and (3) RCF Study designStudy and population a possible 754 of at one 1 month in 1 second; expiratory volume =forced and plant location. 1 ≥ Cross-sectional and Cross-sectional longitudinal study: Of male workers, manufacturing, hired who (1) were employed (2) were 361 included study 6/30/90, before and 6/94. U.S. five PFT ses- seven) sible 6/87 between sions (of least five participated in at facilities, age, were in the analysis variables included FEV

weight,

1998 study) Reference Lockey et al. (U.S. Abbreviations: Height-adjusted; with workers. nonproduction Compared milliliters. In years, pack * † ‡ §

Refractory Ceramic Fibers 67 5 n Effects of Exposure (Continued) Comments

†,‡ 1 pack +0.2 +0.2 +0.5 –37.6 –38.9 –99.6 FEV at initial test, years pack

–100.3 –100.2 –104.7 at years pack * ** ** ** †,‡

0.7 +5.3 +0.8 –36.2 –66.3 –55.2 –171.0 –156.0 –168.3 FVC . . . . § § , , ‡ ‡ . . . . .

†† ...... ,

§ § , 7 years ‡ exposure, RCF cumulative followup Results ) ≤ 3 ...... >7 years

‡‡ , years, production RCF followup 3 exposure, RCF cumulative followup ...... ‡‡ , 3 ...... ‡‡ fibers. ceramic RCFs=refractory ,

3 Item plant location. plant location...... Pulmonary longitudinal function: analysis § § , , ‡ ‡ =probability; P

7 years fiber-months/cm weight change, ≤ exposure, RCF cumulative Followup >15–60 fiber-months/cm >60 fiber-months/cm weight change, and European morbidity and European studies with RCFs

exposure, RCF cumulative Initial exposure, RCF cumulative Initial RCF cumulative Followup (fiber-months/cm exposure years, production RCF Initial years production RCF Followup category, production RCF Initial years, production RCF Initial >7 years

Cumulative fiber exposure: Cumulative category, production RCF Initial employment duration: Production Regression coefficient Regression exposure: Cumulative . S U initial test, to exposure RCF cumulative categorical initial test, to years production RCF categorical weight at initial test, initial test, to years production RCF categorical vital capacity; FVC=forced and plant location. weight change weight at initial test, test, each Evaluation methods Evaluation

. 5–7 (Continued) Table weight at initial test, Study designStudy and population test, smoking at each dichotomized in 1 second; expiratory volume =forced 1 age, were in the analysis variables included age, were in the analysis variables included age, were in the analysis variables included 1998 FEV smoking at the time of dichotomized study) Reference Reference

<0.05. (Continued) P et al. Lockey (U.S. Height-adjusted; Height-adjusted;

Height-adjusted; with workers. nonproduction Compared times. seven to five tested male workers for milliliters In Abbreviations: test, smoking at each dichotomized at initial test, years * † ‡ § ** †† ‡‡ initial test,

68 Refractory Ceramic Fibers 5 n Effects of Exposure

- of - - - effects suggest cumula and eye Decre (Continued) and breathless respiratory Comments nose, stuffy of analysis No smoking. fibers to ing that exposure promotes was performed exposures tive with symptoms. respect to to strongly related more were fibers to exposure cumulative exposure cumulative than to inspirable mass. to function in lung were ments smokers current to limited smokers, and former ness. inspirable and respirable dust dry to related were fibers cough, skin irritation, function in lung Changes both to exposures Current ‡ *

15.26)] 3.57)] 8.6)] 2.23)] 8.33)] 5.03)] 3.84)] 5.42)] 5.11)] 3.39)] 3.54)] 2.11)] 4.08)] 6.05)] 2 ≥ Effect

nose Stuffy [2.01 (1.13, irritation Eye [4.78 (2.66, Skin irritation [3.3 (1.8, [5.84 (2.25, [2.25 (1.43, Dyspnea [3.42 (1.41, Skin irritation [3.18 (2.01, [2.01 (1.05, 2 Dyspnea [2.66 (1.31, irritation Eye [2.63 (1.7, nose Stuffy [2.06 (1.25, irritation Eye [2.16 (1.32, Skin irritation [1.25 (0.74, [2.53 (1.25,

† † irritation Eye 3 Dry cough Results 3 Baseline 3 Symptom analyses Symptom Baseline Dry cough Dry cough 3 3 3 0.6 f/cm 3 mg/m

<3 mg/m 2.5 to <2.5 mg/m <0.2 f/cm ≥ <0.6 f/cm 0.2 to

Current concentrations Current of inspirable mass: concentrations Current of respirable fibers:

Exposure - - -

self- -

skin of dyp

and lung

logistic

and and

samples the

method respira WHO/ and

included

and

cough, function CI)

exposure plants.

smok

each

data for

represen

air for

linear

specific in randomly eye mass fiber

wheeze,

dry

cumulative

that

sex,

95%

calculated

from to multiple

separately

performed.

and European morbidity and European studies with RCFs obtain concentration.

current

total from (i.e.,

reference nose,

evaluated collected

fibers. ceramic RCFs=refractory

expanded

to personal were plant, . S

(with

Multiple

Pulmonary also production

coefficients

and

controlled using mass and fiber

regarding

category

Comments digestive urinary SMR=2,614 SMR=913

of workers

increase (1) years male other peritoneum

specific

the power

the the deaths deaths

in 9,471]). ing: and sian for >30 (n=2, CI=246–7,490]); of and Caucasian (n=2, CI=110–3,295]); of male years SMR=3,306 any cant because of Statistically The

) )

518 277 126 3,390 3,155 2,512 1,197 11,175 97 (69–132) 114 (31–291) 467 (57–1,687) 259 (94–563) 121 (60–216) Person-years at risk

205 (25–740) † * Results

† death SMR (95% CI) the urinary the digestive

† † job RCF 10 >5 to 15 >10 to 20 >15 to 25 >20 to 30 >25 to 5 1 to >30 Total first since Years All cancers: cancers 4 lung 6 non-Caucasions 126 (46–274 Demographics: distribution Worker organs organs of Cause All causes: 11 Caucasions 2 non-Caucasions 263 (32–950 and other 2 pneumoconioses of 2 cancers of 6 cancers 40 Caucasions 60 non-Caucasions 624 Caucasions respiratory disease

male Mortality with study RCFs and cumula- mortality SMR=standardized ratio. time, calendar Person-years Evaluation methods Evaluation Table 5–8 . Table tion. Cause-specific SMRs U.S. the total stratified age, by were using calculated were popula- reference race, duration. tive as the population latency,

Of 46

633

between RCFs fibers; ceramic RCFs=refractory Study designStudy and population deceased, (6.7%) were of facture 10/1/50 and 6/1/88. who the 684 workers met the criteria, alive, (92.5%) were lost and 5 (0.7%) were followup. to and former Current at two male workers at employed plant sites in the man least l year Cohort mortalityCohort study:

interval; CI=confidence 1993 Reference study) Abbreviations: cohort. race Combined et al. Lockey (U.S. * †

Refractory Ceramic Fibers 79 5 n Effects of Exposure

the combined-race cohort to have no signifi- warrant continued surveillance of the cohort cant elevations associated with specific causes mortality registry. of death, including cancers of the lung, digestive organs and peritoneum, urinary organs, and Walker et al. [2002] used the same cohort of pneumoconioses and other respiratory disease. male RCF production workers described by Le- The authors noted that the power to detect a masters et al. [2003]. Walker et al. performed a significant increase in mortality for any specific risk analysis comparing the lung cancer and me- cause was low because of the small number of sothelioma in the cohort’s accumulated mortal- deaths in the cohort and generally short laten- ity experience to that which would have been expected if RCFs had a carcinogenic potency cies. However, a statistically significant increase approximating various forms of asbestos. The in deaths from pneumoconioses and other re- authors reported that deaths from lung cancer spiratory disease occurred in Caucasian males in the RCF cohort were statistically significantly with >30 years RCF latency (n=2, SMR=2,614 below that which would be expected if RCFs [95% CI=246–7,490]). A statistically significant had the potency of either crocidolite or amosite. elevation in deaths from cancers of the diges- The mortality was also lower than would be ex- tive organs and peritoneum also occurred for pected if RCFs had the potency of chrysotile, non-Caucasian males (n=2, SMR=913 [95% but the difference is not statistically significant. CI=110–3,295]). In addition, a statistically sig- For mesothelioma, the authors concluded the nificant elevation occurred in the number of anticipated numbers of deaths under hypoth- deaths from cancers of the urinary organs for eses of asbestos-like potency are too small to be male workers with >15 to 20 years of RCF la- rejected by the zero cases seen in the RCF co- tency (n=2, SMR=3,306 [95% CI=311–9,471]). horts [Walker et al. 2002]. NIOSH researchers noted that this analysis by Walker et al. was not Lemasters et al. [2003] published a subsequent based on the most current update of the RCF analysis of current and former male workers cohort. In addition, the asbestos risk assessment employed between 1952 and 2000 at the two models used by Walker et al. [2002] were fitted RCF manufacturing facilities (942 subjects) to studies with longer followup periods than investigating a possible excess in mortality. the cohort of RCF workers. Because these mod- The mortality analytic methods included (1) els do not specify length of followup, it is not standardized mortality ratios comparing this possible to adjust for these differences. Conse- cohort with the general and State populations quently, it is likely that the RCF cohort has not and (2) a proportional hazards model that re- been followed for a sufficient length of time to lates risk of death to the lifetime cumulative 3 demonstrate the risks that were observed in the fiber-months/cm exposure among the RCF asbestos cohorts. NIOSH believes the mortal- cohort, adjusted for age at hire and for race. ity study by Lemasters et al. [2003] and the risk The analysis found no excess mortality relat- analysis by Walker et al. [2002] have insufficient ed to all deaths, all cancers, or malignancies power for detecting lung cancer risk based on or diseases of the respiratory system (includ- what would be predicted for asbestos. ing mesothelioma) but found a statistically significant association with cancers of the urinary organs [SMR=344.8 (95% confidence 5.3.6 NIOSH HHEs limits of 111.6, 805.4)]. Based on the small As part of its mission as a public health agen- size of the cohort, the young average age cy, NIOSH performs HHEs at the request of (51 years), and a mean latency of 21 years, workers, employers, or labor organizations to the researchers concluded that the findings investigate occupational hazards associated

80 Refractory Ceramic Fibers 5 n Effects of Exposure

with a workplace or work-related activity. One to have been exposed to RCFs (n=19, including such HHE involved evaluating worker expo- 9 with pleural abnormalities, 13 with fibrosis, sures to ceramic fibers at a company manufac- and 1 with mesothelioma). However, no at- turing steel forgings [Kominsky 1978]. At the tempt was made to analyze further for an as- facility, furnaces for heat-treating steel ingots sociation of the cases with exposure to RCFs. were lined with RCF felt and batting, and this The report implied that occupational exposure lining required regular maintenance and re- to kaolin dust and to asbestos caused many or placement. Among the workers interviewed all of the job-related conditions. were six bricklayers involved in furnace lining maintenance. Four of the bricklayers reported having experienced irritation of exposed skin 5.3.7 Discussion areas and of the throat during the handling The radiographic analyses of the U.S. and 1996 and installation of the RCF-containing insula- European worker groups suggest an association. On the basis of the reported symptoms tion between pleural abnormalities, including and their consistency with known effects of pleural plaques, and RCF exposure [Lemas- RCFs, the symptoms of irritation were attrib- ters et al. 1994; Lockey et al. 1996; Cowie et uted to RCF exposure. No attempt was made to al. 1999]. From Rossiter et al. [1994] it is less measure airborne fiber concentrations. Anoth- apparent whether such an association was in- er NIOSH HHE [Lyman 1992] resulted from vestigated. Trethowan et al. [1995] report that an OSHA inspection that identified 18 cases of pleural abnormalities were not independently occupational lung disease recorded in 1 year at related to RCF exposure. Differences between a plant manufacturing fire bricks, ceramic fiber the findings of the U.S. studies and those of products, and other thermal insulation com- the initial European studies may be related to ponents from kaolin. About 600 workers were the long latency before pleural abnormalities potentially exposed to respiratory hazards that are detectable, in particular, pleural plaques included not only RCFs but also kaolin dust, following RCF exposure. Workers exposed to crystalline silica dust, and (for maintenance asbestos developed asbestos-associated pleu- workers) asbestos. A total of 38 workers had ral plaques after a latency period of more than been referred to a pulmonary physician for 15 years after initial exposure [Hillerdal 1994] evaluation based on 2 rounds of chest X-ray and in some cases, after 30 to 57 years [Begin screening of the workforce in 1980 and 1986. et al. 1996]. The European RCF industry de- Diagnoses were related to pleural thickening veloped more than a decade after the U.S. in- (n=10), pleural plaques (n=3), diffuse pulmo- dustry. As a result, workers in the U.S. group nary fibrosis (n=21), mesothelioma (n=1), and are slightly older with a longer average em- other miscellaneous conditions. At least 20 of ployment duration in RCF manufacturing and these cases were classified as work-related by time since first exposure to RCFs. Historical the pulmonologist who evaluated the cases. air sampling data also indicate that airborne The nonoccupational classification of some fiber concentrations were much higher in of the remaining 18 cases was questioned by a early U.S. RCF manufacturing. These factors NIOSH physician who performed a retrospec- might explain the finding of RCF-associated tive record review. The 38 cases were reclas- pleural abnormalities in the U.S. workers but sified on the basis of job histories into those not in the European workers. A further pos- who were likely to have been exposed to RCFs sible explanation may involve differences in (n=19, including 4 with pleural abnormalities the radiographic surveillance methodologies. and 8 with diffuse fibrosis) and those unlikely Both the U.S. and the European studies used

Refractory Ceramic Fibers 81 5 n Effects of Exposure

the 1980 ILO classification systems for pneu- adjusting for the effects of age, sex, and smok- monoconioses to review posteroanterior view ing. Exposure to RCF concentrations in the chest radiographs for study subjects. However, range of 0.2 to 0.6 f/cm3 was associated with Lockey et al. [2002] began to supplement these statistically significant increases in eye irrita- views with left and right 45o oblique view films tion (OR=2.16, 95% CI=1.32–3.54), stuffy nose as a standard practice for radiographic surveil- (OR=2.06, 95% CI=1.25–3.39), and dry cough lance. This methodology, known as a film tri- (OR=2.53, 95% CI=1.25–5.11) compared with ad, was evaluated against the posteroanterior- exposure concentrations lower than 0.2 f/cm3 only view to determine reliability, sensitivity, [Trethowan et al. 1995]. Increasing ORs were and specificity of each method [Lawson et al. demonstrated for RCF exposure concentra- 2001]. The evaluation, involving 652 subjects tions greater than 0.6 f/cm3 compared with ex- in the RCF study, showed the film triad had posure concentrations <0.2 f/cm3 for wheeze considerably higher interreader reliabil- (P<0.0001), dyspnea (P<0.05), eye irritation ity (κ=0.59) than the posteroanterior-only (P<0.0001), skin irritation (P<0.0001), and dry method (κ=0.44). The authors concluded that cough (P<0.05) but not stuffy nose or chronic the film triad method provides an optimum bronchitis [Trethowan et al. 1995]. Lockey et approach. al. [1993] found that dyspnea was significantly The U.S. and 1986 European studies yielded associated with exposure to >15 fiber-months/ 3 3 little evidence of an association between radio- cm (that is, >1.25 fiber-years/cm ) relative to 3 graphic parenchymal opacities and RCF expo- exposure to ≤15 fiber months/cm (dyspnea sure. In the U.S. study, small opacities were rare grade 1COR=2.1, 95% CI=1.3–3.3; dyspnea [Lockey et al. 1996]. Small opacities of profu- grade 2—OR=3.8, 95% CI=1.6–9.4). Lockey et sion category 1/0 or greater were more frequent al. [1993] also found statistically significant as- in the 1986 European study [Trethowan et al. sociations between cumulative RCF exposure 1995], but exposures to silica and other dusts and chronic cough (OR=2.0, 95% CI=1.0–4.0) were believed to account for many of these and pleurisy (OR=5.4, 95% CI=1.4–20.2). Le- cases. The results of statistical analyses did masters et al. [1998] also noted associations not implicate RCF exposure [Trethowan et al. (P<0.05) between employment in an RCF 1995] or yielded results only slightly suggestive production job and increased prevalence of of an RCF exposure effect [Rossiter et al. 1994]. dyspnea and the presence of at least one re- In the 1996 evaluation of the European cohort, spiratory symptom or condition. Recurrent small opacities of category 1/0 or greater were chest illness in the European cohort was asso- positively associated with RCF exposures that ciated with cumulative exposure to respirable occurred before 1971 [Cowie et al. 1999]. Ten fibers and was most strongly associated with of the 51 (19.6%) male workers exposed before cumulative exposure to respirable dust [Cowie 1971 developed category 1/0 or greater opaci- et al. 1999]. tiesC8 had also been exposed to asbestos and 9 In cross-sectional analyses involving spiro- were either current or ex-smokers. metric testing, both the U.S. [Lockey et al. Both the U.S. [Lockey et al. 1993; Lemasters 1998; Lemasters et al. 1998] and 1986 Europe- et al. 1998] and the European [Trethowan et an [Trethowan et al. 1995; Burge et al. 1995] al. 1995; Burge et al. 1995; Cowie et al. 1999] studies found that cumulative RCF exposure studies found that occupational exposure was associated with pulmonary function dec- to RCFs is associated with various reported rements among current and former smok- respiratory symptoms and conditions, after ers. The 1996 European study demonstrated

82 Refractory Ceramic Fibers 5 n Effects of Exposure

decrements in current smokers only [Cowie et occupational exposure limits for specific types al. 1999]. The observed decreased pulmonary of SVFs. As one of the criteria for determin- function in the European workers remained ing the airborne exposure limits for six distinct significantly associated with cumulative RCF types of SVFs, risk assessments were performed exposure, even after controlling for cumula- for each fiber type, including RCFs. The risk tive exposure to inspirable dust [Burge et al. analysis for RCFs was based on the assumption 1995]. A longitudinal analysis of data from that RCFs are a potential human carcinogen as multiple PFTs by Lockey et al. [1998] led the indicated by the positive results of carcinoge- researchers to conclude that exposures to RCFs nicity testing with animals. A health-based rec- between 1987 and 1994 were not associated ommended occupational exposure limit was with decreased pulmonary function. The find- determined using the following rationale: ings from the U.S. and European studies suggest that decrements in pulmonary function 1. If the carcinogenic potential of RCFs is observed in current and former smokers result caused by a nongenotoxic mechanism, from an interactive effect between smoking an occupational exposure limit of 1 re- and RCF exposure. spirable f/cm3 as an 8-hr TWA should be recommended based on an NOAEL of 25 f/cm3 and a safety factor of 25. 5.4 Carcinogenicity Risk 2. If the carcinogenic potential of RCFs is Assessment Analyses linked to a genotoxic mechanism, a mod- The literature contains three significant inde- el assuming a linear relationship between pendent risk analyses of occupational expo- dose and the response (cancer) should be sure to RCFs and potential health effects. In used to establish the occupational expo- each of these analyses, health effects data de- sure limit. rived from multidose and MTD studies with The model indicated that an excess cancer risk rats were used with models to extrapolate risks of 4 ×10-3 is associated with a TWA exposure to human populations. The modeling of ef- to 5.6 respirable f/cm3 based on 40 years of oc- fects observed in experimental animal studies cupational exposure. A cancer risk of 4×10-5 is was necessitated by the lack of adequate data associated with exposure to 0.056 f/cm3, and on adverse health effects in humans with oc- a linear extrapolation indicated that occupa- cupational exposures to RCFs. The three stud- tional exposure to 1 respirable f/cm3 as an 8-hr ies, described in detail below and in Table 5–9, TWA for 40 years is associated with a cancer include the following studies: Dutch Expert risk of 7×10-4. Committee on Occupational Standards (DE- COS) [1995], Fayerweather et al. [1997], and The DECOS analysis relied on the data from a Moolgavkar et al. [1999]. long-term multidose study with rats exposed to kaolin ceramic fibers [Bunn et al. 1993; Mast et al. 1995b]. These data showed that expo- 5.4.1 DECOS [1995] sure by inhalation to 25 f/cm3 (3 mg/m3) for In 1995, DECOS (a workgroup of the Health 24 months produced a negligible amount of Council of the Netherlands) published a report fibrosis (mean Wagner score of 3.2). Conse- evaluating the health effects of occupational quently, the Dutch committee viewed 25 f/cm3 exposure to SVFs. The purpose of the report as the NOAEL for fibrosis. The report also was to establish health-based recommended notes that at the time of publication, no data

Refractory Ceramic Fibers 83 5 n Effects of Exposure

) ) ) ) ) ) -5 -4 -5 -4 -3 -3

TWA=time-weighted

(95% UCL = 0.9×10 (95% UCL = 1.8×10 (95% UCL = 1.2×10 (95% UCL = 1.8×10 (95% UCL = 1.5×10 (95% UCL = 5.8×10 -5 -4 -6 -4 -4 -3

-5 -4 Excess lifetime risk ofExcess cancer–MLE lung industrySteel cohort, industrySteel cohort, industrySteel cohort, 7×10 3.8×10 Exponential: 3.7×10 1.5×10 Quadratic: 4.1×10 1.5×10 Linear: 2.7×10 1.1×10 cohort, nonsmoking ACS cohort, nonsmoking ACS cohort, nonsmoking ACS

fibers; ceramic RCFs=refractory

a 70-year of 40 years 30 years scenario 5 days/week, 5 days/week, 40 years Occupational exposure exposure Occupational lifespan 50 weeks/year, 52 weeks/year, 20–50) of (age lifespan 70-year 4 hr/day, 8 hr/day, 8 hr/day, (TWA) as determined by three independent analyses independent as determined three by (TWA) 3

dose; tolerated MTD=maximum

at 1 f/cm * nonthreshold Extrapolation model model model expansion Linearized model multistage Linearized, clonal Two-mutation

estimate; likelihood MLE=maximum Animal data Animal Risk associated with Risk RCFs to exposure and multidose Long-term and multidose Long-term and multidose Long-term 1995a,b) 1995a,b) MTD rat studies et al. rat (Mast studies et al. rat (Mast studies dose tolerated maximum dose tolerated maximum Society; Cancer ACS=American

Table 5–9 . Table limit. confidence UCL= 95% upper

1997 1999 Study Abbreviations: 1995 Fayerweather Moolgavkar * et al. average; et al. DECOS

84 Refractory Ceramic Fibers 5 n Effects of Exposure

existed from retrospective cohort mortality or case/100,000 exposed persons) were 2 and 3 morbidity and case-control studies of persons orders of magnitude higher. Conversely, the with occupational exposures to RCFs. The lin- risk estimates for exposure to 1 f/cm3 for a ear modeling approach in this analysis of the working lifetime were lower using the Weibull exposure-response relationship using the ani- 1.5-hit nonthreshold and Weibull 2‑hit thresh- mal data does not take into consideration pos- old models. sible differences in dosimetry and lung burden between rats and humans. 5.4.3 Moolgavkar et al. [1999]

5.4.2 Fayerweather et al. [1997] This report describes a quantitative assessment of the risk of lung cancer associated with oc- Fayerweather et al. [1997] conducted a study cupational exposure to RCFs [Moolgavkar et primarily focusing on the risk assessment of al. 1999]. A major premise underlying the risk occupational exposures for glass fiber insula- assessment is that humans are equally suscep- tion installers. They performed risk analyses tible to RCFs as rats, at the tissue level. The risk with several other types of SVFs, including analysis was performed using data from two RCFs. Only the analysis with RCFs is present- chronic inhalation bioassays of RCFs in male ed here. This analysis applied an EPA linear- Fischer 344 rats [Mast et al. 1995a,b]. Dosim- ized multistage model (representing a linear etry in the risk assessment was based on a fiber nonthreshold dose-response) to data from rat deposition and clearance model developed by multidose and MTD chronic inhalation bioas- Yu et al. [1996] that was used to estimate the says [Mast et al. 1995a,b] to determine expo- lung burdens of fibers in humans. The dose- sures at which Ano significant risk@ occurs; i.e., response model used for the risk assessment no more than one additional cancer case per was the two-mutation clonal expansion model, 100,000 exposed persons. Nonlinear models commonly referred to as the Moolgavkar-Ven- were also used for comparison: the Weibull zon-Knudson (MVK) model. The MVK model 1.5-hit nonthreshold model (representing the was fitted to the rat bioassay data to estimate nonlinear, nonthreshold dose-response curve) the proportional increase in the rat lung tumor and Weibull 2-hit threshold model (represent- initiation rate in RCF-exposed rats, relative to ing the nonlinear, threshold dose-response the background initiation rate in nonexposed curve). Fiber inhalation by rats was equated rats. An MVK model for human lung cancer to humans by determining the fibers/day·kg was then created by fitting the model to the of body weight for the animals and using an age-specific lung cancer incidence for either of exposure scenario of 4 hr/day (consistent with two human cohorts. Finally, the human lung insulation installation workers= schedules), for cancer rate for a given tissue dose was esti- 5 days/week and 50 weeks/year over 40 work- mated by increasing the tumor initiation rate ing years of a 70-year lifespan. RCFC interpret- in the human model by the same proportional ed the results of the analysis with the linearized amount that an identical tissue dose would in- multistage model to represent a risk of 3.8×10-5 crease the initiation rate in the MVK model for for developing lung cancer over the work- rats. The assumption was made that, for any ing lifetime at an exposure concentration of given tissue dose, the proportional increase in 1 f/cm3 [RCFC 1998]. Using the nonlinear the lung tumor initiation rate (relative to the models, estimates of nonsignificant expo- background rate) is the same in humans as in sures (i.e., a working lifetime exposure associ- rats. The two human cohorts used for the hu- ated with no more than 1 additional cancer man modeling were a nonsmoking American

Refractory Ceramic Fibers 85 5 n Effects of Exposure

Cancer Society (ACS) cohort [Peto et al. 1992] (i.e., ACS cohort versus Steel Industry cohort; and a cohort of workers from the steel industry MLE and 95% UCL for exponential, quadratic, (not exposed to coke oven emissions) believed and linear models) but with different exposure to be representative of industrial workers. Be- concentrations. The excess risk was also calcu- cause of the difference in the baseline lung lated for exposure concentrations of 0.75 f/cm3, cancer risk, risk estimates based on the Steel 0.5 f/cm3, and 0.25 f/cm3. These risk estimates Industry cohort were approximately 4 times are presented in Table 5–10. higher than those based on the ACS cohort. Both central estimates (maximum likelihood As shown in Table 5–10, the highest risk esti- estimates [MLEs]) and 95% upper confidence mates at each of the three exposure concen- limits (UCLs) were developed. Three equa- trations are associated with the linear model, tions were tested to describe the relationship followed by the exponential model. The lowest between initiation rate for lung cancer and risk estimates are associated with the quadratic lung burden: model. At each exposure concentration, more conservative risk estimates are obtained for the I=A exp(Bd) (exponential) ACS cohort than the Steel Industry cohort. I=A + Bd2 (quadratic) I=A + Bd (linear) At the recommended exposure guideline established by the RCFC (0.5 f/cm3), the highest risk where d = lung burden in fibers per milligram estimate (linear model, Steel Industry cohort) of lung (which can vary with time) and A and is the MLE of 5.3×10-4 or 5.3/10,000 (95% B are constants (different for each model). UCL=2.9×10-3). At 0.5 f/cm3, the risk estimates With each equation, calculations were made for the steel industry cohort are roughly 1 order to determine the excess risk for a worker aged of magnitude (factor of 10) lower with the ex- 20 to 50 to develop lung cancer by age 70 ponential model (MLE=7.3×10-5, 95% UCL= when exposed to RCFs at a concentration of 9.1×10-5), and 2 orders of magnitude lower 1.0 fiber/cm3 for 8 hr/day, 5 days/week. using the quadratic model (MLE=3.5×10-6, 95% UCL=1.1×10-5). At the lowest exposure Using the exponential model, the excess risk of concentration (0.25 f/cm3), the highest risk 3 lung cancer associated with 1.0 f/cm was esti- estimate (Steel Industry cohort, linear model) -5 -5 mated to be 3.7×10 (MLE) and 4.9×10 (95% was the MLE of 2.7×10-4 (95% UCL=1.4×10-3). UCL), based on the ACS cohort. For the same Again, on average, the risk estimates from the 3 -4 conditions the risk of lung cancer was 1.5×10 models using the steel industry cohort are 3 to -4 (MLE) and 1.8×10 (95% UCL) based on the 4 times higher than for corresponding model Steel Industry cohort. Using a quadratic equa- values with the ACS cohort. tion, the researchers reported slightly lower estimates of excess risk of 4.1×10-6 (MLE) and The authors concluded that the risk estimates 1.2×10-5 (95% UCL) for the ACS cohort, and based on the two cohorts “represent bounds on 1.4×10-5 (MLE) and 4.3×10-5 (95% UCL) for risks likely to be seen in occupational cohorts.” the Steel Industry cohort. The highest esti- However, an occupational cohort is unlikely to mates of excess risk resulted with a linear equa- share the nonsmoking status of the ACS cohort. tion: 2.7×10-4 (MLE) and 1.5×10-3 (95% UCL) Therefore, of the two human populations used for the ACS cohort, and 1.1×10-3 (MLE), and for model fitting in the Moolgavkar et al. [1999] 5.8×10-3 (95% UCL) for the Steel Industry co- risk assessment, the steel industry cohort may hort. Additional risk estimates were calculated be the preferable cohort to use for estimating according to the conditions described above the risks from occupational exposures to RCFs.

86 Refractory Ceramic Fibers 5 n Effects of Exposure

Table 5–10 .Estimates (MLE* and 95% UCL) of excess risk of lung cancer at three exposure concentrations using exponential, quadratic, and linear models for an ACS cohort and a steel industry cohort

ACS cohort Steel industry cohort Exposure Exponential Quadratic Linear Exponential Quadratic Linear

0.75 f/cm3: -6 -4 MLE 2.8×10-5 2.3×10-6 2.0×10-4 1.1×10-4 7.9×10 8.0×10 95% UCL 3.7×10-5 6.8×10-6 1.1×10-3 1.4×10-4 2.4×10-5 4.3×10-3 0.5 f/cm3: MLE 1.8×10-5 1.0×10-6 1.3×10-4 7.3×10-5 3.5×10-6 5.3×10-4 95% UCL 2.5×10-5 3.0×10-6 7.3×10-4 9.1×10-5 1.1×10-5 2.9×10-3

0.25 f/cm3: MLE 9.2×10-6 2.5×10-7 6.7×10-5 3.6×10-5 8.8×10-7 2.7×10-4 95% UCL 1.2×10-5 7.5×10-7 3.6×10-5 4.6 ×10-5 2.7×10-6 1.4×10-3

Adapted from Moolgavkar et al.[ 1999]. *Abbreviations: ACS=American Cancer Society; MLE=maximum likelihood estimate; UCL= 95% upper confidence limit.

The Moolgavkar et al. [1999] report also indi- a methodological improvement over the risk cates airborne fiber concentrations estimated assessment for RCFs by Fayerweather et al. to result in excess lifetime risk for cancer of 10-4 [1997], which was based solely on the inhaled (1 in 10,000) based on the approaches used by fiber concentration. Modeling lung burden do- DECOS [1995] and Fayerweather et al. [1997] simetry should, in theory, compensate for the and using the MVK model for both the ACS known differences between rats and humans cohort and the steel industry cohort. With in fiber deposition and clearance. Similarly, us- the DECOS [1995] linearized, nonthreshold ing an MVK model for dose-response estima- model approach, an excess lifetime cancer risk tion could compensate for differences in cell of 10-4 was calculated to result from a fiber mutation and proliferation rates in rats and concentration of 0.14 f/cm3. Using the linear- humans. However, some key parameter val- ized, multistage model approach described in ues in the MVK and lung dosimetry models Fayerweather et al. [1997], a fiber concentra- are poorly known. For example, the dosimetry tion of 2.6 f/cm3 was estimated to correspond model for humans has been validated with to the excess lifetime cancer risk of 10-4. With only three human tissue samples taken from the MVK exponential model, an excess lifetime workers whose exposures to RCFs were not cancer risk of 10-4 was determined for fiber measured [Yu et al. 1997]. 3 concentrations of 0.7 f/cm for the Steel In- A review and comparison of risk modeling 3 dustry cohort and 2.7 f/cm for the ACS cohort approaches for RCFs by Maxim et al. [2003] [Moolgavkar et al. 1999]. describes the three models here as well as additional more sophisticated variations of quantitative risk analyses for RCFs. Using approach- 5.4.4 Discussion es such as benchmark dose modeling, Maxim The estimated lung fiber burden for dosimetry et al. [2003] produced RCF unit potency values in the analysis by Moolgavkar et al. [1999] is ranging from 1.4×10-4 to 7.2×10-4.

Refractory Ceramic Fibers 87 5 n Effects of Exposure

A common weakness among all three of the risk assessment is based on the assumption risk analyses stems from uncertainty about that a worker is exposed to RCFs for 8 hr/day, possible differences in the sensitivity of human 5 days/week, 52 weeks/year, from age 20 to 50 lungs to fibers, as compared with rat lungs. The [Moolgavkar et al. 1999]. An alternative analy- possibility of such a difference is acknowledged sis, in which the assumption was changed to in the report by Moolgavkar et al. [1999], but 8 hr/day, 5 days/week, 50 weeks/year from age the effect of this uncertainty on the risk esti- 20 to 60, was also described but not presented mates is not explored quantitatively. As an ex- in detail. In both cases, the breathing rate for ample, Pott et al. [1994] estimated that in the light work was assumed to be 13.5 liters/min- case of asbestos fibers, humans are approxi- ute. Additional information might be gained mately 200-fold more sensitive than rats, on from assuming an exposure period of 8 hr/day, the basis of fiber concentration in air. Pott et al. 5 days/week, 50 weeks/year, from age 20 to 65, [1994] further noted that a crocidolite inhala- with a breathing rate matching the Interna- tion study that was negative in the rat resulted tional Commission on Radiological Protection in a rat lung fiber concentration that was more “Reference Man” value for light work, which than 1,000-fold greater than the fiber concen- is 20 liters/minute [ICRP 1994]. In addition, trations in the lungs of asbestos workers with the cumulative excess risk of lung cancer was mesotheliomas. In support of this analysis, re- calculated only through age 70 [Moolgavkar et sults of a study by Rödelsperger and Woitowitz al. 1999]. This practice may underestimate the [1995] led the authors to conclude that humans lifetime risk of lung cancer in the exposed co- are at least 6,000 times more sensitive than rats hort, since a substantial fraction of the cohort to a given tissue concentration of amphibole fi- may be expected to survive beyond age 70. The bers. Although amphibole asbestos fibers have excess risk might also be calculated in a com- physicochemical characteristics which differ peting-risks framework using actuarial meth- from those of RCFs, these findings raise ques- ods until most or all of the cohort is presumed tions about using experimental animal data for to have died because of competing risks (gener- predicting human health effects and assum- ally 85 years). Finally, risk estimates derived by ing that target tissues in humans and rats are Moolgavkar et al. [1999] were based solely on equally sensitive to RCF toxicity. data from studies with rats, ignoring data from The lung cancer risk estimates for RCFs derived studies of hamsters [McConnell et al. 1995]. by Moolgavkar et al. [1999] may also be under- Because 42% of the hamsters in these studies estimated for occupationally exposed workers developed mesotheliomas, using this database because of several basic assumptions made for the risk assessment would produce higher in the lung tissue dosimetry. Tissue dosim- estimates of risk than the analysis based on the etry modeling in the Moolgavkar et al. [1999] rat data.

88 Refractory Ceramic Fibers Discussion and Summary of 6 Fiber Toxicity 6.1 Significance of Studies comparative measures of toxicity for different with RCFs agents. RCFs implanted into the pleural and abdominal cavities of various strains of rats Three major sources of data contributing to and hamsters have produced mesotheliomas, the literature on RCFs are (1) experimental sarcomas, and carcinomas at the sites of fiber studies with animals and in vitro bioassays, (2) implantation [Wagner et al. 1973; Davis et al. epidemiologic studies of populations with oc- 1984; Pott et al. 1987]. Similar tumorigenic cupational exposure to RCFs (primarily dur- responses have been observed following intra- ing manufacturing), and (3) exposure assess- tracheal instillation of RCFs [Manville Cor- ment studies that provide quantitative and poration 1991]. These data provide additional qualitative measurements of exposures as well evidence of the carcinogenic effects of RCFs in as the physical and chemical characteristics of exposed laboratory animals. airborne RCFs. Each of these sources of infor- Epidemiological data have not associated oc- mation is considered integral to this criteria cupational exposure to RCFs under current ex- document for providing a more comprehen- posure conditions with increased incidence of sive evaluation of occupational exposure to pleural mesothelioma or lung cancer [Lockey RCFs and their potential health consequences. et al. 1993; Lemasters et al. 1998]. However, Data from inhalation studies with animals ex- in epidemiologic studies of workers in RCF posed to RCFs have demonstrated statistically manufacturing facilities [Lemasters et al. 1994; significant increases in the induction of lung Lockey et al. 1993, 1996; Rossiter et al. 1994; tumors in rats and mesotheliomas in hamsters Trethowan et al. 1995; Burge et al. 1995; Cowie [Mast et al. 1995a,b; McConnell et al. 1995]. et al. 1999], increased exposures to airborne fi- Other inhalation studies with RCFs have bers have been linked to pleural plaques, small shown pathobiologic inflammatory responses radiographic parenchymal opacities, decreased in lung and pleural tissues [Gelzleichter et al. pulmonary function, respiratory symptoms and conditions (pleurisy, dyspnea, cough), and 1996a,b]. Implantation and instillation meth- skin and eye irritation. ods have also been used in animal studies with RCFs to determine the potential effects Many of the respiratory effects showed a statis- of these fibers on target tissues. These studies tically significant association with RCF expo- have recognized limitations for interpreting re- sure after controlling or adjusting for potential sults because the exposure techniques bypass confounders, including cigarette smoking and the natural defense and clearance mechanisms exposure to nonfibrous dust. Yet in PFTs, the associated with the normal route of exposure interactive effect between smoking and RCF (i.e., inhalation). However, they are useful for exposure was especially pronounced, based demonstrating mechanisms of toxicity and on the finding that RCF-associated decreases

Refractory Ceramic Fibers 89 6 n Discussion and Summary of Fiber Toxicity

in pulmonary function were limited to cur- average exposure concentrations ranging from rent and former smokers [Lockey et al. 1998; 0.05 to 2.6 f/cm3 in RCF facilities in the middle Lemasters et al. 1998; Trethowan et al. 1995; to late 1970s. Rice et al. [1994, 1996, 1997] sug- Burge et al. 1995]. The interactive effect be- gest average concentrations in manufacturing tween exposure to airborne fibers and ciga- ranging from

90 Refractory Ceramic Fibers 6 n Discussion and Summary of Fiber Toxicity

it is important to consider recognized factors Other studies have shown that for processes that contribute to fiber toxicity for RCFs and associated with higher concentrations of air- other SVFs in general. The major determi- borne respirable fibers, there is also a greater nants of fiber toxicity have been identified as concentration of total and respirable dust [Es- fiber dose (or its surrogate, airborne fiber ex- men et al. 1979; Krantz et al. 1994]. posure), fiber dimensions (length and diameter), and fiber durability (especially as it affects fiber biopersistence in the lungs) [Bignon et 6.2.2 Fiber Dimensions al. 1994; Bunn et al. 1992; Bender and Hadley Throughout the literature, studies support 1994; Christensen et al. 1994; Lockey and Wi- the theory that fiber toxicity is related to fi- ese 1992; Moore et al. 2001]. ber dimensions [Timbrell 1982, 1989; Harris and Timbrell 1977; Stanton et al. 1977, 1981; 6.2.1 Fiber Dose Lippmann 1988]. Initially, fiber dimensions (length and diameter) play a significant role in The measurement of airborne fiber concentra- determining the deposition site of a fiber in the tions is frequently used as a surrogate for as- lungs. Longer and thicker (>3.5 μm in diam- sessing dose and health risk to workers. Analy- eter) fibers are preferentially deposited in the ses of historical and current air sampling data upper airways by the mechanisms of impac- indicate that occupational exposure concen- tion [Yu et al. 1986] or interception. Timbrell trations of airborne RCFs have decreased dra- [1965] suggested that direct interception plays matically in the manufacturing sector [Maxim an important role in the deposition of fibers, et al. 1997; Rice et al. 1997]. In chronic inhala- as the fiber comes into contact with the airway tion studies of RCFs [Mast et al. 1995a,b; Mc- wall and is deposited. Fibers being deposited in Connell et al. 1995], both rats and hamsters the larger ciliated airways are generally cleared were exposed to a range of size-separated RCF via the mucociliary escalator. Thinner fibers concentrations in a nose-only inhalation pro- tend to maneuver past airway bifurcations into tocol. When airborne RCFs are generated, half smaller and smaller airways until their dimen- or more of the aerosol is composed of respi- sions dictate deposition either by sedimenta- rable particles of unfiberized material that was tion or diffusion [Asgharian and Yu 1989]. An- formerly a component of the fiber [Mast et al. other factor that may enhance deposition is the 1995a,b]. Because of the nature of this mixed electrostatic charge a fiber can accumulate dur- exposure, it is difficult to determine the rela- ing dust-generating processes in occupational tive contributions of the airborne fibers and settings [Vincent 1985]. The fiber charge may nonfibrous particulates to the adverse effects affect its attraction to the lung surface, causing observed in humans and animals. It has been the fiber to be deposited by electrostatic pre- postulated that the nonfibrous particulates may cipitation. have contributed to an overload effect in the Mast et al. [1995a,b] animal studies with RCFs Although the dimensional characteristics and [Yu et al. 1994; Mast et al. 1995a,b; Maxim et geometry of a fiber influence its deposition al. 1997; Brown et al. 2000]. Burge et al. [1995] in the respiratory tract, the fiber’s length and have suggested that the health effects seen in chemical properties dictate its clearance and RCF-exposed workers are a consequence of retention once it has been deposited within combined particulate and fiber exposure, but the alveolar region. For the fiber that traverses the decrements in lung function are more re- the respiratory airways and is deposited in the lated to fiber exposure combined with smoking. gas exchange region, possible fates include

Refractory Ceramic Fibers 91 6 n Discussion and Summary of Fiber Toxicity

dissolution, clearance via phagocytic cells (al- moved by the macrophages toward the muco- veolar macrophages) in the alveoli, or trans- ciliary escalator and removed from the respi- location through membranes into interstitial ratory tract. The ability of the macrophages tissues. Both test animals and workers have to clear fibers is size-dependent. Short fibers been exposed to RCFs of similar length and di- (<15 µm long) can usually be phagocytized ameter [Allshouse 1995], and these exposures by one rat alveolar macrophage [Luoto et al. include fibers of respirable dimensions [Es- 1994; Morgan et al. 1982; Oberdörster et al. men et al. 1979; Lockey et al. 1990; Cheng et al. 1988, Oberdörster 1994], whereas longer fibers 1992]. Since rats and other rodents are obligate may be engulfed by two or more macrophages. nasal breathers, fibers greater than about 1 μm Blake et al. [1998] have suggested that incom- in diameter are too large for deposition in their plete or frustrated phagocytosis may play a role alveoli [Jones 1993]. By comparison, humans in the increased toxicity of longer fibers. Fiber can inhale and deposit fibers up to 3.5 μm in length has been correlated with the cytotoxicity diameter in the thoracic and gas exchange re- of glass fibers [Blake et al. 1998], with greatest gions of the lung. This physiological difference cytotoxicity for fibers 17 and 33 μm long com- prevents the evaluation of fibers with diameters pared with shorter fiber samples. Long fibers of about 1 to 3.5 μm (which would have hu- (17 µm average length) tend to be a more po- man relevance) in rodent inhalation studies. tent inducer of TNF production and transcription factor activation than short fibers (7 µm The role of fiber size in inducing biological ef- average length) [Ye et al. 1999]. fects is well documented and reviewed in the literature [Stanton et al. 1977, 1981; Pott et al. When comparing the dimensions of airborne 1987; Warheit 1994]. Stanton et al. [1977] hy- fibers with those found in the lungs, it is im- pothesized that glass fibers longer than 8 μm portant to consider the preferential clearance with diameters thinner than 0.25 μm had high of shorter fibers as well as the effects of fiber carcinogenic potential. In a review of the sig- dissolution and breakage. Yu et al. [1996] eval- nificance of fiber size to mesothelioma etiol- uated these factors in a study that led to the ogy, Timbrell [1989] concluded that the thin- development of a clearance model for RCFs in ner fibers with an upper diameter limit of rat lungs. Results of that study confirmed that 0.1 μm are more potent for producing diseases fibers 10 to 20 μm long are cleared more slowly of the parietal pleura (e.g., mesothelioma and than those <10 μm long because of the incom- pleural plaques) than thicker fibers. That value plete phagocytosis of long fibers by macro- for fiber diameter is cited by Lippmann [1988] phages. The preferential clearance of shorter fi- in his asbestos exposure indices for mesothe- bers has also been documented in studies with lioma. Oberdörster [1994] studied the effects chrysotile asbestos and other mineral fibers, in of both long (>10–16 µm) and short (<10 µm) which the average length of retained fibers in- fibers on alveolar macrophage functions, con- creased during a discrete period following de- cluding that both will lead to inflammatory re- position [Coin et al. 1992; Churg 1994]. This actions—although a distinct difference exists increase might also be explained by the lon- in the long-term effects because of differential gitudinal cleavage pattern of asbestos fibers, clearance of fibers of different sizes. Alveolar which results in longer fibers of decreasing macrophages constitute the first line of defense diameters [Coin et al. 1992]. By contrast, any against particles deposited in the alveoli; they breakage of RCFs would occur perpendicular migrate to sites where fibers are deposited and to the longitudinal plane, resulting in shorter phagocytize them. The engulfed fibers are then fibers of the same diameter. For the clearance

92 Refractory Ceramic Fibers 6 n Discussion and Summary of Fiber Toxicity

model developed by Yu et al. [1996], the ef- 6.2.3 Fiber Durability fect of fiber breakage was also assessed from Biopersistence (and specifically the retention experimental data and incorporated into the time of the fiber in the lungs) is considered model. The authors concluded that the simul- to be an important predictor of fiber toxicity. taneous effect of fiber breakage and differential Fiber solubility affects the biopersistence of clearance leads only to a small change in fiber fibers deposited within the lung and is a key size distribution in the lung. This result sug- determinant of fiber toxicity. Bender and Had- gests that the dimensions of fibers in the lung are closely related to the dimensions of fibers ley [1994] suggest that some of the important measured in the airborne samples (adjusted considerations of fiber durability include the for deposition); thus, most short fibers in the following: lungs originated as short fibers in airborne ex- ■ Fiber size—particularly length as it re- posures. lates to the dimensions of the alveolar The dimensions of airborne fibers have also macrophages been characterized for workers with occupa- ■ Fiber dissolution rate tional exposure to RCFs. One study of domestic RCF manufacturing facilities found that ap- ■ Mechanical properties of the fibers, in- proximately 90% of airborne fibers were <3 µm cluding partially dissolved and/or digest- in diameter, and 95% of airborne fibers were ed fibers <4 µm in diameter and <50 µm long [Esmen ■ Overloading of the normal clearance et al. 1979]. The study showed that diameter mechanisms of the lung and length distributions of airborne fibers in

the facilities were consistent with a GMD of 0.7 Bignon et al. [1994] argue that fibers that are µm and a GML of 13 µm. Another air sampling biopersistent in vivo and in vitro are more bio- study of domestic RCF manufacturing sites re- logically active than less durable fibers. ported that 99.7% of the fibers had diameters of <3 µm and 64% had lengths >10 µm [Alls- The durability of RCFs [Hammad et al. 1988; house 1995]. Measurements of airborne fibers Luoto et al. 1995] provides a basis for suggest- in the European RCF manufacturing industry ing that these fibers might persist long enough are comparable: Rood [1988] reported that all to induce biological effects similar to those of fibers observed were in the thoracic and respi- asbestos. In vitro durability tests have shown rable size range (i.e., diameter <3 µm), with RCFs to be highly resistant to dissolution in bi- median diameters ranging 0.5 to 1.0 µm and ologically relevant mixtures such as Gamble’s median lengths from 8 to 23 µm. During resolution [Scholze and Conradt 1987]. The per- moval of RCF products, Cheng et al. [1992] sistence of RCFs in both the peritoneal cavity found that 87% of airborne fibers were within [Bellman et al. 1987] and the lung [Hammad the respirable size range, with fiber diameters et al. 1988] has been recognized in experimen- ranging from 0.5 to 6 µm (median diame- tal studies. Hammad et al. [1988] sacrificed ter=1.6 µm) and fiber lengths ranging from 5 rats exposed to either slag wool or ceramic to 220 µm. Another study [Perrault et al. 1992] fibers via inhalation at 5, 30, 90, 180, or 270 of airborne fiber dimensions measured dur- days after exposure. The lungs of the animals ing installation and removal of RCF materials were ashed in a low-temperature asher, and in industrial furnaces reported GMD values of the fiber content of the lungs was evaluated by 0.38 and 0.57 µm, respectively. PCM. The researchers found that 24% of the

Refractory Ceramic Fibers 93 6 n Discussion and Summary of Fiber Toxicity

deposited RCFs persisted in the lungs of rats time periods. However, degradation and dis- sacrificed 270 days following exposure. In the solution of deposited RCFs may still occur, same study, the lungs of rats exposed to slag based on the findings of higher dissolution wool contained only 6% of the slag wool fibers of aluminum (Al) from RCFs (0.8 to 2.4 mg 270 days after exposure compared with those Al/m2) in alveolar macrophages than from sacrificed 5 days following inhalation. From the other SVFs [Luoto et al. 1995]. In another these results, it was concluded that RCFs fol- study, SEM analysis of fibers recovered from low a clearance pattern of relatively durable the lungs of rats 6 months after inhalation of fibers that persist, translocate, or are removed RCFs revealed an eroded appearance, causing by some mechanism other than dissolution. the researchers to conclude that dissolution Similar results were obtained in the study by of Si in the fibers is a plausible mechanism Mast et al. [1995b], which shows that RCFs for long-term fiber clearance [Yamato et al. are persistent in the lungs of rats exposed by 1994]. inhalation. Specifically, compared with the fiber burden in the lungs of animals sacrificed 3 SVFs in general are less durable than asbestos months after exposure (recovery), the lungs of fibers. RCFs are more durable than many other animals sacrificed after 21 months of recovery SVFs, with a dissolution rate somewhat higher contained approximately 20% of the deposited than chrysotile asbestos. Under the extracellu- fibers. Of the retained fibers (measured with lar conditions in the lung, chrysotile—the most both SEM and TEM techniques) 54% to 75% soluble form of asbestos—has a dissolution rate of <1 to 2 ng/cm2/hr. RCFs have a similar dis- had diameters <0.5 μm, and more than 90% 2 were 5 to 20 μm long. solution rate of about 1 to 10 ng/cm /hr under conditions experienced in pulmonary intersti- Researchers have suggested that fibers depos- tial fluid. Other more soluble SVFs can be 10 ited in the gas exchange region with lengths to 1,000 times less durable [Scholze and Con- less than the diameter of an alveolar macro- radt 1987; Christensen et al. 1994; Maxim et phage are phagocytized and cleared via the al. 1999b; Moore et al. 2001]. At the measured mucociliary system or the lymph channels. solubility rate, an RCF with a 1-μm diameter Dissolution of fibers within the AlM occurs if would take more than 1,000 days to dissolve the fibers are not resistant to the acidic intra- completely [Leineweber 1984]. cellular conditions or a pH~5 [Nyberg et al. 1989]. Fibers that are not engulfed by alveolar macrophages are subjected to a pH of 7.4 6.3 Summary of RCF Toxicity in the extracellular fluid of the lung. A study and Exposure Data of SVF durability in rat alveolar macrophages reports that RCFs are much less soluble than In addition to the main determinants of fi- glass wool and rock wool fibers based on the ber toxicity (dose, dimension, and durability), amounts of silicon (Si) and iron (Fe) dis- other factors such as elemental composition, solved from the fibers in vitro [Luoto et al. surface area, and composition can also influ- 1995]. RCFs in rat alveolar macrophage cul- ence the toxicity of the fiber. Thus, it is difficult ture dissolved less than 10 mg Si/m2 of fiber to predict a fiber’s potential for human toxicity surface area and less than 1 mg Fe/m2 of fiber based solely on in vitro or in vivo tests. Based surface area. Glass wool dissolved more than on consideration of these factors, the major 50 mg Si/m2, and rock wool dissolved nearly findings from the RCF animal and human 2 mg Fe/m2 when measured over comparable studies are as follows:

94 Refractory Ceramic Fibers 6 n Discussion and Summary of Fiber Toxicity

■ Toxicologic evidence from experimental — may induce free radicals, micronu- inhalation studies indicates that RCFs clei, polynuclei, chromosomal break- are capable of producing lung tumors age, and hyperdiploid cells [Brown et in laboratory rats and mesotheliomas in al. 1998; Dopp et al. 1997; Hart et al. hamsters [Mast et al. 1995a,b; McCon- 1992]. nell et al. 1995]. However, interpreting ■ Exposure monitoring results indicate these studies with regard to RCF potency that airborne fibers measured in both the and its implication for occupationally ex- manufacturing and end-use sectors of posed human populations is complicated the RCF industry have dimensions that by the issue of coexposure to fibers and fall within the thoracic and respirable nonfibrous respirable particulate. size ranges [Esmen et al. 1979; Lockey et ■ The durability of RCFs contributes to the al. 1990; Cheng et al. 1992]. biopersistence of these fibers both in vivo ■ Epidemiologic studies of workers in the and in vitro [Bellmann et al. 1987; Schol- RCF manufacturing industry report an ze and Conradt 1987; Lockey and Wiese association between increased exposures 1992]. to airborne fibers and the occurrence of pleural plaques, other radiographic ■ Cytotoxicity and genotoxicity studies indicate that RCFs abnormalities, respiratory symptoms, decreased pulmonary function, and eye — are capable of inducing enzyme re- and skin irritation [Lemasters et al. 1994, lease and cell hemolysis [Wright et al. 1998; Lockey et al. 1996; Trethowan et al. 1986; Fujino et al. 1995; Leikauf et al. 1995; Burge et al. 1995]. Current occupa- 1995; Luoto et al. 1997], tional exposures to RCFs have not been linked to decreases in pulmonary func- — affect mediator release [Morimoto et tion of workers [Lockey et al. 1998]. al. 1993; Ljungman et al.1994; Fujino et al. 1995; Leikauf et al. 1995; Hill et ■ Worker exposure to airborne fiber in the al. 1996; Cullen et al. 1997; Gilmour RCF industry over the past 20 years or et al. 1997; Luoto et al. 1997; Wang et more have decreased substantially, re- al. 1999], portedly as the result of increased hazard awareness and the design and implemen- — may decrease cell viability and in- tation of engineering controls [Rice et al. hibit proliferation [Yegles et al. 1995; 1997; Maxim et al. 1997]. Okayasu et al. 1999; Hart et al. 1992], These observations warrant concern for the — affect cell viability and proliferation continued control and reduction of occupa- [Hart et al. 1992], and tional exposures to airborne RCFs.

Refractory Ceramic Fibers 95 Existing Standards and 7 Recommendations Standards and guidelines for controlling work- lengths ≥10 µm for up to 10 hr/day during a er exposures to RCFs vary in the United States. 40-hr workweek. NIOSH also recommended Other governments and international agencies that airborne concentrations determined as to- have also developed recommendations and oc- tal fibrous glass be limited to a 5 mg/m3 of air cupational exposure limits that apply to RCFs. as a TWA. At that time, NIOSH concluded that Table 7–1 presents a summary of occupational exposure to glass fibers caused eye, skin, and exposure limit standards and guidelines for respiratory irritation. NIOSH also stated that RCFs. until more information became available, this recommendation should be applied to other Within the United States, the RCFC formally SVFs. established a recommended exposure guideline of 0.5 f/cm3 as an element of its product The Agency for Toxic Substances and Disease stewardship program known as PSP 2002, Registry (ATSDR) calculated an inhalation which was endorsed by OSHA as a 5-year vol- minimal risk concentration of 0.03 f/cm3 for untary program [OSHA 2002]. As part of that humans based on extrapolation from animal program, the RCFC recommends that workers studies [ATSDR 2002]. The Agency used mac- wear respirators whenever the workplace fiber rophage aggregation, the most sensitive indica- concentration is unknown or when airborne tor of inflammation from RCFs, as the basis for concentrations exceed 0.5 f/cm3. This exposure this value. Calculation of this value is based on guideline was established by the RCFC on Oc- exposure assumptions for general public health tober 31, 1997, and replaces the previous ex- that differ from those used in models for deter- posure guideline of 1 f/cm3 set by the RCFC in mining occupational exposure limits. 1991. ACGIH proposed a TLV of 0.1 f/cm3 as an 8-hr Before this agreement, the OSHA General In- TWA for RCFs under its notice of intended dustry Standard was most applicable to RCFs, changes to the 1998 TLVs [ACGIH 1998]. On requiring that a worker’s exposure to airborne further review, this concentration was revised dust containing <1% quartz and no asbestos to 0.2 f/cm3 [ACGIH 2000]. ACGIH also clas- be limited to an 8-hr PEL of 5 mg/m3 for respi- sifies RCFs as a suspected human carcinogen rable dust and 15 mg/m3 for total dust [29 CFR (A2 designations) [ACGIH 2005]. On the ba- 1910.1000]. sis of a weight-of-evidence carcinogenic risk assessment, the EPA [1993] classified RCFs as NIOSH has not previously commented on a Group B2 carcinogen (probable human car- occupational exposure to RCFs. However, in cinogen based on sufficient animal data). addressing health hazards for another SVF (fibrous glass), NIOSH [1977] recommended ACGIH and EPA designations are consistent an exposure limit (REL) of 3 f/cm3 as a TWA with that of the International Agency for Re- for glass fibers with diameters ≤3.5 µm and search on Cancer (IARC), which classified

96 Refractory Ceramic Fibers 7 n Exisiting Standards and Recommendations

Table 7–1 . Occupational exposure limits and guidelines pertaining to RCFs*, by country

Country Regulated substance Exposure limit†

Australia Synthetic mineral fibers 0.5 f/cm3 Inspirable dust 2 mg/m3 Austria Total dust (lists superfine fibers as suspected carcinogen) 10 mg/m3 Canada RCFs 0.5 f/cm3 Denmark Manmade mineral fibers 1 f/cm3 Total dust (nonstationary work site) 5 mg/m3 Finland Glass wool and mineral wool 10 mg/m3 France General dust, mineral wool 10 mg/m3 Germany Synthetic vitreous fibers 0.5 f/cm3 Netherlands RCFs 1 f/cm3 New Zealand Synthetic mineral fibers 1 f/cm3 Norway Synthetic mineral fibers 1 f/cm3 Poland Glass wool 2 f/cm3 Total dust 4 mg/m3 Sweden Synthetic inorganic fibers 1 f/cm3 United Kingdom [HSE 2004] Machine-made mineral fibers (except RCFs and special- 2 f/cm3 purpose fibers)

RCFs 1 f/cm3 Total dust (gravimetric limit) 5 mg/m3 United States: ACGIH RCFs 0.2 f/cm3 ATSDR [2002]‡ RCFs 0.03 f/cm3 NIOSH§ RCFs 0.5 f/cm3 Glass fibers, other SVFs [NIOSH 1977] 3 f/cm3 Total fibrous glass 5 mg/m3 OSHA [2002] RCFs 0.5 f/cm3 Respirable dust (<1% quartz, no asbestos) 5 mg/m3 Total dust (<1% quartz, no asbestos) 15 mg/m3

Source: Adapted and updated from U.S. Navy [DOD 1997]. *Abbreviations: ACGIH=American Conference of Governmental Industrial Hygienists; ATSDR=Agency for Toxic Substances Disease Registry; NIOSH=National Institute for Occupational Safety and Health; OSHA=Occupational Safety and Health Administration; RCFs=refractory ceramic fibers; REL=recommended exposure limit; TWA=time-weighted average. †8-hr TWA unless otherwise specified. ‡Inhalation minimal risk level based on general public health assumptions, not occupational exposure. §The NIOSH REL is established as a TWA for up to a 10-hr work shift in a 40-hr workweek.

Refractory Ceramic Fibers 97 7 n Exisiting Standards and Recommendations

ceramic fibers, including RCF, as “possibly car- dusts [Pott 1997] classifying RCFs as category cinogenic to humans (Group 2B)” [IARC 1988, IIIA2, citing “positive results (for carcinogenic- 2002]. The IARC characterization was based ity) from inhalation studies (often supported on “sufficient evidence for the carcinogenicity by the results of other studies with intraperito- of ceramic fibers in experimental animals” and neal, intrapleural, or intratracheal administra- a lack of data on the carcinogenicity of ceramic tion).” fibers to humans [IARC 1988, 2002]. DECOS [1995] determined that “RCFs may pose a car- In the United Kingdom, the Health and Safety cinogenic risk for humans,” and set a health- Commission of the Health and Safety Execu- based recommended occupational exposure tive has established a maximum exposure limit limit of 1 f/cm3. for RCFs of 1.0 f/ml of air, with the additional advisory to reduce exposures to the lowest as The German Commission for the Investigation reasonably practicable concentration based of Health Hazards of Chemical Compounds in on the category 2 carcinogen classification for the Work Area published a review of fibrous RCFs [HSE 2004].

98 Refractory Ceramic Fibers Basis for the Recommended 8 Standard 8.1 Background ditions (including dyspnea, wheezing, coughing, and pleurisy) observed in RCF workers to be In the Occupational Safety and Health Act of adverse health effects associated with exposure 1970 (Public Law 91–96), Congress mandated to airborne RCFs [Lemasters et al. 1998; Lock- that NIOSH develop and recommend criteria ey et al. 1993; Trethowan et al. 1995; Burge et for identifying and controlling workplace haz- al. 1995; Cowie et al. 1999]. ards that may result in occupational illness or injury. In fulfilling this mission, NIOSH con- An association between inhaling RCFs and fi- tinues to investigate the potential health ef- brotic or carcinogenic effects has been docu- fects of exposure to naturally occurring and mented in animals, but no evidence of such synthetic airborne fibers. This interest stems effects has been found in workers in the RCF from the results of research studies confirm- manufacturing industry. The lack of such an ing asbestos fibers as human carcinogens. Sig- association could be influenced by the small nificant increases in the production of RCFs population of workers in this industry, the long during the 1970s and concerns about potential latency period between initial exposure and health effects led to experimental and epide- development of measurable effects, the limited miological studies as well as worker exposure number of persons with extended exposure monitoring. Chronic animal inhalation stud- to elevated concentrations of airborne fibers, ies demonstrated the carcinogenic potential of and declining occupational exposure concen- RCFs, with a statistically significant increase in trations. However, the evidence from animal the incidence of lung cancer or mesothelioma studies suggests that RCFs should be consid- in two laboratory species—rats and hamsters ered a potential occupational carcinogen. This [Bunn et al. 1993; Mast et al. 1995a; McConnell classification is consistent with the conclusions et al. 1995]. Evidence of pleural plaques ob- of ACGIH, EPA, DECOS, and IARC. (See dis- served in persons with occupational exposures cussion in Chapter 7.) to airborne RCFs is clinically similar to that observed in asbestos-exposed persons after the Given these considerations, the NIOSH ob- initial years of their occupational exposures jective in developing an REL for RCFs is to to asbestos [Hourihane et al. 1966; Becklake reduce the possible risk of lung cancer and et al. 1970; Dement et al. 1986]. NIOSH con- mesothelioma. In addition, maintaining ex- siders the discovery of pleural plaques in U.S. posures below the REL will also help to pre- studies of RCF manufacturing workers to be vent other adverse effects, including irritation a significant finding because the plaques were of the skin, eyes, and respiratory tract in ex- correlated with RCF exposure [Lemasters et al. posed workers. To establish an REL for RCFs, 1994; Lockey et al. 1996]. In addition, NIOSH NIOSH took into account not only the animal considers the respiratory symptoms and con- and human health data but also exposure

Refractory Ceramic Fibers 99 8 n Basis for the Recommended Standard

information describing the extent to which Maintaining airborne RCF concentrations at or RCF exposures can be controlled at differ- below the REL requires the implementation of a ent workplaces. On the basis of this evalua- comprehensive safety and health program that in- tion, NIOSH considers an REL of 0.5 f/cm3 cludes routine monitoring of worker exposures, (as a TWA for up to 10 hr/day during a 40-hr installation and routine maintenance of engi- workweek) to be achievable for most work- neering controls, and worker training in good places where RCFs or RCF products are man- work practices. To ensure that worker expo- ufactured, used, or handled. Maintaining ex- sures are routinely maintained below the REL, 3 posures at the REL will minimize the risk for NIOSH recommends that an AL of 0.25 f/cm be part of the workplace exposure monitoring lung cancer and reduce the risk of irritation strategy to ensure that all exposure control ef- of the eyes and upper respiratory system. The forts (e.g., engineering controls and work prac- residual risks of lung cancer at the REL are tices) are in place and working properly. The estimated to be 0.073 to 1.2 per 1,000 based purpose of the AL is to indicate when worker on extrapolations of risk models from Mool- exposures to RCFs may be approaching the gavkar et al. [1999] and Yu and Oberdörster REL. Exposure measurements at or above the [2000]. AL indicate a high degree of certainty that concentrations of RCFs exceed the REL. The AL The risk for mesothelioma at the REL of 3 is a statistically derived concept permitting the 0.5 f/cm is not known but cannot be dis- employer to have confidence (e.g., 95%) that counted. Evidence from epidemiologic studies if exposure measurements are below the AL, showed that higher exposures in the past re- only a small probability exists that the expo- sulted in pleural plaques in workers, indicating sure concentrations are above the REL. When that RCFs do reach pleural tissue. Both implan- exposures exceed the AL, employers should tation studies in rats and inhalation studies in take immediate action (e.g., determine the hamsters have shown that RCF fibers can cause source of exposure, identify measures for con- mesothelioma. Because of limitations in the trolling exposure) to ensure that exposures are hamster data, the risk of mesothelioma cannot maintained below the exposure limit. NIOSH be quantified. However, the fact that no meso- has concluded that an AL allows for the peri- thelioma has been found in workers and that odic monitoring of worker exposures in the pleural plaques appear to be less likely to oc- workplace so that resources do not need to be cur in workers with lower exposures suggests a devoted to conducting daily exposure mea- lower risk for mesothelioma at the REL. surements. The AL concept has been an integral element of recommended occupational Because residual risks of cancer (lung can- standards in NIOSH criteria documents and cer and pleural mesothelioma) and irrita- in comprehensive standards promulgated by tion may exist at the REL, NIOSH further OSHA and MSHA. recommends that all reasonable efforts be made to work toward reducing exposures to less than 0.2 f/cm3. At this concentration, the risks of lung cancer are estimated to be 0.03 8.2 Rationale for the REL to 0.47 per 1,000 based on extrapolations The recommendation to limit occupation- of risk models from Sciences International al exposures to airborne RCFs to a TWA of [1998], Moolgavkar et al. [1999], and Yu and 0.5 f/cm3 is based on data from animal and Oberdörster [2000]. human studies, risk assessments, and the

100 Refractory Ceramic Fibers 8 n Basis for the Recommended Standard

availability of methods to control RCF ex- monitoring period covered under the consent posures at the REL in many workplaces. agreement between RCFC and EPA [Maxim et Establishing the REL for RCFs is consistent al. 1994, 1997, 1998]. NIOSH used the expo- with the mission of NIOSH mandated in sure information to evaluate the feasibility of the Occupational Safety and Health Act of controlling workplace exposures at manufac- 1970. This Act states that NIOSH is obli- turing and end-use facilities where RCFs and gated to “develop criteria dealing with toxic RCF products are handled. materials and harmful physical agents and substances which will describe exposure levels that are safe for various periods of employment, including but not limited to 8.2.1 Carcinogenesis in the exposure levels at which no employee Animal Studies will suffer impaired health or functional Chronic inhalation studies with RCFs demon- capacities or diminished life expectancy as strate significant increases in the incidence of a result of his work experience.” The carci- mesothelioma in hamsters and lung cancer in nogenicity findings from the chronic nose- rats. Tables 8–1 through 8–4 present a synop- only inhalation assays of RCF1 in rats and sis of the major findings of these studies [Mast hamsters [Mast et al. 1995a,b; McConnell et al. 1995] warrant concern about pos- et al. 1995a,b; McConnell et al. 1995]. Results sible health effects in workers exposed to from chronic animal inhalation studies with RCFs. Although no increase in lung can- chrysotile and amosite are also presented (i.e., cer or mesothelioma mortality has been results for the positive control groups); these observed in worker populations exposed data provide a reference point for determining to RCFs, radiographic analyses indicate an the relative toxicity of RCFs [Mast et al. 1995a; association between pleural changes (in- McConnell et al. 1999]. cluding pleural plaques) and RCF exposure [Lemasters et al. 1994; Lockey et al. Chronic inhalation exposure to RCF1 at 3 3 1996; Cowie et al. 1999, 2001]. Both the 30 mg/m (187 WHO f/cm ) induced a U.S. [Lockey et al. 1993; Lemasters et al. 13% (16/123) incidence of lung tumors in F344 1998] and the European [Trethowan et al. rats [Mast et al. 1995a]. The incidence of lung 1995; Burge et al. 1995; Cowie et al. 1999, cancer at lower doses did not show a statisti- 2001] studies have found RCF-associated cally significant difference from the unexposed respiratory symptoms, pulmonary func- control group. Lung fiber burdens in the multi- tion reductions, and pleural abnormalities dose chronic rat study revealed a dose-response among RCF production workers. relationship [Mast et al. 1995b]. In the rat, 16 mg/m3 (120 WHO f/cm3) appeared to Several independent evaluations have quantita- be the NOAEL for lung cancer and 3 mg/m3 tively estimated the risk of lung cancer for work- 3 ers exposed to RCFs at various concentrations (26 WHO f/cm ) appeared to be the NOAEL [DECOS 1995; Fayerweather et al. 1997; Mool- for fibrosis. Although it has been suggested gavkar et al. 1999]. NIOSH evaluated these stud- that fibrosis in animals is a precursor to carci- ies to determine whether an appropriate quali- nogenesis, a definite link has not been shown tative or quantitative assessment of lung cancer for RCFs or other fibers. No lung cancers were risk could be achieved. In addition, exposure found in hamsters exposed to RCF1 [McCon- information was collected during the 5-year nell et al. 1995].

Refractory Ceramic Fibers 101 8 n Basis for the Recommended Standard

Table 8–1 . Doses and dimensions of RCFs* in chronic inhalation studies with F344 rats

Mean fiber Mean fiber WHO Total % Fibers diameter† length† Dose >20 µm Reference Fiber type (mg/m3) f/cm3 SD f/cm3 SD long µm SD µm SD

Mast et al. RCF1 30 187 53 234 35 43 0.98 0.61 22.3 17.0 1995a Mast et al. RCF1 6 120 35 162 37 43 0.98 0.61 22.3 17.0 1995b — — — — — 9 75 35 91 34 — — — — — 3 26 12 36 17 — — — — —

0 0 — 0 — 4 5 Mast et al. Chrysotile 10 1.06 +1.14 × 10 1 × 10 NR 0.10 0.15 2.2 3.0 1995a asbestos

*Abbreviations: NR=not reported; RCFs=refractory ceramic fibers; SD=standard deviation; WHO=World Health Organization. †Arithmetic mean.

Table 8–2 . Results of RCF* chronic inhalation studies with F344 rats

Time of first occurrence (months) Lung Pleural Dose WHO Interstitial Pleural neoplasms mesotheliomas Reference Fiber type (mg/m3) f/cm3 SD fibrosis fibrosis Number % Number %

Mast et al. RCF1 30 187 53 6 9 16/123 13 2/123 1.6 1995a

Mast et al. RCF1 16 120 35 12 12 2/124 1.6 0 — 1995b 9 75 35 12 18 5/127 3.9 1/127 0.8

3 26 12 None None 2/123 1.6 0 —

0 0 — None None 1/129 0.8 0 —

4 Mast et al. Chrysotile 10 1.06 +1.14 × 10 3 9 13/69 18.8 1/69 1.4 1995a asbestos

*Abbreviations: RCF=refractory ceramic fiber; SD=standard deviation; WHO=World Health Organization.

102 Refractory Ceramic Fibers 8 n Basis for the Recommended Standard

Table 8–3 . Doses and dimensions of RCF* in chronic inhalation studies with Syrian golden hamsters

% Fibers Mean fiber Mean fiber WHO Total >20 µm long diameter† length† Dose Reference Fiber type (mg/m3) f/cm3 SD f/cm3 SD % f/cm3 SD µm SD µm SD

McConnell RCF1 30 215 56 256 58 43 — — 0.94 0.63 22.1 16.7 et al. 1995

3 3 4 4 McConnell Chrysotile 10 3.0 × 10 1.4 × 10 8.4 × 10 9.0 × 10 NR — — 0.09 0.06 1.68 2.71 et al. 1995 asbestos

McConnell Amosite 7.1 263 90 NR — ~26 69 24 0.60 0.25 13.4 16.7 et al. 1999 asbestos 3.7 165 61 — — ~23 38 14 — — — — 0.8 36 23 — — ~28 10 6 — — — —

*Abbreviations: NR=not reported; RCFs=refractory ceramic fibers; SD=standard deviation; WHO=World Health Organization. †Arithmetic mean.

Table 8–4 . Results of RCF* chronic inhalation studies with Syrian golden hamsters

Time of first occurrence Hamsters with pleural Dose WHO Interstitial Pleural mesotheliomas† Reference Fiber type (mg/m3) f/cm3 SD fibrosis fibrosis Number %

McConnell RCF1 30 215 56 6 months 6 months 42/123 41.6 et al. 1995

3 3 McConnell Chrysotile 10 3.0 × 10 1.4 × 10 3 months 6 months 0 — et al. 1995 asbestos

McConnell Amosite 7.1 263 90 13 weeks 13 weeks 17/87 19.5 et al. 1995 asbestos 3.7 165 61 13 weeks 13 weeks 22/85 25.9 0.8 36 23 13 weeks 13 weeks 3/83 3.6

*Abbreviations: RCF=refractory ceramic fiber; SD=standard deviation; WHO=World Health Organization. †No lung neoplasms were detected.

Refractory Ceramic Fibers 103 8 n Basis for the Recommended Standard

Chronic inhalation exposure to RCF1 at respectively [McConnell et al. 1999]. A concen- 30 mg/m3 induced a 41% (42/102) incidence tration of 215 RCF1 WHO f/cm3 induced me- of mesotheliomas in Syrian golden ham- sotheliomas in 41% of hamsters [McConnell et sters [McConnell et al. 1995]. Determining al. 1995]. Interstitial and pleural fibrosis were a dose-response relationship for inducing first observed at 13 weeks following amosite mesothelioma is not possible based on current- asbestos exposure and at 6 months following ly available data because of the single exposure RCF1 exposure. Although average fiber dimensions for the RCF1 and amosite asbestos sam- dose tested in the hamster and because of the ples were similar, the RCF1 sample contained a low, sporadic occurrence of mesothelioma in higher percentage of fibers longer than 20 μm the exposed rats [Mast et al. 1995a]. Yet the oc- [McConnell et al. 1995, 1999]. Longer fibers currence of mesotheliomas in the rat and the have been associated with increased toxicity in high incidence in the hamster are biologically experimental animal studies [Davis et al. 1986; significant because the spontaneous incidence Pott et al. 1987; Davis and Jones 1988; Warheit of mesotheliomas in rats and hamsters has his- 1994; Blake et al. 1998]. torically been very low [Analytical Sciences In- corporated 1999]. Results from a dose-response analysis using the mesothelioma data from the RCF and amosite To assess the significance of the mesothelioma asbestos hamster studies [McConnell et al. 1995, incidence observed in RCF-exposed ham- 1999] indicated that the carcinogenic potency sters, these results were compared with those estimates for RCFs ranged from about half to obtained from hamsters that were exposed to two times the carcinogenic potency estimates for chrysotile asbestos and were used as positive amosite asbestos [Dankovic 2001] (see Section controls for the study [McConnell et al. 1995] 5.1.2). This analysis may not predict the meso- (see Tables 8–3 and 8–4). However, the chrysotile- thelioma risk in humans, since RCF1 contained exposed hamsters failed to develop any tumors a greater percentage of fibers longer than 20 µm and therefore could not be considered true and because of differences in fiber durability. positive controls. Based on these results, a po- Amosite asbestos is a more durable fiber with a tency ranking could not be assigned for RCFs longer in vivo half-life than RCF1 [Maxim et al. relative to chrysotile, since the carcinogenic re- 1999b; Hesterberg et al. 1993] (see Table 8–5). sponse rate for the latter was zero in this study. Yet RCFs are more durable and less soluble than The two fibers tested also differed with regard many other types of SVFs that have not demon- to their dose and fiber dimension. strated carcinogenicity in experimental studies. The McConnell et al. [1999] study of hamsters This characteristic is significant, as the durabil- exposed to amosite asbestos provides dose- ity of asbestos and SVFs (including RCFs) may response data for comparison with the RCF1 be linked to the risk of lung cancer in animals data of McConnell et al. [1995] (See Tables 8–3 chronically exposed to these fibers [Bignon et al. and 8–4.). These separate studies examined the 1994; Bender and Hadley 1994; Hammad et al. effects of RCF1 or amosite asbestos in hamsters 1988; Luoto et al. 1995]. Because of the long la- using relatively similar exposure conditions, tency period for the development of mesothelio- experimental conditions, and fiber dimensions mas in humans, Berry [1999] hypothesized that [McConnell et al. 1995, 1999]. Exposure to fibers of sufficient durability are needed to cause 263 WHO f/cm3 and 165 WHO f/cm3 of this disease in humans. Extrapolation of the RCF amosite asbestos induced pleural mesothe- dose-response data for lung cancer and meso- liomas in 20% and 26% of the hamsters, thelioma in exposed rodents should take into

104 Refractory Ceramic Fibers 8 n Basis for the Recommended Standard

Table 8–5 . Dissolution constant (Kdis) and weighted in vivo half-life (t0 .5) of amosite asbestos and RCF1

2 Fiber type Kdis (ng/cm per hr) t0 .5(days)

RCF1 7.6 89.6 Amosite asbestos 1.3 418.0

Source: Adapted from Maxim et al. [1999]. *Abbreviation: RCF=refractory ceramic fiber.

account the durability of RCFs in humans. This work is refuted by other scientists who Some evidence indicates that rats are less sen- favor the rat as an appropriate model for eval- sitive than humans to the development of lung uating the risk evaluation of lung cancer in cancer and mesothelioma from exposure to as- humans [Maxim and McConnell 2001]. Limi- bestos and may therefore represent an inappro- tations of the Rödelsperger and Woitowitz priate model for human risk assessment. Pott et [1995] and Pott [1994] analyses (discussed ear- al. [1994] hypothesized that in chronic inhala- lier) include the lack of a dose-response analy- tion studies, rats may have a lower sensitivity to sis, analysis of only one epidemiologic study, inorganic fiber toxicity than humans. The lung inadequate comparisons of exposure duration, cancer risk from inhaling asbestos may be two lack of accounting for the potentially multipli- orders of magnitude lower in rats than in hu- cative effect of smoking and asbestos exposure, mans, and the mesothelioma risk from inhaling lack of consideration of latency and clearance, asbestos may be three orders of magnitude low- and different fiber measurement techniques. er for rats. Rödelsperger and Woitowitz [1995] measured amphibole fiber concentration in In summary, multiple factors affecting the the lung tissues of humans with mesothelioma comparability of different fiber types and ani- and compared the results with fiber burdens mal models reported in the literature make it reported in rats. A significantly increased OR difficult to determine whether the carcinogen- (OR=4.8, 95%; CI=1.05–21.7) for mesothe- ic potency of RCFs in animals is similar to that lioma was seen in humans with amphibole in humans. Extrapolation of the animal data concentrations between 0.1 and 0.2 fiber/μg to humans is further complicated by a limited of dried lung tissue. The lowest tissue con- understanding of the mechanisms of fiber tox- centration reported to produce a significant icity. Consequently, any extrapolation of the carcinogenic response in rat inhalation stud- cancer risk found in animals exposed to RCFs ies of amphiboles (specifically crocidolite) was must account for differences between humans 1,250 fibers/μg of dried lung tissue. By compar- and rodents with regard to fiber deposition ing these results, Rödelsperger and Woitowitz and clearance patterns, uncertainty about the [1995] estimated that humans are at least 6,000 role of RCF durability for potentiating an ad- times more sensitive than rats to a given tissue verse effect, and possible species differences in concentration of amphibole fibers. sensitivity to fibers.

Refractory Ceramic Fibers 105 8 n Basis for the Recommended Standard

8.2.2 Health Effects Studies of 8.2.2.1 Pleural changes in humans Workers Exposed to RCFs The radiographic analyses of the U.S. and 1996 Two major research efforts evaluated the European populations in RCF manufactur- morbidity of workers exposed to airborne fi- ing detected an association between pleural bers in the RCF manufacturing industry. One changes and RCF exposure [Lemasters et al. study was conducted in the United States and 1994; Lockey et al. 1996; Cowie et al. 1999, the other in Europe. The objective of each was 2001]. In the initial European studies, Tre- to evaluate the relationship between occupa- thowan et al. [1995] found pleural abnormalities in a small number of RCF workers who tional exposure to RCFs and potential ad- had other confounding exposures that did verse health effects. These studies contained not include asbestos. Differences observed in multiple components including standardized pleural abnormalities between the U.S. and respiratory and occupational history ques- European worker populations may be related tionnaires, chest radiographs, pulmonary to the latency of exposure required to cause function tests of workers, and air sampling to specific pleural changes [Hillerdal 1994; Begin estimate worker exposures. The first studies et al. 1996], especially the formation of pleu- of European plants were conducted in 1986 ral plaques, which were first observed in stud- and included current workers at seven RCF ies of the U.S. RCF manufacturing industry, manufacturing plants [Rossiter et al. 1994; with its longer latency period. Historical air Trethowan et al. 1995; Burge et al. 1995]. A sampling data also indicate that airborne fiber followup cross-sectional study conducted in concentrations were much higher in early U.S. 1996 evaluated the same medical endpoints RCF manufacturing. Therefore, in addition in workers from six of these seven European to their longer overall latency, RCF manufac- manufacturing plants (one plant had ceased turing workers in the United States probably operation) [Cowie et al. 1999, 2001]. Cur- had generally higher exposures in the early rent as well as former workers were included years of the industry than did their European as study subjects in the followup study. The counterparts. These factors might explain the studies of U.S. plants began in 1987 and in- appearance of RCF-associated pleural plaques volved evaluations of current workers at five in the U.S. studies before their detection in RCF manufacturing plants and former work- the European studies. The U.S. and 1986 Eu- ropean studies yielded little evidence of an as- ers at two of the plants [Lemasters et al. 1994, sociation between radiographic parenchymal 1998, 2003; Lockey et al. 1993, 1996, 1998, opacities and RCF exposure. In the U.S. study, 2002]. In the United States, the earliest com- small opacities were rare, with only three cases mercial production of RCFs and RCF prod- noted in one report [Lockey et al. 1996] and ucts began in 1953. In Europe, RCF produc- only one case (with small rounded opacities tion began in 1968. The demographics of the of profusion category 3/2 attributable to prior U.S. and European populations were similar kaolin mine work) noted in the other [Lemas- at the time they were studied, but the aver- ters et al. 1994]. Small opacities of profusion age age and duration of employment for the category 1/0 or greater were more frequent in U.S. workers were slightly higher than for the the European study by Trethowan et al. [1995], workforce in the 1986 European studies be- but confounding exposures were believed cause of the earlier development of this in- to account for many of these cases. The re- dustry in the United States. sults of statistical analyses indicated either no

106 Refractory Ceramic Fibers 8 n Basis for the Recommended Standard

association with RCF exposure [Trethowan et noted associations (P<0.05) between employ- al. 1995] or an association slightly suggestive ment in an RCF production job and increased of an RCF exposure effect [Rossiter et al. 1994]. prevalence of dyspnea and the presence of at In a more comprehensive evaluation of the Eu- least one respiratory symptom after adjusting ropean study population, small opacities of the data for confounders. Recurrent chest illness category 1/0 or greater were positively associ- in the European study population was associated with RCF exposures that occurred before ated with the estimated cumulative exposure to 1971 [Cowie et al. 1999]. thoracic-sized fibers but was more strongly associated with estimated cumulative exposure to 8.2.2.2 Respiratory symptoms, irritation, thoracic-sized dust [Cowie et al. 1999, 2001]. and other conditions in humans In cross-sectional analyses, both the U.S. [Lock- In both the U.S. [Lockey et al. 1993; Lemasters et ey et al. 1998; Lemasters et al. 1998] and the al. 1998] and the European [Trethowan et al. 1995; 1986 European [Trethowan et al. 1995; Burge Burge et al. 1995; Cowie et al. 1999, 2001] stud- et al. 1995] studies found that cumulative RCF ies, occupational exposure to RCFs was associated exposure is associated with pulmonary func- with various reported respiratory conditions or tion decrements among current and former irritation symptoms after adjusting for the effects smokers. Lemasters et al. [1998] also found of smoking. Worker exposure to RCFs at concen- statistically significant deficits in pulmonary trations of 0.2 to 0.6 f/cm3 was associated with function measures for nonsmoking female statistically significant increases in eye irritation workers. The decreased pulmonary function (OR=2.16, 95% CI=1.32–3.54), stuffy nose in the European study population remained (OR=2.06, 95% CI=1.25–3.39), and dry cough significantly associated with cumulative RCF (OR=2.53, 95% CI=1.25–5.11) compared with exposure, even after controlling for cumulative exposure to concentrations lower than 0.2 f/cm3 dust exposure [Burge et al. 1995]. The 1996 [Trethowan et al. 1995]. Between the 0.2 to European study found pulmonary function 0.6 f/cm3 and >0.6 f/cm3 RCF exposure groups, decrements only in current smokers [Cowie a statistically significant increase occurred in et al. 1999, 2001]. In a longitudinal analysis ORs for wheezing (P<0.0001), grade 2 dyspnea of data from multiple serial pulmonary func- (P<0.05), eye irritation (P<0.0001), and skin ir- tion tests, Lockey et al. [1998] concluded that ritation (P<0.0001)—but not for stuffy nose [Tre- the more recent RCF concentrations occurring thowan et al. 1995]. Lockey et al. [1993] found that after 1987 were not associated with decreased dyspnea was significantly associated with cumula- pulmonary function; rather, decreases in pul- tive exposure to >15 fiber-months/cm3 (i.e., >1.25 monary function were more closely related to fiber-year/cm3) relative to cumulative exposure to typically higher concentrations that occurred ≤15 fiber-months/cm3 (dyspnea grade 1–OR=2.1, before this time period. The U.S. and European 95% CI 1.3–3.3; dyspnea grade 2–OR=3.8, 95% studies suggest that decrements in pulmonary CI 1.6–9.4) after adjusting for smoking and other function observed primarily in current and potential confounders. Lockey et al. [1993] also former smokers are evidence of an interactive found a statistically significant association be- effect between smoking and RCF exposure. tween cumulative RCF exposure and pleurisy (OR=5.4, 95% CI=1.4–20.2), and an elevated but 8.2.3 Carcinogenic Risk in Humans nonsignificant association between cumulative RCF exposure and chronic cough (OR=2.0, Moolgavkar et al. [1999] derived risk estimates 95% CI=1.0–4.0). Lemasters et al. [1998] also for lung cancer in humans on the basis of the

Refractory Ceramic Fibers 107 8 n Basis for the Recommended Standard

results from the two chronic bioassays of RCFs risk for mesothelioma from the high incidence in male Fischer 344 rats [Mast et al. 1995a,b]. (41%) of mesothelioma in hamsters cannot be Several models (linear, quadratic, exponen- appropriately modeled since only a single ex- tial) were used to estimate and compare risks posure was administered in the study. Primar- using reference populations comprised of ei- ily on the basis of chronic animal inhalation ther a nonsmoking ACS cohort or a cohort of studies [Mast et al. 1995a,b; McConnell et al. steel workers not exposed to coke oven emis- 1995], NIOSH concludes that RCFs are a po- sions (see Table 5–10 for risk estimates). The tential occupational carcinogen. Furthermore, exponential model provided the best statistical the evidence of pleural plaques [Lemasters et fit of the data. The linear model provided the al. 1994; Lockey et al. 1996] observed in per- highest estimates of human lung cancer risks sons with occupational exposures to airborne from exposure to RCFs when used with the RCFs is clinically similar to that observed in as- referent steel workers cohort (considered to bestos-exposed persons after the initial years of be most representative of workers exposed to their occupational asbestos exposures [Houri- RCFs because it includes blue collar workers hane et al. 1966; Becklake et al. 1970; Dement who smoke). Lung cancer risk estimates were et al. 1986]. NIOSH considers the discovery of calculated using each model at exposure con- pleural plaques in U.S. studies of RCF manu- 3 3 3 centrations of 0.25 f/cm , 0.5 f/cm , 0.75 f/cm , facturing workers to be a significant finding and 1.0 f/cm3. The RCF-related lung cancer risk because the plaques were correlated with RCF determined from the linear model for the low- exposure [Lemasters et al. 1994; Lockey et al. est concentration (0.25 f/cm3) was 0.27/1,000 1996]. In addition, NIOSH considers the re- for the cohort of steel workers compared with spiratory symptoms and conditions (including 0.036/1,000 using the exponential model and dyspnea, wheeze, cough, and pleurisy) [Lemas- 0.00088/1,000 for the quadratic model when ters et al. 1998; Lockey et al. 1993; Trethowan using the same referent population. et al. 1995; Burge et al. 1995; Cowie et al. 1999] The risk estimates incorporated multiple as- in RCF workers to be adverse health effects that sumptions, including a human breathing rate have been associated with exposure to airborne of 13.5 L/min (considered light work) and fibers of RCFs. multiple criteria for defining the length of time a worker could be exposed to RCFs over Insufficient evidence exists to document an a working lifetime. Higher risk estimates could association between fibrotic or carcinogenic be expected if the assumptions more closely effects and the inhalation of RCFs by workers represented those used by NIOSH and OSHA: in the RCF manufacturing industry though (1) a human breathing rate of 20 L/min and these effects have been demonstrated in ani- (2) a worker exposure duration of 8 hr/day, mal studies. The lack of an observed associa- 5 days/wk, 50 wk/yr, from age 20 to 65 with tion between RCF exposure and these effects the risk calculated beyond age 70 (e.g., to age among workers could be affected by one or 85). Furthermore, the calculated risk estimates more factors, including several relating to the could be an underestimation of the lung cancer study population: the relatively small cohort, risk to humans because the models assumed the proportion of workers with short tenure that the tissue sensitivity to RCFs in the rat is relative to what might be expected (on the ba- equal to that in humans. Although the lung sis of an asbestos analogy) in terms of disease cancer risk estimates derived from the rat data latency, and workers with limited cumulative are reason for concern, estimates of human exposures to RCFs.

108 Refractory Ceramic Fibers 8 n Basis for the Recommended Standard

8.2.4 Controlling RCF Exposures in to implement. For some operations, such as the Workplace removal of RCF insulation tiles from furnaces, the release of high airborne fiber concentra- Table 8–6 summarizes exposure monitoring tions can occur. However, removal of RCF in- data collected by the RCFC under a consent sulation tiles is not routine and is generally ac- agreement with the EPA [Everest 1998; Maxim complished in a short period of time. Workers et al. 1997]. These data indicate that exposures almost universally wear PPE and respiratory to RCFs during 1993–1998 had an AM fiber protection during these job tasks [Maxim et al. concentration of about 0.3 f/cm3 for manu- 1997, 1998]. facturing and nearly 0.6 f/cm3 for end users. Maxim et al. [1997, 1999a] reported results for NIOSH acknowledges that the frequent use of both manufacturing and end-use sectors in PPE, including respirators, may be required which airborne fiber concentrations through for some workers handling RCFs or RCF prod- 1997 were reduced to an AM <0.3–0.6 f/cm3. ucts. The frequent use of PPE may be required during job tasks for which (1) routinely high The exposure monitoring data collected as part airborne concentrations of RCF (e.g., finish- of the RCFC/EPA consent agreement provide assurance that when appropriate engineering ing, insulation removal) exist, (2) the airborne controls and work practices are used, airborne concentration of RCF is unknown or unpre- exposure to RCFs can be maintained for most dictable, and (3) job tasks are associated with functional job categories (FJCs) at the REL of highly variable airborne concentrations be- 0.5 f/cm3. For many manufacturing processes, cause of environmental conditions or the man- reductions in exposures have resulted from the ner in which the job task is performed. In all improved ventilation, engineering or process work environments where RCFs or RCF prod- changes, and product stewardship programs ucts are handled, control of exposure through [Rice et al. 1996; Maxim et al. 1999b]. These the engineering controls should be the highest data provide the basis for the NIOSH determi- priority. nation that a REL of 0.5 f/cm3 as a TWA can be achieved. 8.3 Summary Although many RCF manufacturing and end- user job tasks have exposures to RCFs at con- The following summarize the relevant infor- centrations below 0.5 f/cm3, exposure moni- mation used as the basis for the NIOSH assess- toring data also indicate that not all FJCs may ment of occupational exposures to RCFs: be able to achieve these RCF concentrations ■ Airborne concentrations of RCFs have consistently. FJCs that currently experience been characterized as containing fibers 3 airborne AM fiber concentrations >0.5 f/cm of dimensions in the thoracic and respi- include finishing (manufacturing and end use) rable size ranges. RCFs are among the and removal (end use). Typical processing dur- most durable types of SVFs. In tests of ing finishing operations (e.g., sawing, drilling, solubility, RCFs are nearly as durable as cutting, sanding) often requires high-energy chrysotile asbestos but significantly less sources that tend to generate larger quantities durable than amphibole asbestos fibers of airborne dust and fibers. For RCF insulation such as amosite. removal, activities are performed at remote sites where conventional engineering controls ■ Chronic, nose-only inhalation studies and fixed ventilation systems are more difficult with RCFs in animals show a statistically

Refractory Ceramic Fibers 109 8 n Basis for the Recommended Standard

Table 8–6 . Airborne fiber concentrations in the RCF* industry during 1993–1998, by functional job category and production status† (f/cm3 as TWA)

Functional job category Minimum First Geometric Arithmetic Third Maximum and production status value quartile Median mean mean quartile value

Total: Manufacturing 0.001 0.070 0.186 0.16 0.313 0.407 7.700 End use 0.002 0.052 0.173 0.16 0.560 0.524 30.000

Assembly: Manufacturing 0.001 0.110 0.208 0.18 0.281 0.366 2.154 End use 0.002 0.050 0.159 0.14 0.316 0.402 2.837

Auxiliary: Manufacturing 0.001 0.019 0.038 0.05 0.112 0.132 1.347 End use 0.002 0.021 0.066 0.07 0.198 0.198 2.678

Fiber: Manufacturing 0.004 0.063 0.145 0.14 0.257 0.299 7.700 End use — — — — — — —

Finishing: Manufacturing 0.004 0.316 0.488 0.47 0.663 0.803 4.044 End use 0.006 0.124 0.383 0.35 0.991 0.986 30.000

Installation: Manufacturing — — — — — — — End use 0.003 0.084 0.236 0.20 0.434 0.559 3.371

Mixing/forming: Manufacturing 0.004 0.090 0.184 0.17 0.292 0.364 1.445 End use 0.010 0.074 0.159 0.17 0.319 0.369 4.109

Other: Manufacturing 0.007 0.027 0.070 0.07 0.112 0.177 1.900 End use 0.003 0.013 0.030 0.04 0.194 0.102 6.400

Removal: Manufacturing — — — — — — — End use 0.010 0.373 1.914 0.82 1.816 2.340 16.000

Source: Adapted from Everest [1998]. *Abbreviations: RCF = refractory ceramic fiber; TWA = time-weighted average. †Fiber concentrations were determined during monitoring performed over a 5-year period (1993–1998) under the Refractory Ceramic Fibers Coalition/Environmental Protection Agency (RCFC/EPA) consent agreement. Concentrations were determined by NIOSH method 7400 “B” counting rules.

110 Refractory Ceramic Fibers 8 n Basis for the Recommended Standard

significant increased incidence of lung ■ RCF exposure data gathered under a con- tumors in rats and pleural mesothelio- sent agreement between RCFC and EPA, mas in hamsters. These data support the which included a 5-year comprehensive air NIOSH determination that RCFs are a monitoring program (1993–1998), indi- potential occupational carcinogen. cate that airborne exposure concentrations to RCFs have been decreasing. Monitoring ■ Epidemiologic studies of workers in the results show that 75% to >95% of TWA RCF manufacturing industry show an exposure concentration measurements in increased incidence of pleural plaques, all FJCs (with one exception) were below respiratory symptoms (dyspnea and 1.0 f/cm3. In all but two of the eight FJCs, cough), skin and eye irritation, and de- >70% of TWA measurements were below creased pulmonary function related to the RCFC recommended exposure guide- increasing exposures to airborne fibers. line of 0.5 f/cm3. On the basis of the re- Some of these conditions are docu- view of these data, NIOSH has concluded mented for exposure concentrations in a that the REL of 0.5 f/cm3 can be achieved 3 range as low as 0.2 to 0.6 f/cm . Studies of in most work places where RCFs or RCF workers exposed to airborne RCFs show products are manufactured or used. no evidence of excess risk for lung cancer or mesothelioma. However, the inability ■ The combined effect of mixed exposures to detect such an association could be be- to fibers and nonfibrous particulates may cause of (1) the low statistical power for contribute to increased irritation of the re- detecting an effect, (2) the short latency spiratory tract, skin, and eyes. Engineering period for most workers occupationally controls and appropriate work practices exposed, and (3) the historically low and used to keep airborne RCF concentrations decreasing fiber exposures that have oc- below the REL should help to minimize air- curred in this industry. borne exposures to nonfibrous particulates as well. Because the ratio of fibers to non- ■ Risk assessment analyses using data from fibrous particulate in airborne exposures chronic inhalation studies in rats indi- may vary among job tasks, exposure moni- cate that the excess risk of developing toring should include efforts to characterize lung cancer when exposed to RCFs at a particulate composition and to control and 3 TWA of 0.5 f/cm for a working lifetime minimize airborne fibers and nonfibrous is 0.073 to 1.2/1,000. However, on the particulate accordingly. basis of the assumptions used in the risk analyses, NIOSH concludes that this risk From the assessment described above and estimate is more likely to underestimate throughout this document, NIOSH concludes than to overestimate the risk to RCF- that on a continuum of fiber toxicity, RCFs exposed workers. Reduction of the RCF relate more closely to asbestos than to fibrous TWA concentration to 0.2 f/cm3 would glass and other SVFs and should be handled ac- reduce the risk for lung cancer to 0.03 to cordingly. NIOSH considers all asbestos types 0.47/1,000. OSHA attempts to set PELs to be carcinogens and has established a REL for carcinogens that reflect an estimated of 0.1 f/cm3 for airborne asbestos fibers. This risk of 1/1,000 but is limited by consider- value was determined on the basis of extensive ations of technologic and economic fea- human and animal health effects data and the sibility. recognized limits of analytical methods.

Refractory Ceramic Fibers 111 8 n Basis for the Recommended Standard

Recognizing that RCFs are carcinogens in ani- From the analysis of historical exposure data mal studies and given the limitations in deriv- (see Chapter 4) and the exposure data collecting an exposure value that reflects no excess risk ed as part of the RCFC/EPA consent agreement of lung cancer or mesothelioma for humans, monitoring program (Table 8–6), NIOSH has NIOSH recommends that every effort be made determined that compliance with the REL for RCFs is achievable in most manufacturing and to keep exposures below the REL of 0.5 f/cm3 end-use job categories. Although routine at- as a TWA for up to 10 hr/day in a 40-hr work- tainment of TWA exposures below the REL week. These efforts will further reduce the risk may not currently occur at all job tasks, it does for malignant respiratory disease and protect represent a reasonable objective that can be workers from conditions and symptoms de- achieved through modification of the job task riving from irritation of the respiratory tract, or the introduction or improvement of venti- skin, and eyes. lation controls.

112 Refractory Ceramic Fibers Guidelines for Protecting 9 Worker Health The following guidelines for protecting worker ■ Establish procedures for reporting haz- health and minimizing worker exposures to ards and giving feedback about actions RCFs are considered minimum precautions taken to correct them. that should be adopted as a part of a site-specific safety and health plan to be developed and over- ■ Instruct workers about using safe work seen by appropriate and qualified personnel. practices and appropriate PPE. ■ Inform workers about practices or oper- 9.1 Informing Workers ations that may generate high concentra- about Hazards tions of airborne fibers (such as cutting and sanding of RCF boards and other 9.1.1 Safety and Health Training RCF products). Program ■ Make workers who remove refractory in- Employers should establish a safety and health sulation materials aware of the following: training program for all workers who manufacture, use, handle, install, or remove RCF — Their potential for exposure to respi- products or perform other activities that bring rable crystalline silica them into contact with RCFs. As part of this — Health effects related to this exposure training program, employers should do the following: — Methods for reducing their exposure ■ Inform all potentially exposed workers (in- — Types of PPE that may be required cluding contract workers) about RCF- (including respirators) associated health risks such as skin, eye, and respiratory irritation and lung cancer. 9.1.2 Labeling and Posting ■ Provide MSDSs on site: Although workers should have received train- — Make MSDSs readily available to ing about RCF exposure hazards and methods workers. for protecting themselves, labels and signs serve — Instruct workers how to interpret in- as important reminders and provide warnings formation from MSDSs. to workers who may not ordinarily work in the ■ Teach workers to recognize and report area. Employers should do the following: adverse respiratory effects associated with ■ Post warning labels and signs about RCF- RCFs. associated health risks at entrances and ■ Train workers to detect hazardous situa- inside work areas where airborne concen- tions. trations of RCFs may exceed the REL.

Refractory Ceramic Fibers 113 9 n Guidelines for Protecting Worker Health

■ State the need to wear appropriate respi- 9.2.1 Engineering Controls ratory protection and protective clothing Engineering controls should be the principal in areas where airborne RCFs may exceed method for minimizing exposure to RCFs in the REL. the workplace. ■ If respiratory protection is required, post the following statement: 9.2.1.1 Ventilation Achieving reduced concentrations of air- RESPIRATORS REQUIRED IN THIS AREA. borne RCFs depends on adequate engineering controls such as local exhaust ventilation systems that are properly constructed and maintained. Local exhaust ventilation systems that ■ Print all labels and warning signs in both employ hoods and ductwork to remove fibers English and the predominant language from the workplace atmosphere have been of workers who do not read English. used by RCF manufacturers. One example is ■ If workers are unable to read the labels a slotted-hood dust collection system placed and signs, inform them verbally about over a mixing tank so that airborne fibers are the hazards and instructions printed on captured and collected in a bag house with the labels and signs. HEPA filters [RCFC 1996]. Other types of local exhaust ventilation or dust collection systems may be used at or near dust-generating 9.2 Hazard Prevention and systems to capture airborne fibers. Band saws Control used in RCF manufacturing and finishing operations have been fitted with such engineer- Proper use and maintenance of engineering ing controls to capture fibers and dust during controls, work practices, and PPE are essential cutting operations and thereby reduce expo- for controlling concentrations of airborne fi- sures for the band saw operator [Venturin bers during the manufacturing, use, and han- 1998]. Disc sanders fitted with similar local dling of RCF products. Minimizing exposure exhaust ventilation systems are effective in to RCFs may be accomplished through a com- reducing airborne RCF concentrations dur- bination of the following work practices and ing sanding of vacuum-formed RCF products controls: [Dunn et al. 2004]. For quality control laboratories or laboratories where production sam- ■ Engineering controls and ventilation ples are prepared for analyses, exhaust venti- ■ Product reformulation lation systems should be designed to capture and contain dust. For guidance in designing ■ Worker isolation local exhaust ventilation systems, see Indus- trial Ventilation—A Manual of Recommended ■ PPE (such as protective clothing and equipment and respirators) Practice, 25th edition [ACGIH 2005], Rec- ommended Industrial Ventilation Guidelines ■ Proper decontamination and waste dis- [Hagopian and Bastress 1976], and the OSHA posal ventilation standard [29 CFR 1910.94].

114 Refractory Ceramic Fibers 9 n Guidelines for Protecting Worker Health

Additional engineering controls have been ■ A test of hand sawing versus the use of evaluated by the Bureau of Mines for mini- a powered jigsaw showed an 81% reduc- mizing airborne dust in underground mining tion in concentrations of airborne dust operations and at industrial sand plants. These generated. controls may also have applications for RCF finishing, installation, and removal operations. ■ A comparison of hand sanding versus The use of air showers (also known as a canopy power sanding showed a 90% reduction air curtain or an overhead air supply island) in- in concentrations of airborne dust gen- volves positioning an air supply over the head erated. of a worker to provide a flow of clean, filtered ■ When a light water mist is applied to the air to the worker’s breathing zone [Volkwein surface of a vacuum-formed board be- et al. 1982, 1988]. Proper design and evalua- fore sanding, airborne dust concentration are critical for ensuring that filtration is tion is reduced by 89% for hand sanding adequate to remove airborne fibers from the and 88% for powered sanding. air supply. Also, selection of the air supply flow rate is important to make sure that the velocity ■ The use of a cork bore (core drill) versus delivered to the worker’s breathing zone is suf- an electric drill with a twist bit for cutting ficient to overcome cross drafts and maintain a holes in RCF product forms reduces air- clean air flow. borne dust concentrations by about 85%.

9.2.1.2 Tool selection and modification 9.2.1.3 Engineering controls for RCF finishing operations The RCFC has reported that using hand tools instead of powered tools can significantly re- Researchers at NIOSH have been working with duce airborne concentrations of dust. How- industrial hygienists at RCFC member facili- ever, hand tools often require additional physi- ties to study the effectiveness of engineering cal effort and time, and they may increase the controls designed and applied to RCF finishing risk of musculoskeletal disorders. Employers operations. Because hand tools are not always should therefore use ergonomically correct a practical solution to manufacturing and end- tools and proper workstation design to avoid use facilities requirements, engineering con- ergonomic hazards. trols are being designed and evaluated for use with powered tools. The additional physical effort required to use hand tools may also increase the rate and A joint project between NIOSH and RCFC was depth of breathing and may consequently af- initiated in 1998 and involved investigating en- fect the inhalation and deposition of fibers. gineering controls for use with a pedestal belt/ For operations such as cutting, sawing, grind- disc sander [Dunn et al. 2000, 2004]. These ing, drilling, and sanding, the high level of me- units are frequently used by the manufactur- chanical energy applied to RCF products with ers as well as the customer facilities to produce power tools increases the potential for elevated vacuum-formed boards sized to the required concentrations of airborne fiber. Examples dimensions. A continuous misting nozzle and [Carborundum 1992] of how airborne fiber simple local exhaust ventilation system were concentrations are affected by the equipment integrated for use on the pedestal sanding unit. used to process RCF products include the fol- The mister consisted of a standard atomization lowing: nozzle that was set for a low-water flow rate to

Refractory Ceramic Fibers 115 9 n Guidelines for Protecting Worker Health

minimize part degradation. The local exhaust 1-ft length of 2- by 4-ft lumber against ventilation system used two hoods or pickup the modules and tapping it lightly with points with a total airflow of 700 ft3/min. a hammer. The process helps ensure that the RCF modules are installed tightly in During production of vacuum-formed boards, place. When a light water spray is applied these two controls reduced fiber concentra- to the surface of the modules before tions in the breathing zone as follows: tamping, airborne fiber concentration is % decrease in airborne fibers: reduced by about 75% [Carborundum 1993]. The water is applied with a gar- Disc sanding using water mist ...... 88 den-type sprayer that is set on mist us- Disc sanding using local exhaust ing about 1 gal of water/100 ft2 of surface ventilation ...... 99 area. However, caution is advised regard- Belt sanding using water mist ...... 50 ing the dampening of refractory-linings during installation. The introduction Belt sanding using local exhaust of water can damage refractory-lined ventilation ...... 99 equipment during heating with explosive These studies highlight the potential for sig- spalling from the generation of steam. nificant reductions in worker exposure using ■ After-service RCF insulation removed well designed and maintained engineering from furnaces may be wetted to reduce controls, but their effectiveness needs to be the release of fibers. validated in the field. 9.2.1.5 Isolation

9.2.1.4 Wet methods for dust Some manufacturing processes may be en- suppression closed to keep airborne fibers contained and separated from workers. Fiber counts are lower in more humid atmo- spheres. Examples of using water to suppress ■ When possible, isolate workers from di- dust concentrations are described as follows: rect contact with RCFs by using auto-

■ At one RCF textile facility, misters have mated equipment operated from a closed been added above broad looms and tape control booth or room. looms to decrease fiber concentrations. ■ Maintain the control room at greater air ■ Water knives are high-pressure water jets pressure than that surrounding the pro- that are used to cut and trim edges of RCF cess equipment so that air flows out rath- blanket while suppressing dust and limit- er than in. ing the generation of airborne fibers. ■ Make sure workers take special precau- ■ During the installation of RCF modules tions (such as using PPE) when they in a furnace, a procedure called tamp- must enter the general work area to per- ing is typically performed. After modules form process checks, adjustments, main- are put in place on the furnace wall, the tenance, assembly-line tasks, and related modules are compressed by placing a operations.

116 Refractory Ceramic Fibers 9 n Guidelines for Protecting Worker Health

9.2.2 Product Reformulation ■ Clean work areas regularly with HEPA- filtered vacuums or with wet sweeping One factor that contributes to the toxicity of an inhaled fiber is the durability of the fiber methods to minimize the accumulation and its resistance to degradation in the respira- of debris. tory tract. The chemical characteristics of RCFs ■ Limit access to areas where workers may make them one of the most durable types of be exposed to airborne RCFs: permit only SVFs. As a result, an inhaled RCF of specific di- workers who are essential to the process mensions will persist longer in the lungs. Mod- or operation. ifying the physical characteristics of RCFs or reformulating their chemistry to produce less Use good hygiene and sanitation to protect durable fibers are recommended options for workers as well as people outside the work- reducing the hazard for exposed workers. Such place who might be contaminated with take- an approach has been taken by one RCF manu- home dust and fibers: facturer in developing two more soluble types of SVF [Maxim et al. 1999b]. However, caution ■ Do not allow workers to smoke, eat, or is advised for developing and distributing such drink in work areas where they contact modified fibers. Possible adverse health effects RCFs. of newly developed fibers should be evaluated before introducing them into commerce. Ap- ■ If RCFs get on the skin, wash with warm propriate testing of these fibers should be per- water and mild soap. formed to provide information about the fiber ■ Apply skin moisturizing cream as needed toxicology and potential adverse health effects to avoid dryness or irritation from re- associated with exposure to these fibers. peated washing.

9.2.3 Work Practices and Hygiene ■ Vacuum contaminated clothes with a HEPA-filtered vacuum before leaving the Use good work practices to help reduce expo- work area. sure to airborne fibers. These practices include the following: — Do not use compressed air to clean the work area or clothing. ■ Limit the use of power tools unless they are equipped with local exhaust or dust — Do not shake clothes to remove dust. collection systems. When possible, use hand tools, which generate less dust and ■ Do not wear contaminated clothes out- fewer airborne particles. side the work area. Instead, take the following measures to prevent taking con- ■ Use HEPA-filtered vacuums, wet sweeping, taminants home: or a properly enclosed wet vacuum system for cleaning up dust containing RCFs. — Change into street clothes before go- ing home. ■ During removal of RCF products, dampen insulation with a light water spray to — Leave contaminated clothes at the keep fibers and dust from becoming air- workplace to be laundered by the em- borne. ployer.

Refractory Ceramic Fibers 117 9 n Guidelines for Protecting Worker Health

— Store street clothes in separate areas ■ Procedures for selecting respirators of the workplace to keep from con- taminating them. ■ Medical evaluations of workers required to wear respirators ■ Provide workers with showers and have them shower before leaving work. ■ Fit testing procedures

■ Prohibit removal of contaminated clothes ■ Routine use procedures and emergency or other items from the workplace [NIOSH respirator use procedures 1995b]. ■ Procedures and schedules for cleaning, 9.2.4 Personal Protective Equipment disinfecting, storing, inspecting, repairing, discarding, and maintaining respi- Wear long sleeves, gloves, and eye protection rators when performing dusty activities involving RCFs. ■ Procedures for ensuring adequate air quality for supplied-air respirators 9.2.5 Respiratory Protection ■ Training in respiratory hazards NIOSH recommends using a respirator for any task involving RCF exposures that are unknown ■ Training in the proper use and mainte- or have been documented to be higher than the nance of respirators NIOSH REL of 0.5 f/cm3 (TWA). Respirators should not be used as the primary means of ■ Program evaluation procedures controlling worker exposures. Instead, NIOSH recommends using other exposure-reduction ■ Procedures for ensuring that workers methods, such as product substitution, engi- who voluntarily wear respirators (exclud- neering controls, and changes in work prac- ing filtering facepiece respirators known tices. However, respirators may be necessary as dust masks) comply with the medi- when available engineering controls and work cal evaluation and cleaning, storing, and practices do not adequately control worker ex- maintenance requirements contained in posures below the REL for RCFs. NIOSH rec- Appendix D of the OSHA respiratory ognizes this control to be a particular challenge protection standard in the finishing stages of RCF product manufacturing as well as during the installation and ■ A designated program administrator who removal of RCF insulation materials. is qualified to administer the respiratory protection program If respiratory protection is needed, the employer should establish a comprehensive re- The written respiratory protection program spiratory protection program as described in should be updated as necessary to account for the OSHA respiratory protection standard changes in the workplace that affect respirator [29 CFR 1910.134]. Elements of a respiratory use. All equipment, training, and medical eval- protection program should be established and uations required under the respiratory protec- described in a written plan that is specific to tion program should be provided at no cost to the workplace. This respirator program must workers. Workers should use only respirators include the following: that have been certified by NIOSH [2002].

118 Refractory Ceramic Fibers 9 n Guidelines for Protecting Worker Health

When airborne RCF concentrations exceed the exception is that filtering-facepiece respirators REL, NIOSH recommends the following respi- (for example, any 95 or 100 series filter) can be ratory protection: used without a respirator protection program when they are used voluntarily. ■ At a minimum, use a half-mask, air- purifying respirator equipped with a For information and assistance in establishing 100 series particulate filter (this res- a respiratory protection program and selecting pirator has an assigned protection factor appropriate respirators, see the OSHA Respi- (APF) of 10. ratory Protection Advisor on the OSHA Web site at http://www.osha.gov. Additional infor- ■ For a higher level of protection and for mation is available from the NIOSH Respira- prevention of facial or eye irritation, use tor Selection Logic [NIOSH 2004] document at a full-facepiece, air-purifying respirator http://www.cdc.gov/niosh/docs/2005–100 and (equipped with a 100 series filter) or any the NIOSH Guide to the Selection and Use of powered, air-purifying respirator equipped Particulate Respirators Certified under 42 CFR with a tight-fitting full facepiece. 84 [NIOSH 1996].

■ For greater respiratory protection when the work involves potentially high air- 9.3 Exposure Monitoring borne fiber concentrations (such as removal of after-service RCF insulation 9.3.1 Workplace Exposure Monitoring such as furnace insulation), use a sup- Program plied-air respirator equipped with a full facepiece, since airborne exposure to The workplace exposure monitoring program RCFs can be high and unpredictable. for worksites where RCFs or RCF products are manufactured, handled, or used should include A comprehensive assessment of workplace ex- routine environmental and personal monitor- posures should always be performed to deter- ing of airborne fiber concentrations. The moni- mine the presence of other possible contami- toring strategy should be designed to assess nants (such as silica) and to ensure that the the effectiveness of engineering controls, work proper respiratory protection is used. Table 9–1 practices, PPE, training, and other factors in provides additional guidance for selecting ap- controlling airborne fiber concentrations. The propriate respiratory protection with regard to monitoring program should also be used to airborne fiber concentrations and the NIOSH identify areas or tasks that are associated with REL for RCFs. higher exposures to RCFs and that therefore require additional efforts to reduce them. Workers may voluntarily choose to use respiratory protection even when airborne fiber The goal of an RCF exposure monitoring concentrations are below the NIOSH REL program is to ensure a more healthful work or other applicable Federal or State stan- environment where worker exposure (mea- dards. When respirators are used voluntarily sured by full-shift samples) does not exceed by workers, employers need to establish only the REL. Because adverse respiratory health those respiratory protection program ele- effects can occur at the REL for RCFs, achiev- ments necessary to assure that the respirator ing lower concentrations is desirable when- itself is not a hazard [29 CFR 1910.134]. The ever possible. For work involving potential

Refractory Ceramic Fibers 119 9 n Guidelines for Protecting Worker Health

Table 9–1 . Respiratory protection for exposure to RCFs*

Airborne concentration of RCFs or conditions of use Minimum respiratory protection options

5.0 f/cm3 (10 × REL) Any air-purifying, elastomeric half-mask respirator equipped with a 100 series (NH, R, or P) filterI

Any negative pressure (demand), suppled-air respirator equipped with a half mask

12.5 f/cm3 (25 × REL) Any powered, air-purifying respirator equipped with a hood or helmet and a high-efficiency particulate air filter (HEPA filter)

Any continuous-flow, supplied-air respirator equipped with a hood or helmet

25 f/cm3 (50 × REL) Any air-purifying, full-facepiece respirator equipped with a 100 series (NH, R, or P) filterI

Any powered, air-purifying respirator equipped with a tight-fitting facepiece (half or full facepiece) and a HEPA filter

Any negative pressure (demand), supplied-air respirator equipped with a full facepiece

Any continuous flow, supplied-air respirator equipped with a tight- fitting facepiece (half or full facepiece)

Any negative pressure (demand), self-contained respirator equipped with a full facepiece

500 f/cm3 (1,000 × REL) Any pressure demand, supplied-air respirator equipped with a half-mask

*Abbreviations: APFs=assigned protection factors; HEPA=high-efficiency particulate air; NIOSH=National Institue for Occupational Safety and Health; RCFs=refractory ceramic fibers. HN-100 series particulate filters should not be used in environments where there is potential for exposure to oil mists. IAssigned protection factors (APFs) for other half-mask and full-facepiece particulate respirators certified under 42 CFR Part 84 are being studied by NIOSH. Recommended APFs for these respirators will be revised accordingly.

120 Refractory Ceramic Fibers 9 n Guidelines for Protecting Worker Health

exposure to airborne RCFs, perform the expo- the employer to have confidence (for example, sure sampling survey as follows: 95%) that if the measured exposure concentration is below the AL, only a small probability ■ Collect representative personal samples exists that the exposure concentration is above for the entire work shift using NIOSH the REL. NIOSH has concluded that the use Method 7400 (B rules) [NIOSH 1977a, of an AL permits employers to monitor RCF 1998]. exposures in the workplace without devot- ■ Perform periodic sampling at least an- ing unnecessary resources to conducting daily nually and whenever any major process exposure measurements. The AL concept has change takes place or another reason ex- served as the basis for defining the elements of ists to suspect that exposure concentra- an occupational standard in NIOSH criteria tions may have changed. documents and in comprehensive standards promulgated by OSHA and MSHA. Employ- ■ Collect all routine personal samples in ers should determine whether the use of an AL the breathing zones of the workers. of 0.25 f/cm3 provides adequate assurance that ■ For workers exposed to concentrations worker exposures are being maintained below above the REL, perform more frequent the REL. In some work environments, the high exposure monitoring until at least two degree of exposure variability for certain job consecutive samples indicate that the tasks may require a lower AL to assure that ex- worker’s exposures no longer exceed the posures are being maintained below the REL. REL. Similar exposure monitoring strategies have been espoused by the American Industrial Hy- ■ Notify all workers about monitoring re- giene Association, which recommends that if sults and any actions taken to reduce measured exposures are less than 10% of the their exposures. designated exposure limit (for example, REL ■ Make sure that any sampling strategy or PEL), there is a high degree of certainty that considers variations in work and pro- the exposure limit will not be exceeded. duction schedules as well as the inherent exposure variability in most workplaces 9.3.3 Sampling Strategies [NIOSH 1995a]. When sampling to determine whether work- 9.3.2 Action Level er exposures are below the REL, a focused sampling strategy may be more practical NIOSH has recommended an action level than random sampling. A focused sampling (AL) of 0.25 f/cm3 for determining when ad- strategy targets workers perceived to have ditional controls are needed or when admin- the highest exposure concentrations [Leidel istrative actions should be taken to reduce and Busch 1994]. A focused strategy is most RCF exposures. The purpose of the AL is to efficient for identifying exposures above the indicate when worker exposures to RCFs may REL if maximum-risk workers and time pe- be approaching the REL. Measurement of ex- riods are accurately identified. Focused sam- posure concentrations at or above the AL in- pling may help identify short-duration tasks dicate that there is a high degree of certainty involving high airborne fiber concentrations that RCF concentrations exceed the REL. The that could result in elevated exposures over a AL is a statistically derived concept permitting full work shift.

Refractory Ceramic Fibers 121 9 n Guidelines for Protecting Worker Health

Sampling strategies such as those used by case of RCFs, medical monitoring is especially Corn and Esmen [1979], Rice et al. [1997], important because achieving compliance with and Maxim et al. [1997] have been derived and the REL of 0.5 f/cm3 does not assure that all used in RCF manufacturing facilities to moni- workers will be free from the risk of respiratory tor airborne fiber concentrations by selecting irritation or chronic respiratory disease caused representative workers for sampling. The rep- by occupational exposure. Early identification resentative workers are grouped according to of respiratory system changes and symptoms dust zones, uniform job titles, or functional job associated with RCF exposures (such as de- categories. These approaches are intended to creased pulmonary function, irritation, dys- reduce the number of required samples while pnea, chronic cough, wheezing, and pleural increasing the confidence of identifying work- plaques) may signal the need for more inten- ers at similar risk. sive medical monitoring and the assessment of existing controls to minimize the risk of long- Area sampling may also be useful in exposure term adverse health effects. An ongoing medi- monitoring for determining sources of air- cal monitoring program also serves to inform borne RCF exposures and assessing the effec- workers of potential health risks and promotes tiveness of engineering controls. an understanding of the need for and support of exposure control activities. 9.4 Medical Monitoring A medical monitoring program serves as an ef- NIOSH recommends periodic medical evalu- fective secondary prevention method on two ation, or medical monitoring, of RCF-exposed levels—screening and surveillance. Medical workers to identify potential health effects and screening in the workplace focuses on the early symptoms that may be related to contact with detection of health outcomes for individual airborne fibers. The following sections describe workers and may involve an occupational his- the objectives of medical monitoring and the tory, medical examination, and application of elements of a medical monitoring program for specific medical tests to detect the presence of workers exposed to RCFs. toxicants or early pathologic changes before the worker would normally seek clinical care The primary goals of a workplace medical for symptomatic disease. By contrast, medical monitoring program are (1) early identifi- surveillance (described in Section 9.5) involves cation of adverse health effects that may be the ongoing evaluation of the health status of related to exposures at work and (2) possible a group of workers through the collection and health trends within groups of exposed work- aggregate analysis of health data for the pur- ers. These goals are based on the premise that pose of preventing disease and evaluating the early detection, subsequent treatment, and effectiveness of intervention programs. workplace interventions will ensure the continued health of the affected workforce. 9.4.2 Criteria for Medical Screening 9.4.1 Objectives of Medical Monitoring To determine whether tests or procedures for Medical monitoring and resulting interven- medical screening are appropriate and relevant tions represent secondary prevention and to a given hazard (in this case, exposure to air- should not replace primary prevention efforts borne RCFs), the following factors should be to minimize worker exposures to RCFs. In the considered:

122 Refractory Ceramic Fibers 9 n Guidelines for Protecting Worker Health

■ Prevalence of an associated disease or may benefit by being included in an occupa- symptoms in the population tional medical monitoring program. Workers should be provided with information about ■ Risk of toxicity associated with the ex- the purposes of medical monitoring, the health posure benefits of the program, and the procedures ■ Consequences of false positive test results involved. The following hierarchy describes workers who should be included in a medical ■ Sensitivity, specificity, and predictive val- monitoring program and who could receive ue of the screening test(s) to be used the greatest benefit from medical screening:

■ Reliability and validity of the screening ■ Workers exposed to elevated fiber con- test(s) centrations (for example, all workers exposed to airborne RCFs at concentrations ■ Ability of the screening test(s) to identify above the AL of 0.25 f/cm3 [described in disease early so that effective treatment Section 9.3]) or intervention may be used to impede disease progression ■ Workers in areas or in jobs and activities in which, regardless of airborne fiber ■ Availability, accessibility, and acceptabil- concentration, one or more workers have ity of followup, further diagnostic tests, recently developed symptoms or respira- and effective management of the disease tory changes apparently related to RCF ■ Benefits of the screening program com- exposure pared with the costs [Wagner 1996]. ■ Workers who may have been previously On the basis of these criteria, NIOSH rec- exposed to asbestos or other respiratory ommends a medical screening program for hazards that place them at an increased RCF-exposed workers that require initial and risk of respiratory disease periodic medical examinations. The elements ■ Workers with potential exposure to air- of the program should include a physical ex- borne RCFs who also smoke cigarettes or amination, occupational history, respiratory other tobacco products (see Section 9.6, symptom questionnaire, spirometric test- Smoking Cessation). ing, and chest radiograph when warranted. If a particular medical screening test indicates 9.4.4 Medical Monitoring Program the presence of exposure-related disease or Director the increased probability that disease will develop, further evaluation and diagnostic test- Oversight of the medical monitoring pro- ing may be needed. Recommended guidelines gram should be assigned by the employer to and schedules for specific medical tests are de- a qualified physician or other qualified health scribed in Section 9.4.5 (Recommended Pro- care provider (as determined by appropriate gram Elements). State laws and regulations) who is informed and knowledgeable about the following: 9.4.3 Worker Participation ■ The administration and management of All workers potentially exposed to RCFs, in a medical monitoring program for occu- both manufacturing and end-use industries, pational hazards

Refractory Ceramic Fibers 123 9 n Guidelines for Protecting Worker Health

■ The establishment of a respiratory protec- course in spirometry or other equivalent tion program based on an understanding training.) of the requirements of the OSHA respi- ■ ratory protection standard and types of A chest X-ray (All chest X-ray films should respiratory protection devices available be interpreted by a NIOSH-certified B at the workplace reader using the standard International Classification of Radiographs of Pneu- ■ The identification and management of moconioses [ILO 2000 or the most recent work-related respiratory effects or illnesses equivalent].)

■ The identification and management of ■ Other medical tests as deemed appropri- work-related skin diseases ate by the attending health care professional 9.4.5 Recommended Program ■ A standardized respiratory symptom ques- Elements tionnaire such as the American Thoracic Society Respiratory Questionnaire [Ferris Recommended elements of a medical moni- 1978 or the most recent equivalent] with toring program for workers exposed to RCFs additional questions to address symptoms include provisions for an initial medical ex- of pleuritic chest pain and pleurisy amination and periodic medical examinations at regularly scheduled intervals. Depending on ■ A standardized occupational history ques- the findings from these examinations, more tionnaire that gathers (1) information frequent and detailed medical examinations about all past jobs (with special emphasis may be necessary. Worker education should on those with potential exposure to dust), also be included as a component of the medi- (2) a description of all duties and poten- cal monitoring program. Specific elements of tial exposures for each job, and (3) a de- the examinations and scheduling are described scription of all protective equipment the below and illustrated in the flow chart diagram worker has used in Figure 9–1. 9.4.5.2 Periodic medical examinations

9.4.5.1 Initial medical examination Periodic examinations (including a physical examination of the respiratory system and the An initial (baseline) examination should be skin, spirometric testing, a respiratory symptom performed as near as possible to the date of be- update questionnaire, and an occupational ginning employment (within 3 months). The history update questionnaire) should be ad- initial medical examination should include the ministered at regular intervals determined by following: the medical monitoring program director. The ■ A physical examination of all systems, frequency of the periodic medical examina- with emphasis on the respiratory system tions should be determined according to the and the skin following guidelines:

■ A spirometric test (Anyone administering ■ For workers with fewer than 10 years spirometric testing as part of the medi- since first exposure to RCFs, periodic ex- cal monitoring program should have aminations should be conducted at least completed a NIOSH-approved training once every 5 years.

124 Refractory Ceramic Fibers 9 n Guidelines for Protecting Worker Health

I. Worker participation? Work involves potential for exposure to RCFs, especially for workers who are ■ exposed to elevated fiber concentrations (for example, above a designated AL),

■ in work areas where one or more workers have May not require medical monitoring recently developed respiratory symptoms or NO changes, ■ previously exposed to asbestos or other respiratory hazards

YES

II. Initial medical examination (within 3 months of beginning employment)

■ Examination of respiratory system and skin ■ Spirometric test ■ Chest X-ray ■ Standardized respiratory symptom questionnaire ■ Standardized occupational history questionnaire

III. Adverse symptoms/health outcomes? V. Periodic medical examination

■ Respiratory symptoms (for example, chronic ■ Examination of respiratory system and skin cough, difficulty breathing, shortness of breath, ■ Spirometric test wheezing) NO ■ Standardized respiratory symptom update ■ Recurrent or chronic dermatitis questionnaire ■ Medically significant reason for additional as- ■ Standardized occupational history update sessment questionnaire YES

IV. More frequent or detailed medical VI. Adverse symptoms/health outcomes? examination and treatment YES (described in III.) (as determined by program director)

VII. Continue with guidance and schedule in V.

Figure 9–1 . Flow chart of medical monitoring guidelines for workers exposed to RCFs. This flow chart is intended as a simplified representation of the minimum requirements of the recommended medical monitoring program guidelines. Administration and management of the program should ultimately rely on the judgment of the medical monitoring program director. Frequency of periodic medical examinations are as follows:

■ If time since first RCF exposure is <10 years, examinations should be conducted at least every 5 years. ■ If time since first RCF exposure is ≥10 years, then examinations should be conducted at least every 2 years.

Refractory Ceramic Fibers 125 9 n Guidelines for Protecting Worker Health

■ For workers with 10 or more years since lung function, or radiographic evidence first exposure to RCFs, periodic examina- of pleural plaques or fibrosis) tions should be conducted at least once ■ every 2 years. History of exposure to other respiratory hazards (for example, asbestos) A chest X-ray and spirometric testing are im- ■ Recurrent or chronic dermatitis portant upon initial examination and may also be appropriate medical screening tests during ■ Other medically significant reason(s) for periodic examinations for detecting respira- more detailed assessment tory system changes—especially in workers with more than 10 years since first exposure 9.4.5.4 Worker education to RCFs. The value of periodic chest X-rays Workers should be provided with sufficient in a medical monitoring program should be training to recognize symptoms associated evaluated by a qualified health care provider in consultation with the worker to assess whether with RCF exposures (such as chronic cough, the benefits of testing warrant the additional difficulty breathing, wheezing, and skin irri- exposure to radiation. As with the frequency tation). Workers should also be instructed to of periodic examinations, the utility of the report these symptoms to designated safety chest X-ray as a medical test becomes greater and health personnel and a physician or other for workers with more than 10 years since first qualified health care provider for appropriate exposure to RCFs (based on the latency period diagnosis and treatment. between first exposure and appearance of no- ticeable respiratory system changes). Because 9.4.6 Written Reports to the Worker persons with advanced fiber-related pleural changes experience difficulty in breathing as Following initial and periodic medical exami- the parietal and visceral surfaces become ad- nations, the physician or other qualified health herent and lose flexibility, it may prove benefi- care provider should provide each worker with cial to detect fibrotic changes in the early stages a written report containing the following: so steps may be taken to prevent further lung damage. Similar recommendations have been ■ Results of any medical tests performed made for asbestos-exposed workers diagnosed on the worker with pleural fibrosis or pleural plaques to pre- ■ Medical opinion in plain language about vent more serious types of respiratory disease any medical condition that would in- [Balmes 1990]. crease the worker=s risk of impairment 9.4.5.3 More frequent medical from exposure to airborne RCFs examinations ■ Recommendations for limiting the Any worker should undergo more frequent and worker=s exposure to RCFs, which may detailed medical evaluation if he or she has any include the use of appropriate PPE, as of the following indications: warranted

■ New or worsening respiratory symptoms ■ Recommendations for further evaluation or findings (for example, chronic cough, and treatment of medical conditions de- difficulty breathing, wheezing, reduced tected

126 Refractory Ceramic Fibers 9 n Guidelines for Protecting Worker Health

9.4.7 Written Reports to the Employer of the medical monitoring program should be provided by the employer at no cost to the Following initial and periodic medical exami- participating workers. If the recommended re- nations, the physician or other qualified health strictions determined by the medical program care provider should provide a written report director include job reassignment, such reas- to the employer containing the following: signment should be implemented with the as- ■ Occupationally pertinent results of the surance of economic protection for the worker. medical evaluation When medical removal or job reassignment is indicated, the affected worker should not suf- ■ A medical opinion about any medi- fer loss of wages, benefits, or seniority. cal condition that would increase the worker=s risk of impairment from expo- The employer should ensure that the medical sure to airborne RCFs monitoring program director communicates regularly with the employer=s safety and health ■ Recommendations for limiting the personnel (such as industrial hygienists), em- worker=s exposure to RCFs (or other ployee representatives, and safety and health agents in the workplace), which may in- committees to identify work areas that may re- clude the use of appropriate PPE or reas- quire evaluation and implementation of con- signment to another job, as warranted trol measures to minimize the risk from expo- ■ A statement to indicate that the worker sure to hazards. has been informed about results of the medical examination and about the medical condition(s) that should have further 9.5 Surveillance of Health evaluation or treatment Outcomes Findings, test results, or diagnoses that have no Standardized medical screening data should bearing on the worker=s ability to work with be periodically aggregated and evaluated by an RCFs should not be included in the report to epidemiologist or other knowledgeable person the employer. Safeguards to protect the con- to identify patterns of worker health that may fidentiality of the worker=s medical records be linked to work activities and practices that should be enforced in accordance with all ap- require additional primary prevention efforts. plicable regulations and guidelines. Routine aggregate assessments of medical screening data should be used in combination 9.4.8 Employer Actions with evaluations of exposure monitoring data to identify changes needed in work areas or ex- The employer should assure that the qualified posure conditions. health care provider=s recommended restrictions of a worker=s exposure to RCFs or to oth- One example of surveillance using analyses of er workplace hazards are followed and that the medical screening data is the ongoing epide- REL for RCFs is not exceeded without requir- miologic study of RCF workers described in the ing the use of PPE. Efforts to encourage worker RCFC product stewardship plan referred to as participation in the medical monitoring pro- PSP 2000 [RCFC 2001]. Elements of this plan gram and to report symptoms promptly to the may be adapted and modified by other employ- program director are essential for the program=s ers to develop medical surveillance programs for success. Medical evaluations performed as part workers who are potentially exposed to RCFs.

Refractory Ceramic Fibers 127 9 n Guidelines for Protecting Worker Health

9.6 Smoking Cessation to smoking in the workplace, NIOSH recommends that employers do the following: NIOSH recognizes a synergistic effect between exposure to RCFs and cigarette smoking that ■ Prohibit workers from smoking in the increases the risk of adverse respiratory health workplace. effects. The combined effects of smoking and ■ Disseminate information about health dust exposures have been recognized as con- promotion and the harmful effects of tributing to the increased risk of respiratory smoking. diseases, including chronic bronchitis, emphy- ■ Offer smoking cessation programs to work- sema, and lung cancer. NIOSH urges employers at no cost to participants. ers to establish smoking cessation programs that (1) inform workers about the increased ■ Collect detailed smoking histories as part hazards of cigarette smoking and exposure to of the medical monitoring program. RCFs and (2) provide assistance and encour- ■ Use training, employee assistance pro- agement for workers who want to quit smok- grams, or health education campaigns to ing. NIOSH recommends that all workers encourage activities promoting physical who are potentially exposed to airborne RCF fitness and other healthy lifestyle prac- fibers and who also smoke should participate tices that affect respiratory and cardio- in a smoking cessation program. With regard vascular health.

128 Refractory Ceramic Fibers 10 Research Needs NIOSH [1993] has developed a fiber research 3. Conduct a series of in vitro and in vivo strategy that proposes the following: animal studies to ensure that fiber toxicity studies share a consistent, stan- ■ Research into the mechanisms for hu- dardized approach. Such studies will man fiber disease ensure comparability of results in a ■ Epidemiologic studies of fiber-exposed variety of experiments that all use well- workers for whom limited or no health characterized, known concentrations data exist of synthetic or natural fibers. A series of controlled, systematic in vitro stud- ■ Toxicologic experiments with fibers for ies of the factors believed to be involved which health effects have not been estab- in RCF pathogenicity should produce lished valuable data on their mechanism of action. In vitro studies provide an ex- The research strategy also considers the useful- cellent opportunity to investigate fiber ness of integrating fiber data from various sci- toxicity factors such as dose, dimen- entific disciplines (toxicology, epidemiology, sion, surface area, and physicochemi- industrial hygiene, occupational medicine) to cal composition. This information is elucidate the characteristics of fibers. an important supplement to data from In addition, NIOSH recommends that the follow- chronic inhalation studies. ing steps be taken with regard to RCF research: 4. Assure that an independent agency or 1. Conduct basic scientific investigations, testing laboratory assembles and keeps including in vitro and in vivo animal a set of reference samples of RCFs (sim- studies, to delineate the mechanism of ilar to the Union Internationale Contre action for RCF toxicity. le Cancer [UICC] asbestos samples). Well-characterized RCF material repre- 2. Conduct comparable studies for oth- sentative of that found in occupational er SVFs and natural fibers so that the exposures could serve as an important mechanistic data can be compared. For component of future animal toxicol- instance, Coffin et al. [1992] examined ogy research into the mechanisms of the ability of different synthetic and fiber-induced disease. Additional SVF natural fibers to induce mesotheliomas. such as fibrous glass, mineral wool, and They suggested that in addition to fiber other ceramic fibers should also be rep- length and width, currently undefined resented in this repository. intrinsic surface characteristics of the fibers are directly related to their meso- 5. Initiate and continue occupational thelioma induction potency. health surveillance for industries that

Refractory Ceramic Fibers 129 10 n Research Needs

manufacture, process, install, or remove concentrations and analyze them to- new fibrous materials. Understanding gether with the health data using epi- of this emerging industry is imperative demiologic research methods. Extend so that exposures to synthetic fibrous surveillance efforts to include assess- materials can be avoided and industry- ments of worker exposure in secondary specific controls can be developed. facilities. 6. Continue and expand surveillance of 7. Assess the effects of variable work sched- RCF exposure in U.S. manufactur- ules (such as shifts longer than 8 hr) on ing facilities. Continue monitoring RCF exposure concentrations and health of airborne fiber and total particulate effects.

130 Refractory Ceramic Fibers References

ACGIH [1998]. 1998 TLVs7 and BEIs7: Thresh- Report prepared for the National Institute of old limit values for chemical substances and Environmental Health Sciences, p. 217. physical agents. Cincinnati, OH: American Conference of Governmental Industrial Hy- Asgharian B, Yu CP [1989]. Deposition of fi- gienists. bers in the rat lung. J Aerosol Sci 20:355–366. Assuncao J, Corn M [1975]. The effects of mill- ACGIH [2000]. Annual reports of the Com- mittees on threshold limit values (TLVs7) and ing on diameters and lengths of fibrous glass and biological exposure indices (BEIs7). Cincinnati, chrysotile asbestos fibers. Am Ind Hyg Assoc J OH: American Conference of Governmental 36(11):811–819. Industrial Hygienists. ATSDR [2002]. Toxicological profile for syn- ACGIH [2004]. Industrial ventilation: a manual thetic vitreous fibers. Atlanta, GA: U.S. Depart- of recommended practice. 25th ed. Cincinnati, ment of Health and Human Services, Public OH: American Conference of Governmental Health Service, Agency for Toxic Substances Industrial Hygienists, Publication No. 2093. and Disease Registry, Draft for public com- ment, September. ACGIH [2005]. 2004 TLVs7 and BEIs7: threshold limit values for chemical substances and physi- Balmes JR [1990]. Medical surveillance for pul- cal agents. Cincinnati, OH: American Confer- monary endpoints. Occup Med: State of the Art ence of Governmental Industrial Hygienists. Rev 5(3):499–513.

Allshouse J [1995]. Memorandum of October Barnhart S [1994]. Irritant bronchitis. In: Rosen- 11, 1995, from John Allshouse, Everest Con- stock L, Cullen MR, eds. Textbook of clinical oc- sulting Associates, Inc., to Kent Hatfield, Ph.D., cupational and environmental medicine. Phila- Education and Information Division, National delphia, PA: W.B. Saunders Company. Institute for Occupational Safety and Health, Baron PA [1996]. Application of the thoracic Centers for Disease Control and Prevention, sampling definition to fiber measurement. Am Public Health Service, U.S. Department of Ind Hyg Assoc J 57(9):820–824. Health and Human Services. Becklake MR, Fournier‑Massey GG, McDon- American Iron and Steel et al. v. OSHA et al. ald JC, Siemiatycki J, Rossiter CE [1970]. Lung [1991]. D.C. Cir. 939 F.2d 975. function in relation to chest radiographic Analytical Sciences Incorporated [1999]. Tu- changes in Quebec asbestos workers. I. Meth- mor incidence in control animals by route ods, results, and conclusions. Bull Physiopathol and vehicle of administration in F344/N rats. Respir 6:637–659.

Refractory Ceramic Fibers 131 References

Begin RO, Samet JM, Shaikh RA [1996]. Asbes- Brain JD, Knudson DE, Sorokin WP, Davis MA tos. Part I. Asbestos‑related diseases, and Part [1976]. Pulmonary distribution of particles II. Asbestos in buildings. In: Harber P, Schen- given by intratracheal instillation or by aerosol ker MB, Balmes JR, eds. Occupational and en- inhalation. Environ Res 11:13–33. vironmental respiratory disease. St. Louis, MO: Mosby‑Year Book, Inc., pp. 293–329. Breysse PN, Lees PSJ, Rooney BC [1999]. Comparison of NIOSH Method 7400 A and B Bellmann B, Muhle H, Pott F, Konig H, Kloppel counting rules for assessing synthetic vitreous H, Spurny K [1987]. Persistence of man‑made fiber exposures. Am Ind Hyg Assoc J 60:526– mineral fibers (MMMF) and asbestos in rat 532. lungs. Ann Occup Hyg 31(4B):693–709. Breysse PN, Rice C, Aubroug A, Komoroski MF, Kalinowski M, Versen R, Goodson J, Carlton R, Bellmann B, Muhle H, Creutzenberg O, Ernst Lees PSJ [1990]. Cowl rinsing procedure for H, Brown RC, Sebastien P [2001].Effects of airborne fiber sampling. Appl Occup Environ nonfibrous particles on ceramic fiber (RCF1) Hyg 5:619–622. toxicity in rats. Inhal Toxicol 13:877–901. Brown DM, Fisher C, Donaldson K [1998]. Bender JR, Hadley JG [1994]. Glass fiber man- Free radical activity of synthetic vitreous fi- ufacturing and fiber safety: the producer’s per- bers: iron chelation inhibits hydroxyl radical spective. Environ Health Perspect 102(Suppl generation by refractory ceramic fiber. J Toxi- 5):37–40. col Environ Health 53:545B561.

Berry G [1999]. Models for mesothelioma in- Brown GM, Cowie H, Davis JMG, Donaldson cidence following exposure to fibers in terms K [1986]. In vitro assays for detecting carcino- of timing and duration of exposure and the genic mineral fibres: a comparison of two as- biopersistence of the fibers. Inhal Toxicol says and the role of fibre size. Carcinogenesis 11:111–130. 17(12):1971–1974. Brown RC, Sébastien P, Bellmann B, Muhle H Bignon J, Saracci R, Touray JC [1994]. Intro- [2000]. Particle contamination in experimen- duction: INSERM‑IARC‑CNRS workshop tal fiber preparations. Inhal Toxicol 12(Suppl on biopersistence of respirable synthetic fi- 3):99–107. bers and minerals. Environ Health Perspect 102(Suppl 5):3–5. Brown SK [1992]. Characterization of the fiber diameter distributions of synthetic mineral fi- Blake T, Castranova V, Schwegler-Berry D, Bar- ber products and their dusts. Am Ind Hyg As- on P, Deye GJ, Li C, Jones W [1998]. Effect of soc J 53(1):27–33. fiber length on glass microfiber cytotoxicity. J Toxicol Environ Health 54:243–259. Buchta TM, Rice CH, Lockey JE, Lemasters GK, Gartside PS [1998]. A comparative study of the Bolton RE, Vincent JH, Jones AD, Addison J, National Institute for Occupational Safety and Beckett ST [1983]. An overload hypothesis for Health 7400 AA@ and AB@ counting rules using pulmonary clearance of UICC amosite fibres refractory ceramic fibers. Appl Occup Environ inhaled by rats. Br J Ind Med 40:264-272. Hyg 13(1):58–61.

132 Refractory Ceramic Fibers References

Bunn WB III, Chase GR, Hesterberg TW, Ver- Cherrie J, Dodgson J, Groat S, Maclaren W sen RA, Anderson R [1992]. Manmade mineral [1986]. Environmental surveys in the Euro- fibers. In: Sullivan JB, Krieger GR, eds. Hazard- pean man-made mineral fiber production in- ous materials toxicology, clinical principles of dustry. Scand J Work Environ Health 12(Suppl environmental health. Baltimore, MD: Wil- 1):18–25. liams and Wilkins, pp. 1139–1150. Christensen VR, Eastes W, Hamilton RD, Struss Bunn WB, Bender JR, Hesterberg TW, Chase AW [1993]. Fiber diameter distributions in GR, Konzen JL [1993]. Recent studies of man- typical MMVF wool insulation products. Am made vitreous fibers. J Occup Med 35(2):101– Ind Hyg Assoc J 54(5):232–238. 113. Christensen VR, Jensen SL, Guldberg M, Kam- Burge PS, Calvert IA, Trethowan WN, Har- strup O [1994]. Effect of chemical composi- rington JM [1995]. Are the respiratory health tion of man-made vitreous fibers on the rate of effects found in manufacturers of ceramic fi- dissolution in vitro at different pHs. Environ bers due to the dust rather than the exposure Health Perspect 102(5):83–86. to fibers? Occup Environ Med 52:105–109. Churg A [1994]. Deposition and clearance of Cantor FL, Gorman RW [1987]. Health haz- chrysotile asbestos. Ann Occup Hyg 38(4):625– ard evaluation report: Niagara Mohawk Power 633. Corporation, Lycoming, NY. Cincinnati, OH: Coffin DL, Cook PM, Creason JP [1992]. Rela- U.S. Department of Health and Human Ser- tive mesothelioma induction in rats by mineral vices, Public Health Service, Centers for Dis- fibers: comparison with residual pulmonary ease Control, National Institute for Occupa- mineral fiber number and epidemiology. Inhal tional Safety and Health, NIOSH Report No. Toxicol 4:273–300. HETA 85–493–1786. Coin PG, Roggli VL, Brody AR [1992]. Deposi- Carborundum [1992]. Workplace quality news: tion, clearance, and translocation of chrysotile evaluation of machining techniques. Niagara asbestos from peripheral and central regions of Falls, NY: The Carborundum Company, Fibers the rat lung. Environ Res 58:97–116. Division, No. 4, Oct. 27. Corn M, Esmen NA [1979]. Workplace expo- Carborundum [1993]. Workplace quality news: sure zones for classification of employee expo- tamping procedure for RCF module furnace sures to physical and chemical agents. Am Ind linings. Niagara Falls, NY: The Carborundum Hyg Assoc J 40:47–57. Company, Fibers Division, No. 6, March 16. Corn M, Lees PSJ, Breysse PN [1992]. Final re- CFR. Code of Federal regulations. Washington, port. Characterization of end‑user exposures DC: U.S. Government Printing Office, Office to industrial (RCF) insulation products. Balti- of the Federal Register. more, MD: Johns Hopkins University. Cheng RT, McDermott HJ, Gia GM, Cover TL, Cornett MJ, Rice C, Hertzberg VS, Lockey JE Duda MM [1992]. Exposures to refractory ce- [1989]. Assessment of fiber deposition on the ramic fiber in refineries and chemical plants. conductive sampling cowl in the refractory ce- Appl Occup Environ Hyg 7(6):361–367. ramic fiber industry. Appl Ind Hyg 4:201–204.

Refractory Ceramic Fibers 133 References

Cowie HA, Beck J, Wild P, Massin N, Aubur- Davis JMG, Addison J, Bolton RE, Donaldson tin G, Piekarski C, Hutchison PA, Russell M, K, Jones AD, Smith T [1986]. The pathogenici- Tomain J-P, Cherrie JW, Groat S, Hurley JF, ty of long versus short fibre samples of amosite Soutar CA [1999]. A study of the respiratory asbestos administered to rats by inhalation health of workers in the European RCF indus- and intraperitoneal injection. Br J Exp Pathol try. Vol. 2. IOM Research Report TM/99/01. 67:415–430.

Cowie HA, Wild P, Beck J, Auburtin G, Piekar- Davis JMG, Jones AD [1988]. Comparisons ski C, Massin N, Cherrie JW, Hurley JF, Miller of the pathogenicity of long and short fibres BG, Groat S, Soutar CA [2001]. An epidemio- of chrysotile asbestos in rats. Br J Exp Pathol logical study of the respiratory health of work- 69:717–737. ers in the European refractory ceramic fibre industry. Occup Environ Med 58:800–810. Davis JMG, Brown DM, Cullen RT, Donaldson Creutzenberg O, Bellmann B, Muhle H [1997]. K, Jones AD, Miller BG, McIntosh C, Searl A Biopersistence and bronchoalveolar lavage in- [1996]. A comparison of methods of deter- vestigations in rats after a subacute inhalation mining and predicting the pathogenicity of of various man-made mineral fibres. Ann Oc- mineral fibers. Inhal Toxicol 8:747–770. cup Hyg 41(Suppl 1):213–218. DECOS [1995]. Man made mineral fibers: Cullen RT, Miller BG, Davis JMG, Brown DM, health based recommended occupational ex- Donaldson K [ 1997]. Short-term inhalation posure limits. The Haag, Netherlands: The and in vitro tests as predictors of fiber patho- Health Council of the Netherlands, Dutch Ex- genicity. Environ Health Perspect 105(Sup- pert Committee on Occupational Standards. pl 5):1235–1240. Dement JM, Wallingford KM [1990]. Com- Dankovic DA [2001]. Memorandum of January parison of phase contrast and electron micro- 22, 2001, from DA Dankovic, Education and In- scopic methods for evaluation of occupational formation Division, Risk Evaluation Branch, to asbestos exposures. Appl Occup Environ Hyg Kathleen MacMahon, Education and Informa- 5:242–247. tion Division, Document Development Branch, National Institute for Occupational Safety and Dement JM, Merchant JA, Green FHY [1986]. Health, Centers for Disease Control and Preven- Asbestosis. In: Merchant JA, ed. Occupational tion, Public Health Service, U.S. Department of respiratory diseases. Cincinnati, OH: U.S. De- Health and Human Services. partment of Health and Human Services, Pub- Davis JMG, Addison J, Bolton RE, Donaldson lic Health Service, Centers for Disease Control, K, Jones AD, Wright A [1984]. The pathogenic National Institute for Occupational Safety and effects of fibrous ceramic aluminum silicate Health, DHHS (NIOSH) Publication No. 86– glass administered to rats by inhalation and 102, pp. 287–327. peritoneal injection. In: Biological effects of man‑made mineral fibers. Proceedings of a DOD [1997]. Man-made vitreous fibers. Nor- WHO/IARC Conference. Vol. 2. Copenhagen, folk, VA: U.S. Department of Defense, U.S. Denmark: World Health Organization/ Inter- Navy, Navy Environmental Health Center, national Agency for Research on Cancer, pp. Navy Environmental Health Center Technical 303–322. Manual NEHC–TM6290.91–1, Rev. A.

134 Refractory Ceramic Fibers References

Dopp E, Schuler M, Schiffmann D, Eastmond Everest [1998]. Review of fifth year workplace DA [1997]. Induction of micronuclei, hyperdip- monitoring. Cranbury, NJ: Everest Consulting loidy and chromosomal breakage affecting the Associates (ECA). centric/pericentric regions of chromosomes 1 Everitt JI, Gelzleichter TR, Bermudez E, Mang- and 9 in human amniotic fluid cells after treat- um JB, Wong BA, Janszen DB, Moss OR [1997]. ment with asbestos and ceramic fibers. Mutat Comparison of pleural responses of rats and Res 377:77–87. hamsters to subchronic inhalation of refracto- Driscoll KE, Costa DL, Hatch G, Henderson ry ceramic fibers. Environ Health Perspect 105 R, Oberdorster G, Salem H, RB Schlesinger (Suppl 5):1209–1213. [2000]. Intratracheal instillation as an expo- Fayerweather WE, Bender JR, Hadley JG, East- sure technique for the evaluation of respira- es W [1997]. Quantitative risk assessment for tory tract toxicity: uses and limitations. Toxicol a glass fiber insulation product. Regul Toxicol Sci 55(1):24–35. Pharmacol 25:103–120. Dunn KH, Venturin DE, Chen SH, and Tread- Ferris BG [1978]. Part II. Recommended re- way JC, Cecala AR, Shulman SA, Cleary JN spiratory disease questionnaires for use with [2000]. Engineering controls for sanding of adults and children in epidemiologic research: refractory ceramic fiber (RCF) materials. Cin- epidemiology standardization project. Am Rev cinnati, OH: U.S. Department of Health and Respir Dis 118(6, Part II):7–53. Human Services, Public Health Service, Cen- Fujino A, Hori H, Higashi T, Morimoto Y, Tana- ters for Disease Control and Prevention, Na- ka I, Kaji H [1995]. In-vitro biological study to tional Institute for Occupational Safety and evaluate the toxic potentials of fibrous materi- Health, ECTB Slides 246–05b. Presented at the als. Int J Occup Environ Health 1:21–28. American Industrial Hygiene Conference and Exposition, Orlando, FL, May 19–26. Gantner BA [1986]. Respiratory hazard from removal of ceramic fiber insulation from high Dunn KH, Shulman SA, Cecala AB, Venturin temperature industrial furnaces. Am Ind Hyg DE [2004]. Evaluation of a local exhaust venti- Assoc J 47(8):530–534. lation system for controlling refractory ceram- Gelzleichter TR, Bermudez E, Mangum JB, ic fibers during disc sanding. J Occup Environ Wong BA, Everitt JI, Moss OR [1996a]. Pulmo- Hyg 1:D107–D111. nary and pleural responses in Fischer 344 rats EPA [1973]. Refractory ceramic fibers (CASRN following short-term inhalation of a synthetic not found). Integrated Risk Information Sys- vitreous fiber. I: Quantitation of lung and pleu- tem (IRIS) Report 0647. Washinton, DC: U.S. ral fiber burdens. Fund Appl Toxicol 30:31–38. Environmental Protection Agency, last revised Gelzleichter TR, Bermudez E, Mangum JB, 07/01/1993, accessed online 01/27/05 at www. Wong BA, Moss OR, Everitt JI [1996b]. Pulmo- epa.gov/iris/subst/0647.htm. nary and pleural responses in Fischer 344 rats following short-term inhalation of a synthet- Esmen NA, Corn M, Hammad YY, Whittier ic vitreous fiber. II: Pathobiologic responses. D, Kotsko N, Haller M, Kahn RA [1979]. Ex- Fund Appl Toxicol 30:39–46. posure of employees to man-made mineral fibers: ceramic fiber production. Environ Res Gelzleichter TR, Bermudez E, Mangum JB, Wong 19:265–278. BA, Janszen DB, Moss OR, Everitt JI [1999].

Refractory Ceramic Fibers 135 References

Comparison of pulmonary and pleural re- Harris RL Jr, Timbrell V [1977]. The influence sponses of rats and hamsters to inhaled refrac- of fibre shape in lung depositionCmathemati- tory ceramic fibers. Toxicol Sci 49:93–101. cal estimates. In: Walton WH, ed. Inhaled particles, IV. Oxford: Pergamon Press, pp. 75–88. Gilmour PS, Beswick PH, Brown DM, Donald- son K [1995]. Detection of surface free radical Hart GA, Newman MM, Bunn WB, Hesterberg activity of respirable industrial fibres using su- TW [1992]. Cytotoxicity of refractory ceramic percoiled oX174 RF1 plasmid DNA. Carcino- fibres to Chinese hamster ovary cells in culture. genesis 16(12):2973–2979. Toxicol In vitro 6(4):317–326. Gilmour PS, Brown DM, Beswick PH, MacNee Hart GA, Kathman LM, Hesterberg TW [1994]. W, Rahman I, Donaldson K [1997]. Free radi- In vitro cytotoxicity of asbestos and man-made cal activity of industrial fibers: role of iron in vitreous fibers: roles of fiber length, diameter oxidative stress and activation of transcription and composition. Carcinogenesis 15(5):971– factors. Environ Health Perspect 105(Suppl 977. 5):1313–1317. Hesterberg TW, Miiller WC, McConnell EE, Gorman RW [1987]. Health hazard evalua- Chevalier J, Hadley JG, Bernstein DM, Theve- tion report: Morris Bean & Company, Yellow naz P, Anderson R [1993]. Chronic inhalation Springs, OH. Cincinnati, OH: U.S. Department toxicity of size-separated glass fibers in Fischer of Health and Human Services, Public Health 344 rats. Fund Appl Toxicol 20:464–476. Service, Centers for Disease Control, National Hewett DJ [1996]. Health hazard evaluation Institute for Occupational Safety and Health, and technical assistance report: Standard Steel, NIOSH Report No. HETA 86–038–1807. Burnham, PA. Morgantown, WV: U.S. Depart- Groat S, Kauffer E, Lovett M, Miller BG, Kidd ment of Health and Human Services, Public MW, Davies LST, McIntosh C, Vigneron JC, Health Service, Centers for Disease Control Cherrie JW, Johnston A, Robertson A, Hurley and Prevention, National Institute for Occupa- JF [1999]. Workplace concentrations of air- tional Safety and Health, NIOSH Report No. borne dust and fibres. Vol. 1. IOM Research HETA 94–0329. Report TM/99/01. Hill GW [1983]. Fibres, man-made glass and mineral. In: Parmeggiani, L, ed. Encyclopaedia Hagopian JH, Bastress EK [1976]. Recom- of occupational health and safety, Vol. 1. Ge- mended industrial ventilation guidelines. neva, Switzerland: International Labour Office, Cincinnati, OH: U.S. Department of Health, pp. 852–855. Education, and Welfare, Public Health Service, Center for Disease Control, National Institute Hill IM, Beswick PH, Donaldson K [1996]. En- for Occupational Safety and Health, DHEW hancement of the macrophage oxidative burst (NIOSH) Publication No. 76–162. by immunoglobulin coating of respirable fibers: fiber-specific differences between asbestos and Hammad Y, Simmons W, Abdel-Kader H, man-made fibers. Exp Lung Res 22(2):133–148. Reynolds C, Weill H [1988]. Effect of chemical composition on pulmonary clearance of man- Hillerdal G [1994]. The human evidence: pa- made mineral fibers. Ann Occup Hyg 32(Suppl renchymal and pleural changes. Ann Occup 1):769–779. Hyg 38(4):561B567.

136 Refractory Ceramic Fibers References

Hourihane O, Lessof L, Richardson PC [1966]. Jaurand C [1997]. Mechanisms of fiber-in- Hyaline and calcified plaques as an index of ex- duced genotoxicity. Environ Health Perspect posure to asbestos: a study of radiological and 105(Suppl 5):1073–1084. pathological features of 100 cases with a con- Jones AD [1993]. Respirable industrial fibers: sideration of epidemiology. Br Med J [Clin Res deposition, clearance and dissolution in ani- Ed] 1:1069–1074. mal models. Ann Occup Hyg 37(2):211–226. HSE [2004]. Refractory ceramic fibers. Health Kamp DW, Gracefa P, Pryor WA, Weitzman SA and Safety Commission Paper HSC/04/06. Unit- [1992]. The role of free radicals in asbestos- ed Kingdom Health and Safety Executive, Health induced diseases. Free Radical Biol Med 12:293– and Safety Commission, cleared May 17. 315. IARC [1988]. IARC monographs on the evalu- Klaunig JE, Yong X, Isenberg JS, Bachowski S, ation of carcinogenic risks to Humans. Vol. 43. Kolaja KL, Jiang J, Stevenson DE, Walborg EF Man‑made mineral fibers and radon. Lyon, Jr. [1998]. The role of oxidative stress in chem- France: World Health Organization. Inter- ical carcinogenesis. Environ Health Perspect national Agency for Research on Cancer, pp. 106(Suppl 1):289–295. 33–171. Kominsky JR [1978]. Health hazard evaluation IARC [2002]. IARC monographs on the evalu- determination report: Heppenstall Company, ation of carcinogenic risks to humans. Vol. 81. Pittsburgh PA. Cincinnati, OH: U.S. Depart- Man‑made vitreous fibers. Lyon, France: World ment of Health, Education, and Welfare, Pub- Health Organization. International Agency for lic Health Service, Center for Disease Control, Research on Cancer. National Institute for Occupational Safety and Health, NIOSH Report No. HHE 77–115–473. ICRP [1994]. Human respiratory tract model for radiological protection. International Com- Krantz S, Christensson B, Lundgren L, Paulsson mission on Radiological Protection Publica- B, Figler B, Persson A [1994]. Exposure to re- tion 66, Tarrytown, NY: Elsevier Science Inc. fractory ceramic fibers in smelters and foundries. Arbets Miljo Institutet (National Institute ILO [1980]. International classification of ra- of Occupational Health) 34. diographs of pneumoconioses: standard films Kromback F, Müunzing S, Allmeling A-M, and guidelines. Occupational Safety and Health Gerlach JT, Behr J, Dörger M [1997]. Cell size Series No. 22, Rev. ed. Geneva, Switzerland: In- of alveolar macrophages: an interspecies com- ternational Labour Office. parison. Environ Health Perspect 105(Suppl ILO [2000]. International classification of ra- 5):1261–1263. diographs of pneumoconioses, Rev. ed. 2000. Lawson CC, LeMasters MK, LeMasters GK, Geneva, Switzerland: International Labour Reutman SS, Rice CH, Lockey JE [2001]. Office. Reliability and validity of chest radiograph surveillance programs. Chest 120:64–68. Janssen YMW, Heintz NH, Marsh JP, Borm PJA, Mossman BT [1994]. Induction of c-fos Leanderson P, Tagesson C [1989]. Cigarette and c-jun Proto-oncogenes in target cells of the smoke potentiates the DNA-damaging effect lung and pleura by carcinogenic fibers. Am J of manmade mineral fibers. Am J Ind Med Respir Cell Mol Biol 11(5):522–530. 16:697–706.

Refractory Ceramic Fibers 137 References

Leanderson P, Soderkvist P, Tagesson C [1989]. Feldman DJ [1998]. An industry-wide pulmo- Hydroxyl radical mediated DNA base modi- nary study of men and women manufactur- fication by manmade mineral fibres. Br J Ind ing refractory ceramic fiber. Am J Epidemiol Med 46:435–438. 148(9):910–919.

Lees PSJ, Breysse PN, McArthur BR, Miller ME, LeMasters GK, Lockey JE, Yiin JH, Hilbert TJ, Rooney BC, Robbins CA, Corn M [1993]. End Levin LS, Rice CH [2003]. Mortality of workers user exposure to man-made vitreous fibers: I. occupationally exposed to refractory ceramic Installation of residential insulation products. fibers. J Occup Environ Med 45:440–450. Appl Occup Environ Hyg 8(12):1022–1030. Lentz TJ, Rice CH, Lockey JE, Succop PA, Le- masters GK [1999]. Potential significance of Leidel NA, Busch KA [1994]. Statistical design airborne fiber dimensions measured in the and data analysis requirements. In: Harris RL, U.S. refractory ceramic fiber manufacturing Cralley LV, eds. Patty=s industrial hygiene and industry. Am J Ind Med 36:286–298. toxicology. 3rd ed. Vol. 3, Part A. New York: John Wiley and Sons, Inc., pp. 453–582. Lewis TR, Morrow PE, McClellan RO, Raabe OG, Kennedy GL, Schwetz BA, Goehl TJ, Roy- Leikauf GD, Fink SP, Miller ML, Lockey JE, croft JH, Chhabra RS [1989]. Cotemporary is- Driscoll KE [1995]. Refractory ceramic fibers sues in toxicology: establishing aerosol expo- activate alveolar macrophage eicosanoid and sure concentrations for inhalation toxicology cytokine release. J Appl Physiol 78(1):164–171. studies. Toxicol Appl Pharm 99:377–383. Leineweber JP [1984]. Solubility of fibres in Lippmann M [1988]. Asbestos exposure indi- vitro and in vivo. In: Biological effects of man- ces. Environ Res 46:86–106. made mineral fibres. World Health Organiza- Lippman M [1990]. Man-made mineral fibers tion, pp. 87–101. (MMMF): human exposures and health risk Lemaire I, Ouellet S [1996]. Distinctive profile assessment. Toxicol Ind Health 6(2):225–246. of alveolar macrophage-derived cytokine re- Ljungman AG, Lindahl M, Tagesson C [1994]. lease induced by fibrogenic and nonfibrogenic Asbestos fibres and man-made mineral fibres: mineral dusts. J Toxicol Env Health 47:465–478. induction and release of tumour necrosis factor-α from rat alveolar macrophages. Occup Lemasters G, Lockey J, Rice C, McKay R, Environ Med 51:777–783. Hansen K, Lu J, Levin L, Gartside P [1991]. Pleural changes in workers manufacturing Lockey J, Lemasters G, Rice C, McKay R, Han- refractory ceramic fiber and products. Un- sen K, Lu J, Levin L, Gartside P, Dimos J, Con- published report. kel B, Ausdenmoore K, Giles D [1990]. An industry‑wide pulmonary morbidity study of Lemasters G, Lockey J, Rice C, McKay R, Han- workers manufacturing refractory ceramic fi- sen K, Lu J, Levin L, Gartside P [1994]. Radio- bers and RCF fibers. Stamford, CT: Thermal graphic changes among workers manufactur- Insulation Manufacturers Association.Unpub- ing refractory ceramic fibre and products. Ann lished report. Occup Hyg 38(Supp. 1):745–751. Lockey JE, Wiese NK [1992]. Health effects Lemasters GK, Lockey JE, Levin L, McKay of synthetic vitreous fibers. Clin Chest Med RT, Rice CH, Horvath EP, Papes DM, Lu JW, 13(2):329–339.

138 Refractory Ceramic Fibers References

Lockey JE, Lemasters GK, Rice CH, McKay Cincinnati, OH: U.S. Department of Health RT, Gartside PS [1993]. A retrospective cohort and Human Services, Public Health Service, morbidity, mortality, and nested case-control Centers for Disease Control and Prevention, study of the respiratory health of individuals National Institute for Occupational Safety and manufacturing refractory ceramic fiber and Health, NIOSH Report No. HETA 91–185. RCF products. Cincinnati, OH: University of Cincinnati. Unpublished report. MacKinnon PA, Lentz TJ, Rice CH, Lockey JE, Lemasters GK, Gartside PS [2001]. Electron Lockey J, Lemasters G, Rice C, Hansen K, Levin microscopy study of refractory ceramic fibers. L, Shipley R, Spitz H, and Wiot J [1996]. Re- Appl Occup Environ Hyg J 16(10):944–951. fractory ceramic fiber exposure and pleural plaques. Am J Respir Crit Care Med 154:1405– Manville Technical Center [1991]. Section 8e 1410. (Toxic Substances Control Act) report on refractory ceramic fibers. Washington, DC: U.S. Lockey JE, Levin LS, Lemasters GK, McKay RT, Environmental Protection Agency, Office of Rice CH, Hansen Kr, Papes DM, Simpson S, Toxic Substances. EPA Document Control No. Medvedovic M [1998]. Longitudinal estimates 8EHQ–0691–0553. Unpublished document. of pulmonary function in refractory ceramic fiber manufacturing workers. Am J Respir Crit Mast RW, McConnell EE, Anderson R, Che- Care Med 157:1226–1233. valier J, Kotin P, Bernstein DM, Thevenaz P, Glass LR, Miiller WC, Hesterberg TW [1995a]. Lockey JE, LeMasters GK, Levin L, Rice C, Yiin Studies on the chronic toxicity (inhalation) of J, Reutman S, Papes D [2002]. A longitudinal four types of refractory ceramic fiber in male study of chest radiographic changes of workers Fischer 344 rats. Inhal Toxicol 7:425–467. in the refractory ceramic fiber industry. Chest 121:2044–2051. Mast RW, McConnell EE, Hesterberg TW, Che- valier J, Kotin P, Thevenaz P, Bernstein DM, Luoto K, Holopainen M, Savolainen K [1994]. Glass LR, Miiller WC, Anderson R [1995b]. Scanning electron microscopic study on the Multiple-dose chronic inhalation toxicity study changes in the cell surface morphology of of size-separated kaolin refractory ceramic rat alveolar macrophages after their exposure fiber in male Fischer 344 rats. Inhal Toxicol to man-made vitreous fibers. Environ Res 66(2):198–207. 7:469–502. Luoto K, Holopainen M, Savolainen K [1995]. Du- Mast RW, Maxim LD, Utell MJ, Walker AM rability of man-made vitreous fibres as assessed [2000]. Refractory ceramic fiber: toxicology, by dissolution of silicon, iron and aluminium epidemiology, and risk analyses—a review. In- in rat alveolar macrophages. Ann Occup Hyg hal Toxicol 12(5):359–399. 39(6):855–867. Maxim LD, Kelly WP, Walters T, Waugh R Luoto K, Holopainen M, Sarataho M, Savolain- [1994]. A multiyear workplace-monitoring en K [1997]. Comparison of cytotoxicity of program for refractory ceramic fibers. Regul man-made vitreous fibres. Ann Occup Hyg Toxicol Pharmacol 20(3):S200–S215. 41(1):37–50. Maxim LD, Allshouse JN, Kelly WP, Walters Lyman MD [1992]. Health hazard evalua- T, Waugh R [1997]. A multiyear workplace- tion report: Thermal Ceramics, Augusta, GA. monitoring program for refractory ceramic

Refractory Ceramic Fibers 139 References

fibers: findings and conclusions. Regul Toxicol inhalation toxicity of a kaolin-based refractory Pharmacol 26:156–171. ceramic fiber in Syrian golden hamsters. Inhal Toxicol 7:503–532. Maxim LD, Allshouse JN, Chen SH, Treadway J, Venturin D [1998]. The development and McConnell EE, Axten C, Hesterberg TW, Che- use of respirator response functions as part of valier J, Miiller WC, Everitt J, Oberdörster G, a workplace exposure monitoring program for Chase GR, Thevenaz P, Kotin P [1999]. Studies control of potential respiratory hazards. Regul on the inhalation toxicology of two fiberglass- Toxicol Pharmacol 27:131–149. es and amosite asbestos in the Syrian golden hamster. Part II: Results of chronic exposure. Maxim LD, Venturin D, Allshouse JN [1999a]. Inhal Toxicol 11:785–835. Respirable crystalline silica exposure associated with the installation and removal of RCF Middleton AP [1982]. Visibility of fine fibres of and conventional silica-containing refractories asbestos during routine electron microscopical in industrial furnaces. Regul Toxicol Pharma- analysis. Ann Occup Hyg 25(1):53–62. col 29(1): 44–63. Moolgavkar SH, Luebeck EG, Turim J, Hanna Maxim LD, Mast RW, Utell MJ, Yu CP, Boymel L [1999]. Quantitative assessment of the risk of PM, Zoitos BK, Cason JE [1999b]. Hazard as- lung cancer associated with occupational ex- sessment and risk analysis of two new syn- posure to refractory ceramic fibers. Risk Anal thetic vitreous fibers. Regul Toxicol Pharmacol 19(4):599–611. 30:54–74. Moore MA, Brown RC, Pigott G [2001]. Ma- terial properties of MMVFs and their time- Maxim LD, Allshouse JN, Chen SH, Treadway dependent failure in lung environments. Inhal JC, Venturin DN [2000a]. Workplace moni- Toxicol 13:1117–1149. toring for refractory ceramic fiber (RCF) in the United States. Regul Toxicol Pharmacol Morgan A, Holmes A, Davison W [1982]. 32:293–309. Clearance of sized glass fibers from the rat lung and their solubility in vivo. Ann Occup Hyg Maxim LD, Allshouse JN, Venturin DE [2000b]. 25(3):317–331. The random-effects model applied to refractory ceramic fiber data. Regul Toxicol Pharmacol Morgan WKC [1994]. Bronchitis, airways’ ob- 32:190–199. struction and occupation. In: Parkes WR, ed. Occupational lung disorders, 3rd ed. Boston, Maxim LD, McConnell EE [2001]. Interspecies MA: Butterworth-Heinemann Ltd. comparisons of the toxicity of asbestos and synthetic vitreous fibers: a weight-of-evidence Morgan WKC [1995]. Epidemiology and oc- approach. Regul Toxicol Pharmacol 33:319–342. cupational lung disease. In: Morgan WKC, Seaton A, eds. Occupational lung diseases, 3rd Maxim LD, Yu CP, Oberdörster G, Utell MJ ed. Philadelphia, PA: W.B. Saunders Company, [2003]. Quantitative risk analyses for RCF: pp. 82–110. survey and synthesis. Regul Toxicol Pharmacol 38:400–416. Morimoto Y, Kido M, Tanaka I, Fujino A, Hi- gashi T, Yokosaki Y [1993]. Synergistic effects of McConnell EE, Mast RW, Hesterberg TW, mineral fibres and cigarette smoke on the pro- Chevalier J, Kotin P, Bernstein DM, Theva- duction of tumour necrosis factor by alveolar naz P, Glass LR, Anderson R [1995]. Chronic macrophage of rats. Br J Ind Med 50:955–960.

140 Refractory Ceramic Fibers References

Morrow PE [1986]. The setting of particulate for Occupational Safety and Health, DHHS exposure levels for chronic inhalation toxicity (NIOSH) Publication No. 95–106. studies. J Am Coll Toxicol 5(6):533–544. NIOSH [1995b]. Report to Congress on work- Muhle H, Bellmann B [1996]. Interim report: ers= home contamination study conducted study of the in vivo solubility of vitreous silicate under the Workers= Family Protection Act (29 fiber dusts. Hanover, Germany: Fraunhofer In- U.S.C. 671a). Cincinnati, OH: U.S. Department stitute for Toxicology and Aerosol Research. of Health and Human Services, Public Health NIOSH [1974]. Criteria for a recommended Service, Centers for Disease Control and Pre- standard: occupational exposure to crystal- vention, National Institute for Occupational line silica. Washington, DC: U.S. Department Safety and Health, DHHS (NIOSH) Publica- of Health, Education, and Welfare, Center for tion No. 95–123. Disease Control, National Institute for Occu- NIOSH [1996]. NIOSH guide to the selection pational Safety and Health, DHEW (NIOSH) and use of particulate respirators certified un- Publication No. 75–120, pp. 54–55, 60–61. der 42 CFR 84. Cincinnati, OH: U.S. Depart- NIOSH [1977a]. Occupational exposure sam- ment of Health and Human Services, Public pling strategy manual. Cincinnati, OH: U.S. De- Health Service, Centers for Disease Control partment of Health, Education, and Welfare, and Prevention, National Institute for Occu- Center for Disease Control, National Institute pational Safety and Health, DHHS (NIOSH) for Occupational Safety and Health, DHEW Publication No. 96–101. (NIOSH) Publication No. 77–173. NIOSH [1998]. NIOSH manual of analytical NIOSH [1977b]. NIOSH criteria for a recom- methods, 4th ed. Cincinnati, OH: U.S. Depart- mended standard: occupational exposure to fi- ment of Health and Human Services, Public brous glass. Cincinnati, OH: U.S. Department Health Service, Centers for Disease Control of Health, Education, and Welfare, Center for and Prevention, National Institute for Occu- Disease Control, National Institute for Occu- pational Safety and Health, DHHS (NIOSH) pational Safety and Health, DHEW (NIOSH) Publication No. 98–119. Publication No. 77–152. NIOSH [2002]. NIOSH certified equipment NIOSH [1993]. NIOSH occupational fiber list as of November, 2004. Cincinnati, OH: U.S. exposures and lung disease: research strategy. Department of Health and Human Services, Cincinnati, OH: U.S. Department of Health Public Health Service, Centers for Disease Con- and Human Services, Public Health Service, trol and Prevention, National Institute for Oc- Centers for Disease Control and Prevention, cupational Safety and Health, DHHS (NIOSH) National Institute for Occupational Safety and Publication No. 2002–144. [http://www2a.cdc. Health. Version 7.4. gov/drds/cel/cel_form_code.asp] NIOSH [1995a]. NIOSH criteria for a recom- NIOSH [2004]. NIOSH respirator selection logic. mended standard: occupational exposure to Cincinnati, OH: U.S. Department of Health and respirable coal mine dust. Cincinnati, OH: U.S. Human Services, Public Health Service, Centers Department of Health and Human Services, for Disease Control and Prevention, National Public Health Service, Centers for Disease Institute for Occupational Safety and Health, Control and Prevention, National Institute DHHS (NIOSH) Publication No. 2005–100.

Refractory Ceramic Fibers 141 References

http://www.cdc.gov/niosh/docs/2005–100/ Perrault G, Dion C, Cloutier Y [1992]. Sam- default.html pling and analysis of mineral fibers on construction sites. Appl Occup Environ Hyg J Nyberg K, Johansson A, Camner P [1989]. In- 7(5):323–326. traphagolysosomal pH in alveolar macrophages studied with fluorescein-labelled amorphous Peters GA, Peters BJ [1980]. Sourcebook on as- silica particles. Exp Lung Res 15:49–62. bestos diseases: medical, legal, and engineering aspects. Chapter 1. Products. New York: Gar- Oberdörster G [1994]. Macrophage-associated re- land STPM Press, pp. A1–A7. sponses to chrysotile. Ann Occup Hyg 38:601– 615. Peto R, Lopez AD, Boreham J, Thun M, Heath C Jr. [1992]. Mortality from tobacco in developed Oberdörster G, Morrow PE, Spurny K [1988]. countries: indirect estimation from national Size dependent lymphatic short term clearance vital statistics. The Lancet 339:1268–1278. of amosite fibres in the lung. Ann Occup Hyg Piguet PF, Collart MA, Grau GE, Sappino A-P, 32(Suppl 1):149–156. Vassalli P [1990]. Requirement of tumour ne- O’Brien DM, Froehlich PA, Gressel MG, Hall crosis factor for development of silica-induced RM, Valiante D, Bost P, Clark NJ [1990]. Senti- pulmonary fibrosis. Nature 344:245–247. nel event notification system for occupational Pott F [1997]. Fibrous dusts. In: Greim H, ed. risks (SENSOR): recommendations for control Occupational toxicants: critical data evalua- of silica exposure at Ingersoll‑Rand Company, tion for MAK values and classification of car- Phillipsburg, NJ. Cincinnati, OH: U.S. Depart- cinogens. Vol. 8. Weinheim, Germany: Verlags- ment of Health and Human Services, Public gesellschanft mbH (VCH). Health Service, Centers for Disease Control, National Institute for Occupational Safety and Pott F, Ziem U, Reiffer FJ, Huth F, Ernst H, Health, NIOSH Report No. ECTB 171–17b. Mohr U [1987]. Carcinogenicity studies of fibres, metal compounds, and some other dusts Okayasu R, Wu L, Hei TK [1999]. Biological in rats. Exp Pathol 32:129–152. effects of naturally occurring and man-made Pott F, Roller M, Kamino K, Bellman B [1994]. fibres: in vitro cytotoxicity and mutagenesis in Significance of durability of mineral fibers for mammalian cells. Br J Cancer 79(9–10):1319– their toxicity and carcinogenic potency in the 1324. abdominal cavity of rats in comparison with OSHA [2002]. OSHA supports program de- the low sensitivity of inhalation studies. Envi- signed to protect workers exposed to refractory ron Health Perspec 102(Suppl 5):145–150. ceramic fibers. U.S. Department of Labor, Oc- RCFC [1993]. Refractory ceramic fiber (RCF) cupational Safety and Health Administration, monitoring project: quality assurance project Office of Communications, OSHA trade news plan. Washington, DC: Refractory Ceramic Fi- release, February 14. bers Coalition. Perkins RC, Scheule RK, Hamilton R, Gomes G, RCFC [1996]. Refractory ceramic fibers: a sub- Freidman G, Holian A [1993]. Human alveolar stitute study. A report prepared by Everest Con- macrophage cytokine release in response to in sulting Associates, submitted to the U.S. Envi- vitro and in vivo asbestos exposure. Exp Lung ronmental Protection Agency, Washington, DC, Res 19:55–65. for the Refractory Ceramic Fibers Coalition.

142 Refractory Ceramic Fibers References

RCFC [1998]. Refractory ceramic fiber. Re- Rood AP [1988]. Size distribution of airborne port presented to the ACGIH Committee on ceramic fibres as determined by transmission Threshold Limit Values for Chemical Sub- electron microscopy. Ann Occup Hyg 32(2): stances and Physical Agents, July 13. 237–240. RCFC [2001]. PSP 2000 product stewardship Rossiter CE, Gilson JC, Sheers G, Thomas HF, strategic plan for refractory ceramic fibers. Trethowan WN, Cherrie JW, Harrington JM [http://www.rcfc.net/psp2000.htm]. [1994]. Refractory ceramic fibre production RCFC [2004]. RCFC Member companies. workers. Analysis of radiograph readings. Ann [http://www.rcfc.net/about.htm]. Occup Hyg 38(Suppl 1):731–738. Rice C, Lockey J, Lemasters G, Dimos J, Gart- Scholze H, Conradt R [1987]. An in vitro study side P [1994]. Assessment of current fibre and of the chemical durability of siliceous fibres. silica exposure in the U.S. refractory ceramic Ann Occup Hyg 31(4B):683–692. fibre manufacturing industry. Ann Occup Hyg Selikoff IJ, Hammond EC, Churg J [1968]. As- 38(Supp 1):739–744. bestos exposure, smoking, and neoplasia. J Am Rice C, Lockey J, Lemasters G, Levin L, Gartside Med Assoc 204(2):104–110. P [1996]. Identification of changes in airborne Smith DM, Ortiz LW, Archuleta RF, Johnson NF fiber concentrations in refractory ceramic fibre [1987]. Long‑term health effects in hamsters manufacture related to process or ventilation modifications. Occup Hyg 3:85–90. and rats exposed chronically to man‑made vitreous fibres. Ann Occup Hyg 31(4B):731–754. Rice C, Lockey J, Lemasters G, Levin L, Staley P, Hansen K [1997]. Estimation of historical Solomon E, Borrow J, Goddard AD [1991]. and current employee exposure to refractory Chromosome aberrations and cancer. Science ceramic fibers during manufacturing and re- 254:1153–1160. lated operations. Appl Occup Environ Hyg J Stacey MH [1988]. Production and characteri- 12(1):54–61. sation of fibers for metal matrix composites. Rice CH, Levin LS, Borton EK, Lockey JE, Hil- Mater Sci Technol 4(3):227–230. bert TJ, LeMasters GK [2005]. Exposures to re- Stanton MF, Layard M, Tegeris A, Miller E, May M, fractory ceramic fibers in manufacturing and Kent E [1977]. Carcinogenicity of fibrous glass: related operations: A 10-year update. J Occup pleural response in the rat in relation to fiber di- Environ Hyg 2:462–473. mension. J Natl Cancer Inst 58(3):587–603. Rödelsperger K, Woitowitz HJ [1995]. Air- borne fibre concentrations and lung burden Stanton MF, Layard M, Tegeris A, Miller E, May compared to the tumour response in rats and M, Morgan E, Smith A [1981]. Relation of par- humans exposed to asbestos. Ann Occup Hyg ticle dimension of carcinogenicity in amphi- 39(5):715–725. bole asbestos and other fibrous minerals. J Natl Cancer Inst 67:965–975. Rogers A, Yeung P, Berry G, Conaty G, Apthorpe L [1997]. Exposure and respiratory health of Sweeney J, Gilgrist D [1998]. Exposures to workers in the Australian refractory ceramic respirable silica during relining of furnaces fibre production industry. Ann Occup Hyg for molten metals. Appl Occup Environ Hyg 41(Suppl 1):261–266. 13(7):508–510.

Refractory Ceramic Fibers 143 References

Thornton CC, Lehman CB, Wood EW [1984]. Vallyathan V, Shi X, Castranova V [1998]. Reac- Flexible blanket noise control insulation field tive oxygen species: their relation to pneumo- test results and evaluation. In: Maling GC, ed. coniosis and carcinogenesis. Environ Health Inter noise 84, Proceedings of the International Perspect 106(Suppl 5):1151–1155. Conference on Noise Control Engineering. van den Bergen EA, Rocchi PSJ, Boogaard PJ Honolulu, HI: Noise Control Foundation, [1994]. Ceramic fibers and other respiratory pp. 405–408. hazards during the renewal of the refractory TIMA [1993]. Man-made vitreous fibers: no- lining in a large industrial furnace. Appl Oc- menclature, chemical and physical properties. cup Environ Hyg 9(1):32–35. Nomenclature Committee of Thermal Insula- tion Manufacturers Association, Inc., 2nd rev., Venturin DE [1998]. Unit operational code of March 1. practice and engineering controls guidebook: band saws. Amherst, NY: Unifax Inc. prepared Timbrell V [1965]. Human exposure to as- for Refractory Ceramic Fibers Coalition. bestos: dust controls and standards. The inhalation of fibrous dusts. Ann NY Acad Sci Vincent JH [1985]. On the practical signifi- 132(1):255–273. cance of electrostatic lung deposition of iso- metric and fibrous aerosols. J Aerosol Sci Timbrell V [1982]. Deposition and retention 16(6):511–519. of fibres in the human lung. Ann Occup Hyg 26(1–4):347–369. Volkwein JC, Page SJ, Thimons ED [1982]. Canopy-air curtain dust reductions on a gath- Timbrell V [1989]. Review of the significance ering-arm loader. Pittsburgh, PA: U.S. Depart- of fibre size in fibre-related lung disease: a cen- ment of the Interior, Bureau of Mines, Report trifuge cell for preparing accurate microscope- of Investigations 8603. evaluation specimens from slurries used in inoculation studies. Ann Occup Hyg 33(4):483–505. Volkwein JC, Engle MR, Raether TD [1988]. Dust control with clean air from and over- Tran CL, Jones AD, Cullen RT, Donaldson K head air supply island (OASIS). Appl Ind Hyg [1997]. Overloading of clearance of particles and 3(8):236–239. fibres. Ann Occup Hyg 41(Suppl 1):237–243. Treadwell MD, Mossman BT, Barchowsky A Vu VT [1988]. Health hazard assessment of [1996]. Increased neutrophil adherence to en- nonasbestos fibers. Washington, DC: U.S. En- dothelial cells exposed to asbestos. Toxicol Appl vironmental Protection Agency, final draft. Pharmacol 139:62–70. Vu V, Barrett JC, Roycroft J, Schuman L, Trethowan WN, Burge PS, Rossiter CE, Har- Dankovic D, Baron P, Martonen T, Pepelko W, rington JM, Calvert IA [1995]. Study of the Lai D [1996]. Workshop report: chronic inha- respiratory health of employees in seven Eu- lation toxicity and carcinogenicity testing of ropean plants that manufacture ceramic fibres. respirable fibrous particles. Reg Toxicol Pharm Occup Environ Med 52(2):97–104. 24:202–212. Turim J, Brown RC [2003]. A dose-response Wagner GR [1996]. Screening and surveillance model for refractory ceramic fibers. Inhal Toxi- of workers exposed to mineral dust. Geneva, col 15:1103–1118. Switzerland: World Health Organization.

144 Refractory Ceramic Fibers References

Wagner JC, Berry G, Timbrell V [1973]. Me- Castranova V [1999]. Critical role of glass fiber sotheliomata in rats after inoculation with as- length in TNF-α production and transcription bestos and other minerals. Br J Cancer 28:173– factor activation in macrophages. Am J Physiol 185. 276: L426–434.

Walker AM, Maxim LD, Utell M [2002]. Risk Yegles J, Janson X, Dong HY, Renier A, Jaurand analysis for mortality from respiratory tumors M-C [1995]. Role of fibre characteristics on cy- in a cohort of refractory ceramic fiber workers. totoxicity and induction of anaphas/telophase Regul Toxicol Pharmacol 35: 95–104. aberrations in rat pleural mesothelial cells in Wang Q, Han C, Wu W, Wang H, Liu S, Ko- vitro: correlations with in vivo animal findings. hyama N [1999]. Biological effects of man- Carcinogenesis 16(11):2751–2758. made mineral fibers (I): reactive oxygen species production and calcium homeostasis in Yu CP, Oberdörster G [2000]. Dose response alveolar macrophages. Ind Health 37:62–67. and human cancer and non-cancer risk assessment of inhaled refractory ceramic fibers Warheit DB [1994]. A review of some bio- (RCF). physical factors and their potential roles in the development of fiber toxicity. Regul Toxicol Yu CP, Asgharian B, Yen BM [1986]. Impaction Pharmacol 20: S113–S120. and sedimentation deposition of fibers in air- WHO/EURO Technical Committee for Mon- ways. Am Ind Hyg Assoc J 47(2):72–77. itoring and Evaluating Airborne MMMF Yu CP, Zhang L, Oberdörster G, Mast RW, Glass [1985]. The WHO/EURO man-made mineralf fiber reference scheme. Scand J Work Environ LR, Utell MJ [1994]. Clearance of refractory Health 11:123–129. ceramic fibers (RCF) from the rat lung: development of a model. Environ Res 65:243–253. Wright A, Cowie H, Gormley IP, Davis JMG [1986]. The in vitro cytotoxicity of asbestos fi- Yu CP, Ding YJ, Zhang L, Oberdörster G, Mast

bers: I. P388D1 cells. Am J Ind Med 9:371–384. RW, Maxim LD, Utell MJ [1996]. A clearance model of refractory ceramic fibers (RCF) in Yamato H, Hori H, Tanaka I, Higashi T, Morim- the rat lung including fiber dissolution and oto Y, Kido M [1994]. Retention and clearance of inhaled ceramic fibres in rat lungs and de- breakage. J Aerosol Sci 27(1):151–160. velopment of a dissolution model. Occup En- Yu CP, Ding YJ, Zhang L, Oberdörster G, Mast viron Med 51:275–280. RW, Maxim LD, Utell MJ [1997]. Retention Ye J, Shi X, Jones W, Rojanasakul Y, Cheng N, modeling of refractory ceramic fibers (RCF) in Schwegler-Berry D, Baron P, Deye GJ, Li C, humans. Regul Toxicol Pharmacol 25:18–25.

Refractory Ceramic Fibers 145

APPENDIX A

Air Sampling Methods*

*Reprinted from NIOSH Manual of Analytical Methods (NMAM), Fourth Edition, 8/15/94.

Refractory Ceramic Fibers 147 Appendix A n Air Sampling Methods

148 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 149 Appendix A n Air Sampling Methods

150 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 151 Appendix A n Air Sampling Methods

152 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 153 Appendix A n Air Sampling Methods

154 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 155 Appendix A n Air Sampling Methods

156 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 157 Appendix A n Air Sampling Methods

158 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 159 Appendix A n Air Sampling Methods

160 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 161 Appendix A n Air Sampling Methods

162 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 163 Appendix A n Air Sampling Methods

164 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 165 Appendix A n Air Sampling Methods

166 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 167 Appendix A n Air Sampling Methods

168 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 169 Appendix A n Air Sampling Methods

170 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 171 Appendix A n Air Sampling Methods

172 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 173 Appendix A n Air Sampling Methods

174 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 175 Appendix A n Air Sampling Methods

176 Refractory Ceramic Fibers Appendix A n Air Sampling Methods

Refractory Ceramic Fibers 177 Appendix A n Air Sampling Methods

178 Refractory Ceramic Fibers Appendix B

Functional Job Categories for RCF Workers

Table B–1 . Functional job categories for RCF workers

Functional category Definition General examples Additional comments

Fiber manufac- The production or manu- Raw materials, furnace man, None to date turing facture of RCF bulk furnace operator, or assistant or blanket, except in a furnace operator supervisory capacity. Production worker or relief Includes all job functions on the production Blanket line line, from mixing the raw Working leader ingredients to packaging Needler the finished product (bulk Slit/cut/pack or blanket) at the end of the line. Line utility Utility operator Chopper operator End of line, bagging of bulk RCF End of line trimming, rolling, and packaging of RCF blanket

Finishing Cutting or machining RCF Operating die stamp on RCF Working in an area where materials after fiber blanket or paper except for finishing is taking place but manufacture. Hand or automotive applications not personally working with power tools may be used RCFs unless in a supervisory in finishing operations. Sawing, slotting, trimming, capacity or in other auxiliary or filing casting tips or riser operations. sleeves

Cutting blanket for duct wrap

(Continued)

Adapted from Maxim et al. 1997.

Refractory Ceramic Fibers 179 Appendix B n Functional Job Categories for RCF Workers

Table B–1 (Continued) . Functional job categories for RCF workers

Functional category Definition General examples Additional comments

Finishing Cutting or trimming RCF board EXAMPLE: Unloading dry (Continued) or other vacuum-formed RCF forms from the drying material capacity oven and taking them to Sanding RCF board or other the finishing area for final vacuum-formed RCF material shaping, or packaging shapes immediately after finishing Loading sander would be considered finish- Off-line cutting and tandem ing. However, unloading dry rerolling and/or repackaging forms from an oven and tak- of RCF blanket ing them to be packaged, or Cutting or trimming RCF mod- packaging shapes that come ules for use in appliances directly from the drying oven would be considered Milling or routing RCF board auxiliary operations. or other vacuum-formed RCF material Off-site cutting of batten strips from RCF blanket

Installation Building or manufactur- Installing hardware or modules Working inside furnace during industrial furnaces or On-site cutting (trimming) ing the installation of RCF boilers, refinery or petro- modules to fit materials, even though not chemical plant equip- working directly with that ment, kilns, foundries, Caulking and filling gaps material (e.g., a plumber or electric power generators, Wrapping molds with RCF electrician working inside a and industrial incinera- Spraying or pumping RCF cast- furnace during an installa- tors at end user locations. able material inside furnace tion) Includes furnace mainte- Cutting and installing laid-in nance. Does not include blanket factory manufacture of industrial furnace components.

Removal Removal of after-service Unwrapping and knocking out Working inside furnace dur- RCF material from an in- molds ing the removal of RCF dustrial furnace, etc., that Furnace disassembly materials, even though not has completed its eco- working directly with that nomic life. Includes the Furnace maintenance material (e.g., a plumber or removal of RCF material Cleanup and disposal of re- electrician working inside a during furnace mainte- moved material furnace during a removal) nance.

Assembly Combining or assembling Laminating operations RCF material with other Cutting material for modules material (RCF or other), except automotive appli- Encapsulating RCF blanket cations. Includes factory Unpacking blanket and loading assembly of industrial into module folder furnace components. (Continued)

180 Refractory Ceramic Fibers Appendix B n Functional Job Categories for RCF Workers

Table B–1 (Continued) . Functional job categories for RCF workers

Functional category Definition General examples Additional comments

Assembly Installing bands around modules operations Packaging modules at end of line (Continued) Trimming modules and installing hardware Assembling appliances Off-site assembly of industrial furnace components (original equipment manufacture) Changing RCF gaskets, etc. in appliances Cutting and assembling material for sound-proofing exhaust ducts Sewing RCF material Stapling RCF material Ball milling or grinding RCF material Mixing RCF putties, compounds, or castables

Mixing/form- Wet end production of Forming RCF board or shapes Premixing dry materials before ing vacuum-cast shapes, Weighing, batching, or mixing adding to mix tank board, and felt materials to be formed Placing wet parts on conveyor Operating mixing machine Felting

Auxiliary op- Jobs in which workers are Moving RCF-wrapped molds erations passively exposed to RCFs into and out of furnace while performing their Warehouse duties, including normal duties and whose dock work, loading trucks, exposures are not likely to moving materials parallel those of workers working directly with RCF Supervising materials. Includes certain Driving forklift jobs in which RCFs may Making cartons to package RCFs be handled but with small at end of line probability of significant Quality control inspection exposures (e.g., warehouse worker or person Packaging dry parts unloading completed Maintaining or repairing equip- parts for packaging). ment except furnaces

(Continued)

Refractory Ceramic Fibers 181 Appendix B n Functional Job Categories for RCF Workers

Table B–1 (Continued) . Functional job categories for RCF workers

Functional category Definition General examples Additional comments

Auxiliary Cleaning furnaces or plant areas operations where RCFs are used (Continued) Removing vacuum-formed parts from oven and/or packaging them (no finishing)

Other (not All duties performed in the Diecutting parts for automotive Wrapping RCF blanket around elsewhere production of RCF paper, airbag filters, gaskets, mufflers, a hot weld so the weld may classisfied) textiles, and automotive or catalytic converters cool without stress between components or other in- the hot seam or joint and the dustry sectors not covered Wrapping substrate for catalytic cooler surrounding metal in any of the foregoing converter (not elsewhere classified) categories. Also, expo- Operating former to make rov- sures that cannot reason- ing ably be included in the categories listed above Operating tape loom (i.e., not elsewhere classified). Industrial hygienist Operating carding machine personnel should explain tasks and industry sectors Papermaking as fully as possible for observations in this category.

182 Refractory Ceramic Fibers Appendix C

Cellular and Molecular Effects of RCFs (In Vitro Studies)

The cellular and molecular effects of RCF ex- human pulmonary cells. The human alveolar posures have been studied with two different macrophage has a volume several times greater objectives. One purpose of these in vitro stud- than that of the rat alveolar macrophage [Krom- ies is to provide a quicker, less expensive, and back et al. 1997]. Macrophage size and volume more controlled alternative to animal toxic- may affect (1) the size range of fibers that can ity testing. These experiments, which strive to be phagocytized, dissolved, and cleared by the act as screening tests or alternatives to animal lungs and (2) the resulting pathogenicity of the studies, are best interpreted by comparing their fiber. Even the use of a human lung cell line does results with those of in vivo experiments. The not guarantee that in vitro results will be directly second objective of in vitro studies is to provide applicable to the intact human response. The data that may help to explain the pathogenesis in vivo integration of stimuli from the nervous, and mechanisms of action of RCFs at the cel- hormonal, and cardiovascular systems cannot be lular and molecular levels. These cytotoxicity reproduced in vitro. and genotoxicity studies are best interpreted by comparing the effects of RCFs with those of Another point to consider when reviewing these other SVFs and asbestos fibers. In vitro studies data is the number and definitions of variables serve as an important complement to animal used in different studies. Variables include dif- studies and provide important tools for study- ferences in fiber type, fiber length, fiber dose, ing the molecular mechanisms of fibers. It is not cell type, and length of exposure tested, among yet possible to use these data in the derivation others. Disparate results between studies make of an REL. strong conclusions from in vitro studies difficult. At the same time, these studies may pro- Drawing strong conclusions relevant to human vide important data regarding the mechanism health based on these in vitro studies is impossi- of action of RCFs that would not be obtainable ble. One point to consider when reviewing these in other testing venues. data is the relevance of the cell types studied. Many studies to date have examined the effects of RCFs may exert their effects on pulmonary tar- RCFs on rodent cell lines. The cytotoxic effects of get cells via direct or indirect mechanisms. Di- RCFs may vary with cell size, volume, and lineage. rect mechanisms are the resultant effects when Effects observed in the cells from organs other fibers come in direct physical contact with cells. than the lung or effects in species other than the Direct cytotoxic effects of RCFs include effects human may not be similar to those elicited with on cell viability, responses, and proliferation.

Refractory Ceramic Fibers 183 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

Indirect cellular effects of RCFs involve the in- also examined and summarized. Table C–1 de- teraction of fibers with inflammatory cells that scribes RCF cytotoxicity studies involving their may be activated to produce inflammatory direct effects on cells. Table C–2 describes RCF mediators. These mediators may affect target cytotoxicity studies involving the release of cells directly or may attract other cells that act mediators. Table C–3 summarizes RCF geno- on target cells. An inflammatory cell type often toxic studies. used in RCF in vitro studies is the pulmonary macrophage. Pulmonary macrophages are the first line of defense against inhaled material C.1 Direct Cytotoxic Effects that deposits in the alveoli, and among func- of RCFs tions, they attempt to phagocytize particles deposited in the lung. Effects of RCF exposure on RCFs may have a direct cytotoxic effect on tar- macrophages and other inflammatory cells are get cells. Measurements of cell viability and cell assessed by the measurement of inflammatory proliferation are both indications of cytotoxic mediator release in vitro. effects. Cell viability can be assessed through the detection of enzymes released by cells or Three important groups of inflammatory me- dyes taken up by cells that indicate altered cell diators are cytokines, ROS, and lipid media- membrane integrity or permeability. Measure- tors (prostaglandins and leukotrienes). Some ment of cytoplasmic LDH and trypan blue ex- of the cytokines that have been implicated in clusion are two methods used to assess cell via- the inflammatory process include TNF and in- bility. LDH is a cytoplasmic enzyme; its release terleukins (ILs). TNF and many ILs stimulate indicates plasma membrane damage. Trypan the deposition of fibroblast collagen, an initial blue is a dye that can only penetrate damaged step in fibrosis, and prostaglandins (PG)s in- cell membranes. β-glucuronidase is a lysosom- hibit these effects. ROS include hydroxyl radi- al enzyme, it assesses lysosomal permeability cals, hydrogen peroxide, and superoxide anion and membrane viability. It may also be released radicals. Oxidative stress occurs when the ROS when alveolar macrophages are activated by level in a cell exceeds its antioxidant level. Oxi- frustrated phagocytosis. The cytotoxic effects dative stress may result in damage to deoxyribo of RCFs on rat pleural mesothelial cells, por- nucelic acid (DNA), lipids, and proteins. cine aortic endothelial cells, human-hamster

Either direct or indirect effects of RCFs may hybrid (AL) cells, human macrophages, macro- result in genotoxic effects on pulmonary target phage-like P388D1 cells, and human alveolar cells. Changes in the genetic material may be epithelial cells are summarized in Table C–1 important in tumor development [Solomon and C–2 and in the text below. et al. 1991]. Genotoxic effects may be assessed Luoto et al. [1997] evaluated the effects of through the analysis of chromosome changes RCFs, quartz, and several MMVFs on LDH or alterations in gene expression following ex- levels in rat alveolar macrophages and hemo- posure to RCFs. lysis in sheep erythrocytes. RCF1, RCF2, RCF3, The following summary of RCF in vitro stud- and RCF4 at 1.0 mg/ml induced a lower release ies examines their direct effects on cell prolif- of LDH (less than 20% of control) from rat eration and viability and indirect effects via alveolar macrophages compared with quartz release of TNF, ROS, and other inflammatory (approximately 40% of control) [Luoto et al. mediators. The genotoxic effects of RCFs are 1997]. RCF1 stimulated the lowest amount of

184 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

LDH release (less than 10% of control), lower The effect of several fibers on the viability of

even than TiO2 (approximately 15% of con- rat pleural mesothelial cells was investigated trol). RCF1, RCF2, RCF3, RCF4, MMVF10, [Yegles et al. 1995]. On a per weight basis, MMVF11, MMVF21, and MMVF22 at 0.5, the rank order of cytotoxicity was National 2.5, and 5.0 mg/ml induced a dose-depen- Institute for Environmental Health Sciences dent increase in sheep erythrocyte hemolysis. (NIEHS) chrysotile, RCF3, MMVF10 and RCF1 and RCF3 induced slightly more hemo- RCF1, Calidria chrysotile, RCF4, and all others. lysis than other MMVFs. The hemolytic activ- Based on the total number of fibers, the rank

ity of MMVFs was similar to that of TiO2, and order of cytotoxicity was RCF3, MMVF10, much less than that of quartz. RCF1, RCF4, MMVF11, NIEHS chrysotile, amosite, and all others. Cytotoxicity was de- At doses of 100, 300, and 1,000 µg/ml RCFs pendent on fiber dimensions as the longest (unspecified type), an increased release of LDH (RCF3, MMVF10, RCF1, MMVF11) or thick- was induced from rat macrophages [Leikauf est (RCF4, RCF1, MMVF11, RCF3) fibers were et al. 1995]. At equivalent gravimetric doses of the most cytotoxic. 1,000 μg/ml, the effects of RCFs were much less than those of silica. Ceramic fibers (unspecified RCF1, RCF2, RCF3, and RCF4 were found to type) at 50 μg/ml induced no difference in LDH inhibit the proliferation and colony-forming levels compared with negative controls in rat al- efficiency of Chinese hamster ovary cells in veolar macrophages [Fujino et al. 1995]. Chrys- vitro [Hart et al. 1992]. The inhibition was otile, crocidolite, amosite, and anthophyllite as- concentration-dependent. RCF4 was least bestos all induced significant increases in LDH cytotoxic, RCF2 was intermediate, and RCF1 and β-glucuronidase levels. Ceramic fibers also and RCF3 were the most cytotoxic. A correla- induced a significant increase in β-glucuroni- tion existed between average fiber length and dase but much less than that induced by each of toxicity, with the shortest fibers being least

the asbestos fiber types. cytotoxic. LC50s for the RCF ranged from 10 to 30 μg/cm2. In each assay, the RCFs were less In the permanent macrophage-like cell line cytotoxic than those of the positive controls P388D1, an elutriated ceramic fiber (unspeci- 2 of crocidolite (LC50=5 μg/cm ) and chrysotile fied type) at 10 or 50 μg/ml after 24 or 48 hr (LC =1 μg/cm2) asbestos. had no significant effect on cell viability as 50 measured by the trypan blue assay [Wright et At 0 to 80 μg/cm2 RCF1, tremolite, and erion- al. 1986]. The elutriation process used for this ite were significantly less cytotoxic to human- experiment provided mainly respirable fibers. hamster hybrid AL cells than chrysotile as de- All other fibers examined, excluding short- termined by the surviving fraction of colonies fiber amosite, reduced viability. Although the after fiber exposure [Okayasu et al. 1999]. RCF1, specific data on the effect of exposure to fibers crocidolite asbestos, and MMVF10 at 25 μg/cm2 on enzyme release was not presented, an as- induced focal necrosis in rat pleural mesothe- sociation between decreasing cell viability and lial cells after 24 hr that became a more obvious increasing loss of intracellular glucosaminidase necrosis by 72 hr [Janssen et al. 1994]. At 72 hr, and LDH was reported under most conditions the qualitative effects of 25 μg/cm2 RCF1 were investigated. Cytotoxicity was correlated with comparable to those of 5 μg/cm2 crocidolite as- fiber lengths greater than 8 µm when all fiber bestos. In contrast, minimal necrosis was seen types were combined. at 25 µg/cm2 crocidolite asbestos, RCF2, and

Refractory Ceramic Fibers 185 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

MMVF10 fibers in hamster tracheal epithelial These studies suggest that RCFs may have some cells at 24 hr; no necrosis was present at 72 hr. similar direct cytotoxic effects to asbestos. They are capable of inducing enzyme release and cell RCF1, RCF2, RCF3, and RCF4 as well as as- hemolysis. They may decrease cell viability and bestos and other fibers had a dose-dependent inhibit proliferation. In most studies, the ef- effect on cytotoxicity, as measured by cell de- fects of RCFs are much less pronounced than tachment, in the human alveolar epithelial cell the effects of asbestos at similar gravimetric line A549 [Cullen et al. 1997]. Cell detachment concentrations. Fiber length was demonstrated is associated with epithelial damage, an impor- to be an important factor in determining the tant step in the inflammatory process. These cell responses in many studies. cells are a primary target of inhaled fibers. When equivalent doses (10, 25, 50, and 100 μg/ ml) were tested with various fibers, all RCFs C.2 Indirect Effects of RCFs: had a less significant effect than both crocido- Effects on Inflammatory lite and amosite asbestos. When the dose data Cells were adjusted for the number of fibers, RCF1, RCF2, and RCF3 were more cytotoxic than In addition to direct effects on target cells, RCF4 and crocidolite. RCFs may have indirect mechanisms of action by acting on inflammatory cells. Inflammatory In an assay determining the ability of fibers to cells, such as pulmonary macrophages, may re- induce an increase in the diameter of human spond to fiber exposure by releasing inflamma- A549 cells, an elutriated ceramic fiber (unspec- tory mediators that initiate the process of pul- ified type) had a midrange of activity when monary inflammation and fibrosis. Cytokines compared with 12 other fibers [Brown et al. and ROS are among the inflammatory media- 1986]. It was more active than most varieties of tors released. Many studies, summarized below amosite tested (but not UICC amosite) but less and in Table C–2, have investigated this link active than the chrysotile fibers. An association between fiber exposure and mediator release was found between increasing length and assay to try to elucidate the mechanism of action of RCFs. Cytokines are a class of proteins that are activity. When these same fiber samples were involved in regulating processes such as cell se- tested for colony inhibition in V79/4 Chinese cretion, proliferation, and differentiation. One hamster lung fibroblasts, the ceramic fiber had of the cytokines most commonly analyzed in even less effect than the TiO control. Analysis 2 RCF cytotoxicity studies is TNF. TNF has been of all fibers upheld the association between in- implicated in silica- and asbestos-induced pul- creasing length and increased activity. In both monary fibrosis [Piguet et al. 1990; Lemaire assays, fiber diameter was not related to activ- and Ouellet 1996]. TNF and many ILs are as- ity in most cases. sociated with collagen deposition (an initial stage of fibrosis), and PGs inhibit these effects. Chrysotile asbestos at 10 μg/cm2 and crocido- Experiments on the effects of RCF exposure on lite asbestos at 5 µg/cm2 altered porcine aor- TNF production in various cell types have had tic endothelial cell morphology and increased differing results. neutrophil adherence [Treadwell et al. 1996]. RCF1 fibers at 10 μg/cm2 did not change cell TNF secretion has been associated with ex- morphology or increase neutrophil binding. posure to asbestos both in vitro and in vivo

186 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

[Perkins et al. 1993]. In vitro incubation of doses, the effect on the levels of all mediators human alveolar macrophages from normal measured was greater after asbestos exposure volunteers with 25 μg/ml chrysotile asbestos than after RCF exposure. resulted in increased levels of TNF secretion. Chrysotile A, chrysotile B, crocidolite, MMVF21, Alveolar macrophages from 6 human subjects RCF1, and silicon carbide at 100 μg/ml caused a occupationally exposed to asbestos for more significantly increased synthesis of TNF mRNA than 10 years secreted increased amounts of after 90 minutes of incubation with rat alveo- the cytokines TNF, IL–6, PGE , and IL–1ß in 2 lar macrophages [Ljungman et al.1994]. After 4 vitro. Five human subjects occupationally ex- hr of incubation, chrysotile A still had a signifi- posed for less than 10 years did not show in- cantly increased TNF mRNA production, and creases in these cytokines. The two exposure all other fibers were at baseline concentrations. groups were matched for age, smoking history, None of the fibers studied increased TNF re- and diagnosis. The increased TNF secretion lease at 90 minutes. However, an increased TNF in both in vitro and chronic in vivo asbestos- bioactivity occurred after 4 hr of incubation exposed conditions suggests its importance in with chrysotile A, chrysotile B, crocidolite, or the response of the lung to fiber exposure, al- MMVF21. RCF1 at 100 μg/ml did not increase though the small exposure group sizes warrant TNF production under these conditions. caution in drawing strong conclusions. Chrysotile asbestos and alumina silicate ce- 6 When equal numbers (8.2  10 ) of various ramic fibers increased in vitro alveolar macro- fiber types, including RCF1, RCF2, RCF3, and phage TNF production in rats exposed to ciga- RCF4, were incubated separately with rat al- rette smoke in vivo and in rats unexposed to veolar macrophages, silicon carbide whiskers, smoke [Morimoto et al. 1993]. Asbestos at 50 amosite, and crocidolite asbestos stimulated and 100 μg/ml induced a significantly greater the highest TNF release [Cullen et al. 1997]. TNF release in rats exposed to cigarette smoke RCF1, RCF2, RCF3, and RCF4 showed no sig- versus unexposed rats. No significant differ- nificant increase in TNF release compared with ences were found between groups at all doses control. of RCF fibers tested (25, 50 and 100 µg/ml). In contrast, ceramic fibers (unspecified type) at RCF exposure, in contrast to chrysotile, did not have a significant synergistic effect with 50 µg/ml (1.72  105 f) significantly increased cigarette smoke exposure. TNF release compared with controls in rat alveolar macrophages [Fujino et al. 1995]. Po- In addition to the cytokines such as TNF, an- tassium octatitanate whisker, chrysotile, and other group of inflammatory mediators that crocidolite asbestos induced the greatest TNF has been studied in vitro are the ROS. These release. Alveolar macrophages exposed to ei- mediators, also referred to as reactive oxygen ther 300 or 1,000 μg/ml RCFs or 1,000 μg/ml metabolites (ROMs) are normally produced asbestos showed a significant increase in TNF during the process of cellular aerobic metabo- production [Leikauf et al. 1995]. At 300 μg/ml lism and in phagocytic cells in response to par- RCFs, a significant elevation occurred in leu- ticle exposure. One molecular effect of asbestos

kotriene B4. At 1,000 µg/ml RCFs, significant exposure has been demonstrated to be the in-

elevations occurred in leukotriene B4 and pros- duction of ROS [Kamp et al. 1992]. Oxidative

taglandin E2. Levels induced at lower doses stress occurs when the ROS level in a cell ex- were not different from controls. At equivalent ceeds the antioxidant level. ROS may result in

Refractory Ceramic Fibers 187 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

damage to DNA, lipids and proteins and have of lung lining fluid [Hill et al. 1996]. At doses been implicated in having a role in carcinogen- >100 μg RCF1, fibers coated with IgG induced esis [Klaunig et al. 1998; Vallyathan et al. 1998]. a significantly increased superoxide anion re- This research has suggested that free radical ac- lease. This supports the premise that lung lin- tivity may be involved in the pathogenesis of ing fluid and other substances that fibers are fiber-induced lung disease. exposed to in vivo may significantly alter the effect of fibers on cells. IgG-coated long fiber The ability of RCFs to induce the release of free amosite asbestos, in spite of a poor binding radicals has been studied in rodent alveolar affinity for IgG, induced a comparable super- macrophages. RF1 and RF2 (Japanese ceramic oxide anion release response to that of coated fibers) at 200 μg/ml induced a significant pro- RCF1. duction of superoxide anion and a significant increase in intracellular free calcium in guinea Brown et al. [1999] investigated the abil- pig alveolar macrophages [Wang et al. 1999]. ity of RCF1, amosite asbestos, silicon carbide, Both superoxide anion and increased intra- MMVF10, Cole 100/475 glass fiber, and RCF4 cellular calcium are associated with oxidative to cause translocation of the transcription fac- stress. Superoxide anions may generate hydro- tor NF-κB to the nucleus in A549 lung epithe- gen peroxide and hydroxyl radical, classified lial cells. RCF1, amosite asbestos, and silicon as ROS or free radicals. RF2 exposure resulted carbide produced a significant dose-dependent in a significant depletion of glutathione. Glu- translocation of NF-κB to the nucleus; the oth- tathione is a cellular antioxidant that protects er fibers tested did not. Equal fiber numbers cells against oxidative stress; depletion of glu- were tested. tathione is associated with oxidative stress. The These cytotoxicity studies indicate that RCFs RFs did not affect hydrogen peroxide produc- may share some aspects of their mechanism of tion. In each test, the effects of chrysotile were action with asbestos. They both affect the pro- significantly greater than those of the RFs. duction of TNF and ROS as well as cell viabil- RCF1, MMVF10, and amosite asbestos at 8.24 × ity and proliferation. The effects of RCFs have 106 f/ml induced a significant depletion of intra- usually been less pronounced than those of as- cellular glutathione in rat alveolar macrophages bestos. Results of in vitro studies are difficult to [Gilmour et al. 1997]. RCF1 had similar effects compare, even within studies of different fiber to amosite asbestos, whereas MMVF10 caused types, because of different study designs, dif- the greatest depletion of glutathione. ferent fiber concentrations and characteristics, and different endpoints. RCF1, RCF2, and RCF3 induced a greater production of ROMs in human polymorpho- nuclear cell cultures than RCF4 and chrysotile C.3 Genotoxic Effects [Luoto et al. 1997]. A dose-dependent produc- of RCFs tion of ROMs was seen in all RCFs and other MMVFs tested from 25 to 500 μg/ml. Quartz In addition to research assessing the cytotox- had a greater effect on ROM production than icity of RCFs, studies have also assessed the all fibers tested. genotoxicity of RCFs. Most genotoxicity assays assess changes in or damage to genetic mate- RCF1 had a high binding capacity for rat im- rial. Methods that have been used to investigate munoglobulin (IgG), a normal component the genotoxicity of fibers include cell-free or in

188 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

vitro cell systems investigating DNA damage, oxyguanosine (dG) in calf thymus DNA and studies of aneuploidy or polyploidy, studies of solution [Leanderson et al. 1989; Leanderson chromosome damage or mutation, gene mu- and Tagesson 1989]. Ceramic and glasswool fi- tation assays, and investigations of cell growth bers had poor hydroxylating capabilities relative regulation [Jaurand 1997]. Several studies, de- to rockwool and slag wool fibers [Leanderson scribed below and in Table C–3, have exam- et al. 1989]. Hydroxyl radical scavengers, such ined the ability of RCFs to produce genotoxic as dimethyl sulfoxide, decreased the hydrox- changes in comparison with asbestos. ylation. Compounds that increase hydroxyl radical formation, such as hydrogen peroxide, Several fibers, including RCF1 and RCF4, were increased hydroxylation. Rockwool in combi- assayed for free radical generating activity us- nation with cigarette smoke condensate caused ing a DNA assay and a salicylate assay [Brown a synergistic increase in 8-OH-dG formation; et al.1998]. The DNA plasmid assay showed ceramic and glasswool fibers did not have syn- only amosite asbestos to have free radical activ- ergistic effects with cigarette smoke [Leander- ity. The salicylate assay showed amosite as well son and Tagesson 1989]. as RCF1 to have free radical activity. Coating the fibers with lung surfactant decreased their RCF1, RCF2, RCF3, and RCF4 induced nucle- hydroxyl radical generation. Differences in ar abnormalities, including micronuclei and RCF1 results in the two assays were proposed polynuclei, in Chinese hamster ovary cells to be a result of increased iron release from [Hart et al. 1992]. Micronuclei may form when RCF1 in the salicylate assay. An iron chelator chromosomes or fragments of chromosomes inhibited the RCF hydroxylation of salicylate. are separated during mitosis. Polynuclei may RCF4 showed no free radical activity. arise when cytokinesis fails after mitosis. The incidence of micronuclei and polynuclei after When equal fiber numbers were compared, exposure to 20 μg/cm2 RCF was from 22% to RCF1, RCF2, RCF3, and RCF4 had minimal 33%. At 5 µg/cm2, chrysotile and crocidolite free- radical-generating activity on plasmid induced nuclear abnormalities of 49% and DNA compared with crocidolite and amosite 28%, respectively. asbestos [Gilmour et al. 1995]. RCFs and other MMVF produced a small but insignifi- Amosite, chrysotile, and crocidolite asbestos, cant amount of DNA damage. This damage and ceramic fibers caused a significant increase was mediated by hydroxyl radicals. No cor- in micronuclei in human amniotic fluid cells relation was found between iron content and [Dopp et al. 1997]. The response was dose- free radical generation. At 9.3 × 105 fibers per dependent with asbestos fiber exposure but assay, amosite produced substantial free radi- not with ceramic fiber exposure. Significant cal damage to plasmid DNA [Gilmour et al. increases in chromosomal breakage and hy- 1997]. Amosite significantly upregulated the perdiploid cells were noted after asbestos and transcription factors AP–1 and NFkB; RCF1 ceramic fiber exposure. had a much smaller effect on AP–1 upregula- tion only. RCF1, RCF3, and RCF4 did not induce anaphase aberrations in rat pleural mesothelial SVFs, including ceramic fibers (unspecified), cells [Yegles et al. 1995]. Of all fibers tested, were reported to form hydroxyl radicals based UICC chrysotile was the most genotoxic on the on the formation of the DNA adduct 8-hy- basis of weight, number of fibers with a length droxydeoxyguanosine (8-OH-dG) from 2-de- >4 µm and number of fibers corresponding

Refractory Ceramic Fibers 189 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

to Stanton=s and Pott=s criteria [Stanton et al. toxicity of fibers must address these factors. 1981; Pott et al. 1987]. Durability is difficult to assess using in vitro studies because of their acute time course. The effect of fibers on the mRNA levels of c- However, in vitro studies provide an opportu- fos and c-jun proto-oncogenes and ornithine nity to study the effects of varying doses and decarboxylase (ODC) in hamster tracheal epi- dimensions of fibers in a quicker, more efficient thelial (HTE) cells and rodent pleural meso- method than animal testing. Although they thelial (RPM) cells were examined [Janssen et provide important information about mecha- al. 1994]. ODC is a rate-limiting enzyme in the nism of action, they do not currently provide synthesis of compounds involved in cell prolif- data that can be extrapolated to occupational eration and tumor promotion, the polyamines. risk assessment. In HTE cells, crocidolite induced a significant dose-dependent increase in levels of c-jun and The association between fiber dimension and ODC mRNA but not c-fos mRNA. RCF1 in- toxicity has been documented and reviewed duced only small nondose-dependent increases [Stanton et al. 1977, 1981; Pott et al. 1987; War- in ODC mRNA levels. In RPM cells, crocidolite heit 1994]. Fiber length has been correlated 2 fibers at 2.5 μg/cm significantly elevated lev- with the cytotoxicity of glass fibers [Blake et al. els of c-fos and c-jun mRNA. RCF1 increased 1998]. Manville code 100 (JM–100) fiber sam- proto-oncogene expression at cytotoxic levels ples of average lengths of 3, 4, 7, 17, and 33 μm 2 of 25 μg/cm ; no significant effect was seen at were assessed for their effects on LDH activ- concentrations ≤5 μg/cm2. ity and rat alveolar macrophage function. The RCF1 fibers were nonmutagenic in the hu- greatest cytotoxicity was reported in the 17 μm and 33 µm samples, indicating that length is man-hamster hybrid cell line AL [Okayasu et al. 1999]. Chrysotile was a significant inducer of an important factor in the toxicity of this fiber. mutations in the same system. Multiple macrophages were observed attached along the length of long fibers. Relatively short These studies demonstrate that RCFs may fibers, <20 μm long, were usually phagocy- share some similar genotoxic mechanisms with tized by one rat alveolar macrophage [Luoto asbestos including induction of free radicals, et al. 1994]. Longer fibers were phagocytized micronuclei, polynuclei, chromosomal break- by two or more macrophages. Incomplete, or age, and hyperdiploid cells. Other studies have frustrated, phagocytosis may play a role in the demonstrated that, using certain methods and increased toxicity of longer fibers. Long fibers doses, RCFs did not induce anaphase aberra- (17 µm average length) were a more potent tions and induced proto-oncogene expression inducer of TNF production and transcrip- only at cytotoxic concentrations. RCFs were tion factor activation than shorter fibers (7 μm nonmutagenic in human-hamster hybrid cells. average length) [Ye et al. 1999]. These studies demonstrate the important role of length in C.4 Discussion of In Vitro fiber toxicity and suggest that the capacity for Studies macrophage phagocytosis may be a critical factor in determining fiber toxicity. The toxicity The toxicity of fibers has been attributable to of individual fibers of the same type of RCFs their dose, dimensions, and durability. Any test may differ according to their size in relation to system that is designed to assess the potential alveolar macrophages.

190 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

Several RCF in vitro studies reported a di- induce necrosis in rat pleural mesothelial cells. rect association between a longer fiber length They also may induce free radicals, micronuclei, and greater cytotoxicity. Hart et al. [1992] re- polynuclei, chromosomal breakage, and hyper- ported the shortest fibers to be the least cyto- diploid cells in vitro. toxic. Brown et al. [1986] reported an association between length and cytotoxic activity but In vitro studies also provide an excellent oppor- not between diameter and activity. Wright et tunity for investigating the pathogenesis of RCF. al. [1986] reported that cytotoxicity was cor- However, comparisons are difficult to make be- related with fibers >8 μm length. Yegles et al. tween in vitro studies because of differences in [1995] reported that the longest and thickest fiber doses, dimensions, preparations, and com- fibers were the most cytotoxic. The four most positions. Important information, such as fiber cytotoxic fibers had GM lengths ≥13 μm and length distribution, is not always determined. GM diameters >0.5 μm. The production of ab- Even when comparable fibers are studied, the normal anaphases and telophases was associat- cell line or conditions under which they are tested with Stanton fibers with a length >8 μm and ed may vary. Much of the research to date has diameter ≤0.25 μm. Hart et al. [1994] reported been done in rodent cell lines and in cells that that cytotoxicity increased with fiber length up are not related to the primary target organ. In to 20 μm. All of these studies demonstrate the vitro studies using human pulmonary cell lines importance of fiber dimensions on cytotoxicity. should provide pathogenesis data most relevant Other studies have not reported the length dis- to human health risk assessment. tribution of fiber samples used. When studies Short-term in vitro studies cannot take into are done with RCFs for which specific lengths account the influence of fiber dissolution and are assessed for cytotoxicity (such as has been fiber compositional changes that may occur done with glass fibers) [Blake et al. 1998], it will over time. In an in vivo exposure, fibers are be possible to determine the strength of the as- continually modified physically, chemically, sociation between RCF fiber length and toxic- and structurally by components of the lung en- ity and determine whether a threshold length vironment. This complex set of conditions is exists above which toxicity increases steeply. difficult to recreate in vitro. Just as it is unlikely In addition to providing data on the correlation that only one factor will be an accurate predic- between fiber length and toxicity, in vitro stud- tor of fiber toxicity, it is much more unlikely ies have provided data on the relative toxicity of that any one in vitro test will be able to predict RCFs compared with asbestos. Uncertainties ex- fiber toxicity. Best results are obtained by tox- ist in the interpretation of these studies because icity assessment in several in vitro tests and in of differences in fiber doses, dimensions, and comparison with in vivo results. In vitro stud- durabilities. RCFs do appear to share some sim- ies provide an excellent opportunity to inves- ilar mechanisms of action with asbestos. (See tigate factors important to fiber toxicity such references in Tables C–1, C–2, and C–3.) They as dose, dimension, surface area, and physico- have similar direct and indirect effects on cells chemical composition. They provide the abil- and alter gene function in similar ways. They ity to obtain information that is an important are capable of inducing enzyme release and supplement to the data of chronic inhalation cell hemolysis. They may decrease cell viability studies. They do not currently provide infor- and inhibit proliferation. They both affect the mation that can be directly applied to human production of tumor necrosis factor and ROS, health risk assessment and the development of and affect cell viability and proliferation. They occupational exposure limits.

Refractory Ceramic Fibers 191 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies) - -

RCF4, (Continued) and RCF3

had less all RCFs RCF2,

crocidolite, Results

RCF1, least were and amosite amosite). (not UICC most amosites Chrysotile effect> Ceramic fiber had no effect. assays. the two effect> ceramic (alling fiber length types) and activity in both assays. asbestos. fiber numbers, MVF11, cytotoxic; than crocido cytotoxic more were and amosite effect than crocidolite and amosite. lite V79/4 assay: equivalent for adjusted When doses, equivalent At in results Ceramic fiber had different increas between found Association A549 assay:

75,

or 25, Dose 50, 10, 5, 25,

0, 50 µg/ml 25 or 100 µg/ml 100 µg/ml or V79/4 cells: 10, A549 cells:

mean:

direct effects on cells on effects direct * reported Not Diameter ( μ m) Diameter 0.79 ± 2.07 1.13 ± 1.90 0.71 ± 2.12 0.94 ± 1.71 0.15 ± 1.53 0.22 ±1.85 done Not 0.57 ± 2.01 0.81 ± 1.76 0.89 ± 1.78 0.25 ± 1.6 0.47 ± 1.39 Geometric 0.84 ± 2.01 0.26 ± 1.75

Length ( μ m) Length reported Not Geometric mean: 23.91 ± 2.39 12.43 ± 2.66 14.99 ± 2.64 6.82 ± 2.00 ± 2.86 3.03 ± 2.57 4.96 ± 2.62 2.88 3.50 ± 2.17 8.73 ± 2.25 done Not 14.21 ± 2.64 15.66 ± 2.76 13.67 ± 2.34 10.42 ± 2.66

In vitro In cytotoxicity of RCFs:

Fiber typeFiber

Table C–1 . Table (unspecified) amosite Long Elutriated (E) ceramic dioxide Titanium Quartz RCF1 RCF2 RCF3 RCF4 Crocidolite MMVF10 MMVF11 MMVF21 MMVF22 C100/475 glass Short fiber amosite fiber amosite E long UICC amosite chrysotileSuperfine carbide 2 Silicon carbide 1 Silicon E factory amosite 104E glass E brucite UICC crocidolite E UICC chrysotile E tremolite E UICC anthophyllite UICC chrysotile E UICC crocidolite

endpoints Cell line and table. Cell diameter fibroblasts inhibition Colony lung hamster V79/4 Chinese alveolar Human A549 cells epithelial cells Cell detachment

Reference [1997] of See at end footnote [1986] et al. Cullen et al. Brown

192 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

2 2

2 RCF2 (Continued) all others control: is least RCF4 and RCF1 magnesium sulfate Results % of not were and erionite : 50 All caused fibers a significant with control; compared increase chrysotile caused the highest release. Crocidolite Chrysotile 5 μ g/cm 1 μ g/cm most cytotoxic; are RCF3 is intermediate; cytotoxic. LDH levels. in colony decrease dependent proliferation. formation and cell 300 μ g/ml—174% RCFs 1,000 μ g/ml—295% RCFs 1,000 μ g/ml—896% Silica 100 μ g/ml—158% RCFs 10–30 μ g/cm RCFs Ceramic fiber, whiskers, control; from different levels. significantly increased -glucuronidase: LDH: β LC cytotoxicity: RCF in increases moderate induced RCFs a concentration- induced RCFs

or Dose 10, 2, 300, 5, 1, 0, 5 μ g/ml 0 or 1,000 μ g/ml 20 μ g/ml 0, 5 μ g/ml 50 µg/ml RCF1–4: Crocidolite: Chrysotile: 100, direct effects on cells on effects direct * Diameter ( μ m) Diameter Geometric mean: 1.92 ± 2.9 ± 2.2 0.54 0.41 ± 1.5 0.44 ± 1.6 0.31 ± 1.5 0.085 ± 1.4 ± 1.5 0.20 0.32 ± 1.8 0.36 ± 2.3 0.21 ± 1.6 Geometric mean: 1.03 ± 0.73 ± 0.82 1.11 1.22 ± 0.98 1.43 ± 0.79 0.21 ± 0.12 0.12 ± 0.07 Median: 0.59 0.28 0.88 0.18 Length ( μ m) Length Geometric mean: 29.5 ± 3.1 12.8 ± 3.0 2.8 ± 2.0 12.0 ± 2.3 4.9 ± 2.1 0.7 ± 1.9 1.3 ± 2.3 2.7 ± 2.5 2.5 ± 3.5 1.4 ± 2.0 Geometric mean: 21.5 ± 16.12 16.7 ± 15.03 24.3 ± 18.82 9.2 ± 7.08 1.8 ± 1.94 1.65 ± 1.83 Median: 4.9 2.5 available Not available Not In vitro In cytotoxicity of RCFs: Fiber typeFiber Ceramic Glass octatitanate Potassium (long) sulfate Magnesium (short) sulfate Magnesium Chrysotile asbestos asbestos Crocidolite Amosite Anthophyllite Erionite RCF1 RCF2 RCF3 RCF4 UICC Crocidolite UICC Chrysotile RCF asbestos Crocidolite Silica dioxide Titanium

. C–1 (Continued) Table

endpoints Cell line and -glucuronidase table. β Cell proliferation formation Colony macrophages LDH ovary cells LDH macrophages Rat alveolar Rat alveolar hamster Chinese

et al. Reference Hart et al. Leikauf of See at end footnote [1995] [1995] [1992] et al. Fujino

Refractory Ceramic Fibers 193 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies) all -

- - fiber

and had

quartz

(Continued) most

2 2 2 2 doses short-fiber

individual glucosamini viability

cytotoxicity;

of both viability.

usually

and reported.

Results excluding were not

LDH degree

decreasing m fibers

reduced

MMVF 22 had the greatest μ in were

fibers,

concentrations; >8

lowest

dusts. hemolytic than allbut fibers other less than quartz. much overall. in he increase Dose-dependent the least LDH induced RCF1 release; LDH release; fiber-induced LDH release the greatest induced Chrysotile ~4 μ g/cm other amosite, the Ceramic increasing fective all and fibers by induced molysis fibers. than other effects dase 42 μ g/cm Erionite 40 μ g/cm Tremolite 35 μ g/cm RCF1 Hemolysis: slightly more were and RCF3 RCF1 LDH: Chrysotile cytotoxic was more assay: blue Trypan Fibers lethal doses: Mean

or Dose g ± SD:

2.5, -10 0.5, 1.0 mg/ml 5.0 mg/ml 112 ± 10 Hemolysis: LDH: 0–400 µg/ml in Fiber number 0–80 μ g/cm 76 ± 6 185 ± 18 109 ± 10 112 ± 14 144 ± 20 169 ± 15 15 ± 0.4 10 19 ± 1

direct effects on cells on effects direct *

mean:

Diameter ( μ m) Diameter reported 1.30 ± 0.72 1.39 ± 0.98 1.37 ± 0.87 1.33 ± 1.02 1.42 ± 0.78 1.12 ± 0.88 1.18 ± 0.64 0.88 ± 0.45 0.13 ± 3.43 0.23 ± 2.74 0.95 ± 2.6 details not had measured less diameters than 3 μ m; 0.12 ± 0.08 Mean: Geometric allAlmost fibers

Length ( μ m) Length 1.41 ± 2.7 1.31 ± 2.9 13.5 ± 2.7 distribulength tions reported 1.78 ± 2.3 21.29 ± 17.42 20.18 ± 19.63 25.66 ± 25.74 10.34 ± 11.59 23.21 ± 15.57 15.65 ± 13.31 26.02 ± 23.10 20.75 ± 20.52 Mean: Geometric mean: fiber Cumulative In vitro In cytotoxicity of RCFs:

Fiber typeFiber

RCF1 RCF2 RCF3 RCF4 MMVF10 MMVF11 MMVF21 MMVF22 UICC chrysotile Tremolite Erionite RCF1 UICC crocidolite UICC amosite UICC chrysotile E UICC chrysotile E UICC amosite fiber amosite E long E tremolite Short fiber amosite E UICC crocidolite fiber E ceramic E UICC anthophyllite

. C–1 (Continued) Table

cells L A

endpoints Cell line and table. concentrations Sheep erythrocytes Hemolysis Human-hamster P388D1 assay blue Trypan Rat alveolar concentrations macrophages LDH colonies hybrid Glucosaminidase LDH Surviving

et al.

Reference et al. Okayasu Wright of See at end footnote [1997] [1999] [1986] et al. Luoto

194 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

calidria and longest RCFs=refractory Results

with cytotoxicity: most cytotoxic. were fibers thickest per weight basis: Cytotoxicity > > MMVF10 RCF1 RCF3 Chrys, MMVF10=RCF1> >MMVF11 > Chrys, RCF4 >all others NIEHS>amosite correlated and diameter Fiber length fibers: Chrys NIEHS > RCF3> > all others > RCF4 of number per total Cytotoxicity

2 75,

50, Dose Sciences; Health Environmental 25, 100 μ g/cm or 50 μ g/ml 10 or 20 μ g/ml 80 μ g/ml 44 ± 2.7 7.4 ± 0.5 22 ± 1.6 152 ± 14 Fiber dose: 12.5, direct effects on cells on effects direct *

Diameter ( μ m) Diameter of Institute NIEHS=National Mean: 0.07 ± 0.09 0.17 ± 0.11 0.05 ± 0.04 1.30 ± 0.80 0.74 ± 0.50 1.30 ± 0.60 0.55 ± 0.50 1.10 ± 0.90 0.05 ± 0.04 0.04 ± 0.03 0.04 ± 0.06 ± 0.08 0.04 ± 0.03 0.04 ± 0.03 0.05 ± 0.31 ± 0.20 0.19 ± 0.12 0.04 ± 0.03

Length ( μ m) Length

Mean: 2.8 ± 3.0 4.7 ± 5.9 2.3 ± 1.8 19.2 ± 15.0 31.8 ± 36.0 8.9 ± 7.2 21.5 ± 16.8 16.7 ± 12.9 1.7 ± 2.2 2.3 ± 1.9 ± 1.6 ± 1.4 2.4 ± 3.1 4.7 ± 5.2 4.2 ± 1.2 2.4 ± 1.8 2.1 ± 3.6 0.8 ± 0.5

In vitro In cytotoxicity of RCFs:

UICC

vitreous MMVF=man-made fiber;

with leached

Fiber typeFiber Calidria NIEHS short Canadian Canadian superfine phosphorylated, phosphorylated

le Cancer. Contre Internationale UICC=Union Quartz RCF1 RCF3 RCF4 MMVF10 MMVF11 Amosite Crocidolite Attapulgite Chrysotile (chrys), E brucite chrysotile SFA dioxide Titanium Chrys, Canadian milled Canadian Chrys, Chrys 443 (Canadian) UICC chrys, Chrys, Chrys, Chrys, Chrys, Chrys 445 (Canadian) acid oxalic

. C–1 (Continued) Table

dehydrogenase; LDH=lactate endpoints Cell line and integrity mesothelial cells Cell viability: Mitochondrial Rat pleural deviation; SD=standard

E=elutriated;

fibers; ceramic Reference *Abbreviations: [1986] (Continued) et al. Yegles [1995] et al. Wright

Refractory Ceramic Fibers 195 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies) - po - and (Continued) crocidolite, Results SiC2, the stimulated amosite long highest TNF production. TNF activitymore than in cultures. control octatitanate causedtassium the highest TNF production. TNF production creased no-fiber controls; above All significantly fibers in SiC1, did not show RCF1–RCF4 fibers 6 Dose 50 μ g/ml (>5 μ m long) 8.2×10

Diameter ( μ m) Diameter 1.13 ± 1.90 0.25 ± 1.6 0.47 ± 1.39 0.20 ± 1.5 Geometric mean: Geometric mean: 0.84 ± 2.01 0.71 ± 2.12 0.94 ± 1.71 0.26 ± 1.75 0.15 ± 1.53 0.22 ±1.85 done Not 0.57 ± 2.01 0.81 ± 1.76 0.89 ± 1.78 0.54 ± 2.2 0.41 ± 1.5 0.44 ± 1.6 0.31 ± 1.5 0.085 ± 1.4 0.32 ± 1.8 0.36 ± 2.3 0.21 ± 1.6 0.79 ± 2.07 1.92 ± 2.9

effects on mediator release mediator on effects *

Length ( μ m) Length 23.91 ± 2.39 12.8 ± 3.0 2.8 ± 2.0 12.0 ± 2.3 4.9 ± 2.1 0.7 ± 1.9 1.3 ± 2.3 2.7 ± 2.5 2.5 ± 3.5 1.4 ± 2.0 12.43 ± 2.66 14.99 ± 2.64 6.82 ± 2.00 14.21 ± 2.64 15.66 ± 2.76 13.67 ± 2.34 3.03 ± 2.86 4.96 ± 2.57 2.88 ± 2.62 3.50 ± 2.17 8.73 ± 2.25 done Not 10.42 ± 2.66 29.5 ± 3.1 Geometric mean: Geometric mean:

In vitro In cytotoxicity of RCFs:

Fiber type Fiber RCF1 RCF2 RCF3 RCF4 amosite Long Crocidolite SiC1 SiC2 MMVF10 MMVF11 MMVF21 MMVF22 Ceramic Glass Amosite Anthophylite Erionite C100/475 glass octatitanate Potassium 104E glass asbestos Crocidolite (short) sulfate Magnesium Chrysotile asbestos (long) sulfate Magnesium Table C–2 . Table

endpoints Cell line and table alveolar Human Rat alveolar epithelial cells TNF macrophages TNF

Reference et al. Cullen et al. Fujino of See at end footnote [1997] [1995]

196 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

in 2 at

did

g/ml than

their

fibers after μ

doses

PGE

increased and and

at effects significant.

increased binding

(Continued)

, asbestos asbestos 1,000 4

exposure

300 release.

not

response or asbestos was fibers

increased

fibers

high elevated to

RCFs;

LFA LTB

a after

RCFs. doses,

concentrations 300 were

LFA anion production. Results 2

IgG; significantly were had g/ml RCF1

RCF1

TNF, greater greatly a

g/ml a for g of

PGE

doses

exposure elevated μ

fibers levels production

1,000 4

equivalent

were to lower concentrations. RCFs; exposure 1,000 after RCFs duced decrease. effects. not. IgG-coated induced superoxide >100 superoxide affinity intracellular glutathione. At had similar and amosite RCF1 LTB IgG-coated TNF Coating RCF1 All significantly fibers lowered MMVF10 caused the greatest

/ml

300, 1 mg/ml asbestos: LFA Dose 8.24×10 MMVF21: 125 μ g– 20 mg/ml 100/475, Code μ g– SiC:125 15.6 μ g– 5 mg/ml 1,000 μ g/ml RCF1, 100,

0.59 0.28 0.88 0.18 effects on mediator release mediator on effects Median: * reported Not reported Not Diameter ( μ m) Diameter

4.9 2.5 NA NA Median: 5,462 9,015 21,742 70,550 164,705 WHO f/ μ g Length ( μ m) Length reported Not distribution also reported. length Percentage

In vitro In cytotoxicity of RCFs:

Fiber typeFiber 2

RCF1 asbestos Amosite MMVF10 MMVF21 RCF1 asbestos LFA SiC (unspecified) RCF Silica TiO Code Manville Johns asbestos Crocidolite 100/475 (glass)

. C–2 (Continued) Table a normal 2 4 endpoints Cell line and macrophages after release withcoating rat immunoglobulin (IgG), of component lining fluid. lung table Glutathione PGE Rat alveolar macrophages anion Superoxide Rat macrophages TNF LTB Rat alveolar

et al. Reference of See at end footnote et al. Hill Leikauf [1996] [1995] [1997] Gilmour et al.

Refractory Ceramic Fibers 197 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies) -

concen (Continued) A. did RCF1 chrysotile B, chrysotile B, MMVF21, and MMVF21 ex RCF3 or A, A, Results increased and SiCwh RCF2, incubation; crocidolite, RCF1, concentrations TNF mRNA 90 minutes; after sure. production. ROM on in higher resulted posure than production ROM chrysotile or expo RCF4 trations had to returned in all 4 hours baseline after chrysotilebut crocidolite, in TNF an increase induced bioactivity 4 hours after of a significant not induce to response dependent production. ROM increase. Chrysotile a dose- All showed fibers Quartz effect had the greatest Chrysotile RCF1,

or 100, 400, Dose 50, 200, 500 μ g/ml 100 μ g/ml 25,

effects on mediator release mediator on effects * Diameter ( μ m) Diameter Mean: Mean: 0.3 ± 0.2 1.30 ± 0.72 reported Not 1.3 ± 0.8 reported Not 1.2 ± 0.7 1.1 ± 0.8 reported Not 1.39 ± 0.98 1.37 ± 0.87 1.33 ± 1.02 1.42 ± 0.78 1.12 ± 0.88 1.18 ± 0.64 0.88 ± 0.45

Length ( μ m) Length Mean: Mean: 9.9 ± 7.8 reported Not 24.6 ± 19.9 21.29 ± 17.42 reported Not 21.4 ± 17.6 22.4 ± 19.0 reported Not 20.18 ± 19.63 25.66 ± 25.74 10.34 ± 11.59 23.21 ± 15.57 15.65 ± 13.31 26.02 ± 23.10 20.75 ± 20.52

In vitro In cytotoxicity of RCFs:

Fiber typeFiber Crocidolite MMVF22 RCF1 SiCwh RCF1 RCF2 RCF3 RCF4 MMVF10 MMVF11 MMVF21 MMVF22 Chrysotile Chrysotile B MMVF21

. C–2 (Continued) Table endpoints Cell line and table. Rat alveolar macrophages TNF macrophages ROM Rat alveolar

Reference et al. Luoto of See at end footnote [1997] [1994] Ljungman et al.

198 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies)

no SC1,

smoke.

TO1, TNF RCFs.

exposed greater in

cells

to dose- ceramic other

increased signifi anion between

ceramic the

greater TO1, smoke- SC1,

production;

had rats

and

higher controls. than in

significantly

production intracellular fibers

RW1

concentration.

resulted RF2, PT1, smoke-exposed

exposed all from

and

those

in controls; and smoke

TNF ROM=reactive Results significantly

difference MG1

GSH peroxide superoxide not

stimulated RF1, RF2,

and stimulated induced significantly chrysotile

in oxide TO1=titanium rats cells

macrophages.

not.

than increased and

rats

production production MG1, exposure WO1

calcium. than

tests,

fibers did

cigarette

all

than increased dependent alveolar TNF fibers to from production rats fiber significant TNF PT1, decreased production. hydrogen RF WO1, exposed cantly free effects Chrysotile Both Chrysotile Chrysotile Chrysotile, Chrysotile Chrysotile, In vitreous MMVF=man-made fiber;

=titanium dioxide; 2 or RF3=refractory mullite; Dose TiO 50, 100 μ g/ml 25, 200 μ g/ml glass; MG1=micro -

RF2=refractory ceramic; effects on mediator release mediator on effects * B4; =leukotriene 0.88 ± 3.10 1.80 ± 2.32 0.24 ± 2.35 0.77 ± 2.53 1.10 ± 2.00 2.40 ± 1.37 0.35 ± 1.51 0.30 ± 1.58 0.14 ± 1.53 1.00 ± 1.72 reported Not 4 Diameter ( μ m) Diameter = 3.1 μ m ameter di aerodynamic factor; necrosis TNF=tumor median Mass LTB

RF1=refractory ceramic; 20.2 ± 2.58 16.5 ± 2.51 3.0 ± 2.22 12.0 ± 2.36 11.0 ± 1.96 11.0 ± 1.75 6.0 ± 2.04 2.1 ± 2.00 10.5 ± 2.03 reported Not 6.4 ± 2.45 Length ( μ m) Length reported Not carbide; SiC=silicon

fiber amosite; LFA=long [1999];

In vitro In cytotoxicity of RCFs: et al. Wang Fiber typeFiber IgG=immunoglobulin; fibers; ceramic RCFs=refractory Chrysotile glass MG1 = micro = titanium oxide TO1 wool = rock RW1 RF1 = refractory ceramic GW1 = glass wool = wollastonite WO1 RF2 = refractory ceramic RF3 = refractory mullite PT1 = potassium titanate carbide SC1 = silicon chrysotileCanadian : material fibrous Japan ceramic silicate Alumina (Japan)

carbide in SC1=silicon WO1=wollastonite.

GW1=glass wool; . C–2 (Continued) Table

endpoints PT1=potassium titanate; Cell line and ; 2 wool; RW1=rock wh=whiskers; macrophages anion Superoxide peroxide Hydrogen GSH free Intracellular macrophages TNF pig alveolar Guinea Rat alveolar calcium

GSH=glutathione; [1999];

E =prostaglandin 2 Reference metabolites; oxygen et al. Wang PGE Abbreviations: * [1999] [1993] et al. Wang et al. Morimoto

Refractory Ceramic Fibers 199 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies) - - - (Continued) and RCF4

RCF3, Results RCF2, activity. in micronuclei. increase withseen not with but asbestos fiber exposure. ceramic fibers. content. radicalhad a minimal free with asbestos effect compared radical had free amosite Only had free and RCF1 Amosite radical activity. hydroxyl surfactant decreased radical activity. RCF1. by droxylation radicals but hydroxyl by ated withwas not associated iron breakage chromosomal duced cells. and hyperdiploid assay:Salicylate All caused fibers a significant was response Dose-dependent in fibers and ceramic Asbestos RCF1, was medi damage DNA RCF Plasmid assay: withCoating the fibers lung hy inhibited chelator iron An

2 6 or or or f/20 5 5 f/ml

5 10 7 assay:

10 10 × 5.0, × × Dose 1.0, 10.0 μ g/cm 9.249 1.2332 fiber numbers: 9.249×10 8.24×10 6.166 * 0.5, at equal Tested Plasmid assay: Salicylate

0.10 0.90 0.25 0.24 Average: reported Not reported Not Diameter ( μ m) Diameter

27.6 35.25 45.27 17.96 67.17 19.32

2.05 1.71 2.24 12.03 Average: Length ( μ m) Length reported Not distribution Size 60.86 64.75 77.36 59.35 85.24 50.00 >10 >20 In vitro genotoxic effects of effects vitro In RCFs genotoxic

Table C–3 . Table

Fiber typeFiber

fiber amosite Long SiC RCF1 RCF4 MMVF10 asbestos Amosite Gypsum RCF1 RCF2 RCF3 RCF4 MMVF10 MMVF11 MMVF21 MMVF22 100/475 glassCode Ceramic (unspecified) asbestos Crocidolite asbestos Rhodesian chrysotile Short fiber amosite fiber amosite Long asbestos Crocidolite asbestos

DNA

endpoints Test system and system Test table. DNA scission DNA assay Salicylate radical Hydroxyl Hyperdiploidy Chromosomal amniotic Human breakage supercoiled Plasmid oX174RF1 Plasmid oX174RF1 DNA Depletion of generation fluid cells Micronuclei

Reference et al. Brown of See at end footnote [1997] [1998] [1995] et al. Dopp Gilmour et al.

200 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies) -

- 33%. 28%.

AP–1 49%.

induced

mRNA 2 was

(Continued)

ODC only. was

20%

in significant

non-dose-

g/cm mRNA. c-jun increased

AP–1 only.

was μ

2 concentrations

mRNA

2.5 and small

crocidolite chrysotile RCF ODC Results

g/cm , , , at induced had a much RCF1 2 increases

2 2

μ c-jun

c-fos and

concentrations

g/cm g/cm g/cm and at induced

cells: μ

cells:

μ μ

5 5 20

c-jun

radical free less DNA nificantly asbestos. than amosite damage At elevated c-fos At concentrations. concentrations. Crocidolite mRNA dependent transcriptionlated factors and NFkB; smaller effect on At Crocidolite dose-dependent of HTE RCF1

RPM RCF1 abnormality incidence: Nuclear significantly upregu Amosite

sig and MMVF10 induced RCF1

/ml

7 fibers: or

5 2 or

* 2, 10,

Dose g/ml 1, 5, other

5 μ g/cm μ 1.25–25 μ g/cm per numbers assay. 9.3×10 8.24×10 ≤ 0, 5 μ g/ml 0 or 20 μ g/ml 0, 5 Equal fiber assay: DNA assay: Iron All Crocidolite: Chrysotile: Asbestos: 1–4: RCFs

0.27 0.08 1.36 1.07 1.05 0.8 0.8 1.03 ± 0.73 1.11 ± 0.82 1.22 ± 0.98 1.43 ± 0.79 0.21 ± 0.12 0.12 ± 0.07 reported Not Diameter ( μ m) Diameter Mean: Mean:

In vitro genotoxic effects of effects vitro In RCFs genotoxic 11.4 19.8 24.0 — 6.0 1.1 — 21.5 ± 16.12 16.7 ± 15.03 24.3 ± 18.82 9.2 ± 7.08 1.8 ± 1.94 1.65 ± 1.83 Length ( μ m) Length reported Not Mean: Mean:

Fiber typeFiber

. C–3 (Continued) Table MMVF10 RCF1 RCF1 RCF2 RCF3 RCF4 UICC Crocidolite Crocidolite Chrysotile MMVF10 RCF1 beads Polystyrene Riebeckite Erionite UICC Chrysotile asbestos Amosite

- -

tran-

c-fos, c-fos, and ODC and ODC endpoints Test system and system Test c-jun, tions of tions of c-jun, of Activation table. DNA Depletion of induction induction Polynuclei DNA supercoiled Rat alveolar RPM cells concentra mRNA hamster Chinese macrophages ovary cells Micronuclei epithelial (HTE) tracheal Hamster cells concentra mRNA scription factors Plasmid oX174RF1

Reference Gilmour et al. et al. Janssen of See at end footnote [1992] [1994] Hart et al. [1997]

Refractory Ceramic Fibers 201 Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies) - - did on and

(Continued) effect fibers

dG. condensate wool

Results glass synergistic

a smoke

based examined fibers or

have

not cigarette hydroxylation caused less DNA wool. than rock had a synergistic condensate hydroxylation. effect on hydroxylation. nonmutagenic. of hydroxylation capac had poor hydroxylating and wools rock to ity relative slag wools. of genic fiber concentration. on Ceramic Ceramic and glass fibers wool smoke and cigarette wool Rock be to was determined RCF1 Chrysotile was the most muta All caused fibers significant fibers and ceramic Glass wool

2 O 2 2

* Dose 1.0 ml PBS with 0.5 mg or DNA 300 μ l and/or or smoke-PBS 100 μ M H DNA 0.5 mM dG in 1.0 ml PBS with 0.5 mg 10 mg fiber and 10 mg fiber 0–80 μ g/cm

mean:

0.13 ± 3.43 0.23 ± 2.74 0.95 ± 2.6 0.12 ± 0.08 Diameter ( μ m) Diameter reported Not reported Not Geometric

/g): 2

area

In vitro genotoxic effects of effects vitro In RCFs genotoxic Length ( μ m) Length 0.95 0.91 1.10 1.30 0.36 0.60 0.73 1.01 1.28 1.16 1.18 1.14 1.30 1.06 0.90 0.62 1.41 ± 2.7 1.31 ± 2.9 13.5 ± 2.7 1.78 ± 2.3 Geometric mean: reported Not Surface

Fiber typeFiber . C–3 (Continued) Table Ceramic Glass wool Fiber 1 = ceramic wool Rock wool Fiber 5 = rock wool Fiber 6 = rock wool Fiber 7 = rock wool Fiber 8 = rock wool Fiber 9 = rock wool Fiber 10 = rock wool Fiber 13 = rock wool Fiber 16 = rock Fiber 2 = glass wool Fiber 3 = ceramic wool Fiber 4 = rock wool Fiber 11 = rock Fiber 12 = slag wool wool Fiber 14 = rock Fiber 15 = slag wool Tremolite Erionite RCF1 source: European source: European UICC chrysotile

cells L A endpoints 2-dG to DNA thymus DNA thymus Test system and system Test 8–OH–dG of Hydroxylation Hydroxylation of 8–OH–dG dG to of Hydroxylation 8–OH–dG dG to dG solution table. Calf Calf hybrid Human-hamster assay Mutation

[1989] [1989] Reference et Leanderson et al. Okayasu of See at end footnote al. al. [1999] et Leanderson

202 Refractory Ceramic Fibers Appendix C n Cellular and Molecular Effects of RCFs (In Vitro Studies) - meso

fibers

pleural >4 fibers fiber;

Results vitreous

RPM=rat

of and number of number fibers;

and Stanton’s to corresponding criteria. Pott’s aberrations. anaphase induce the basis of on genotoxic weight, μ m long, Ceramic and glass did not fibers UICC chrysotile was the most ceramic

2 MMVF=man-made

75, 50, RNA;

Dose 25, RCFs=refractory

100 μ g/cm or 12.5, saline;

mRNA=messenger

Diameter ( μ m) Diameter 0.04 ± 0.03 0.07 ± 0.09 0.17 ± 0.11 0.05 ± 0.04 1.30 ± 0.80 0.74 ± 0.50 1.30 ± 0.60 0.55 ± 0.50 1.10 ± 0.90 0.05 ± 0.04 0.04 ± 0.03 0.04 ± 0.06 ± 0.08 0.04 ± 0.03 0.05 ± 0.31 ± 0.20 0.19 ± 0.12 0.04 ± 0.03 PBS=phosphate-buffered

In vitro genotoxic effects of effects vitro In RCFs* genotoxic Length ( μ m) Length decarboxylase;

OH–dG=hydroxydeoxy-guanosine;

19.2 ± 15.0 4.7 ± 5.2 4.7 ± 5.9 2.3 ± 1.8 2.8 ± 3.0 31.8 ± 36.0 8.9 ± 7.2 21.5 ± 16.8 16.7 ± 12.9 1.7 ± 2.2 2.3 ± 1.9 ± 1.6 ± 1.4 2.4 ± 3.1 4.2 ± 1.2 2.4 ± 1.8 2.1 ± 3.6 0.8 ± 0.5

epithelial;

ODC=ornithine leached

tracheal

phosphorylated NIEHS short Canadian superfine Calidria Fiber typeFiber

Sciences;

. C–3 (Continued) Table RCF1 RCF3 RCF4 MMVF10 MMVF11 Amosite Crocidolite Attapulgite Canadian milled Canadian UICC chrys, UICC chrysotile Canadian Chrys, Chrys,phosphorylated acid with oxalic Chrys, Chrys 445 (Canadian) Chrys 443 (Canadian) Chrys, Chrys, Chrys, Cancer. Health

HTE=hamster

- Contre f=fibers;

Environmental

for

endpoints Test system and system Test aberrations thelial cells Internationale

Anaphase/telophase Rat pleural meso Institute

dG=deoxyguanosine;

UICC=Union

Reference thelial; NIEHS=National et al. Yegles *Abbreviations: [1995]

Refractory Ceramic Fibers 203