Recent Results 120 in Cancer Research Managing Editors Ch. Herfarth, Heidelberg· H.-J. Senn, St. Gallen Associate Editors M. Baum, London· V. Diehl, Koln E. Grundmann, Munster· F. Gutzwiller, Zurich W. Hitzig, Zurich· M.F. Rajewsky, Essen M. Wannenmacher, Heidelberg Founding Editor P. Rentchnick, Geneva Recent Results in Cancer Research Volume 110 P. Schlag, P. Hohenberger, U. Metzger (Eds.): Combined Modality Therapy of Gastrointestinal Tract Cancer 1988. 105 figures, 122 tables. XVII, 301. ISBN 3-540-18610-7 Volume 111 H. Scheurlen, R. Kay, M. Baum (Eds.): Cancer Clinical Trials: A Critical Appraisal 1988. 37 figures, 53 tables. IX, 272. ISBN 3-540-19098-8 Volume 112 L. Schmid, H.-J. Senn (Eds.): AIDS-Related Neoplasias 1988.23 figures, 35 tables. IX, 97. ISBN 3-540-19227-1 Volume 113 U. Eppenberger, A. Goldhirsch (Eds.): Endocrine Therapy and Growth Regulation of Breast Cancer 1989. 26 figures, 17 tables. IX, 92. ISBN 3-540-50456-7 Volume 114 P. Boyle, C.S. Muir, E. Grundmann (Eds.): Cancer Mapping 1989. 97 figures, 64 tables. IX, 277. ISBN 3-540-50490-7 Volume 115 H.-J. Senn, A. Goldhirsch, R.D. Gelber, B. Osterwalder (Eds.): Adjuvant Therapy of Primary Breast Cancer 1989.64 figures, 94 tables. XVI, 296. ISBN 3-540-1881O-X Volume 116 K.W. Brunner, H. Fleisch, H.-J. Senn (Eds.): Bisphosphonates and Tumor Osteolysis 1989. 22 figures, 6 tables. IX, 78. ISBN 3-540-50560-1 Volume 117 V. Diehl, M. Pfreundschuh, M. Loeffler (Eds.): New Aspects in the Diagnosis and Treatment of Hodgkin's Disease 1989. 71 figures, 91 tables. XIV, 283. ISBN 3-540-51124-5 Volume 118 L. Beck, E. Grundmann, R. Ackermann, H.-D. Roher (Eds.): Hormone-Related Malignant Tumors 1990. 83 figures, 91 tables. XI, 269. ISBN 3-540-51258-6 Volume 119 S. Brunner, B. Langfeldt (Eds.): Advances in Breast Cancer Detection 1990. 53 figures, 51 tables. Xl, 195. ISBN 3-540-52089-9 Pierre Band (Ed.) Occupational Cancer Epidemiology With 23 Figures and 59 Tables Springer-Verlag Berlin Heidelberg New York London Paris Tokyo Hong Kong Barcelona Pierre Band, M.D. Head, Division of Epidemiology Biometry & Occupational Oncology Cancer Control Agency of British Columbia 600 West 10th Avenue Vancouver, B.C. V5Z 4E6, Canada ISBN-13 :978-3-642-84070-8 e-ISBN-13 :978-3-642-84068-5 DOl: 10.1007/978-3-642-84068-5 Library of Congress Cataloging-in-Publication Data Occupational cancer epidemiology/Pierre Band (ed.). p.cm. - (Recent results in cancer research; 120) ISBN-13:978-3-642-84070-8 (U.S.: alk. paper) 1. Cancer-Etiology. 2. Cancer-Risk factors. 3. Occupational diseases. 4. Cancer-Epidemiology. I. Band, P.R. (Pierre R.), 1935- . II. Serie •. [DNLM: 1. Neoplasms--epidemiology. 2. Occupational Diseases- epidemiology. WI RE106P v. 120/QZ 200 015] RC261.R35 vol. 120 [RC268.48] 616.99'4 s-dc20 [614.5'999] DNLM/DLC 90-9570 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in other ways, and storage in data banks. Duplication of this publication or parts thereof is only permitted under the provisions of the German Copyright Law of September 9, 1965, in its current version, and a copyright fee must always be paid. Violations fall under the prosecution act of the German Copyright Law. © Springer-Verlag Berlin Heidelberg 1990 Softcover reprint of the hardcover 1st edition 1990 The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Product Liability: The publisher can give no guarantee for information about drug dosage and application thereof contained in the book. In every individual case the respective user must check its accuracy by consulting other pharmaceutical literature. Typesetting: Macmillan India Ltd., Bangalore-25, India 2125/3140-543210-Printed on acid-free paper. Preface The identification of occupational cancer risk factors and of carcinogens in the workplace is assuming increasing import ance in cancer epidemiology. This book, which contains the proceedings of a symposium held in Vancouver in June 1988, combines overviews by experts on substantive topics and methodologic issues of broad interest in occupational cancer epidemiology. Among the former are state of the art reviews emphasizing recent data and new and innovative analytic approaches. The substantive topics include discussion of cancer risks from exposure to complex organic mixtures, asbestos and man-made fibers, herbicides, radon, and electromagnetic fields. Risk assessment, exposure assessment, and analysis of occupa tional cohort studies are examples of the methodologic issues addressed. This book provides basic information and opens perspectives on new areas of research. In organizing our symposium we were greatly assisted by Betty Fata and Kim Rust of Venue West Conference Ltd. We appreciate the financial support received from the following: Cancer Control Agency of British Columbia Alcan Aluminium Ltd. National Health Research & Development Program Workers' Compensation Board of British Columbia Purdue Frederick Inc. Rhone-Poulenc Pharma Inc. London Drugs Ltd. Ciba-Geigy Canada Ltd. Bristol-Myers Pharmaceutical Group Schering Canada Inc. Hoffmann-La Roche Ltd. Sterling Drug Ltd. Parke-Davis Canada Inc. Beecham Laboratories Pfizer Canada Inc. Roussel Canada Inc. Novopharm Contents D. Krewski, D. Wigle, D.B. Clayson, and G.R. Howe Role of Epidemiology in Health Risk Assessment. . . . .. 1 J. Siemiatycki Discovering Occupational Carcinogens in Population- Based Case-Control Studies: Review of Findings from an Exposure-Based Approach and a Methodologic Comparison of Alternative Data Collection Strategies. . . 25 M. Ghin Recent Approaches to Retrospective Exposure Assessment in Occupational Cancer Epidemiology . . . . . . . . . . . . 39 B. Armstrong Effects of Measurement Errors on Estimates of Exposure-Response Relationships ..... . . ..... 50 J.J. Spinelli, P.R. Band, and R.P. Gallagher Adjustment for Confounding in Occupational Cancer Epidemiology . . . . . . . . . . . . . . . . . . . . . . 64 N. Breslow Statistical Issues in the Analysis of Data from Occupational Cohort Studies. . . . . . . . .. ....... 78 H.F. Stich, A.B. Acton, and B. Palcic Towards an Automated Micronucleus Assay as an Internal Dosimeter for Carcinogen-Exposed Human Population Groups. . . . . . . . . . . . . . . . . . . 94 x.z. J.L. Mumford, He, and R.S. Chapman Human Lung Cancer Risks Due to Complex Organic Mixtures of Combustion Emissions ............. 106 J.e. McDonald Cancer Risks Due to Asbestos and Man-Made Fibres .. 122 VIII Contents A. Blair and S. Hoar Zahm Herbicides and Cancer: A Review and Discussion of Methodologic Issues. ... 132 O. Axelson Cancer Risks from Exposure to Radon Progeny in Mines and Dwellings. . . . . . . . . . . . . . . . . . . . . . 146 G. Theriault Cancer Risks Due to Exposure to Electromagnetic Fields. . . . . . . . . . . . . . . . . . . . . . . . . . .. .. 166 P.R. Band, J.J. Spinelli, R.P. Gallagher, w.J. Threlfall, v.T. Y. Ng, J. Moody, D. Raynor, L.M. Svirchev, D. Kan, and M. Wong Identification of Occupational Cancer Risks Using a Population-Based Cancer Registry 181 Subject Index. . . . . . . . . . . . . . . . . . . 190 List of Contributors* Acton, A.B. 941 McDonald, J.e. 122 Armstrong, B. 50 Moody, J. 181 Axelson, O. 146 Mumford, J.L. 106 Band, P.R. 64, 181 Ng, V.T.Y. 181 Blair, A. 132 Paicic, B. 94 Breslow, N. 78 Raynor, D. 181 Chapman, R.S. 106 Siemiatycki, J. 25 Clayson, D.B. 1 Spinelli, J.J. 64, 181 Gallagher, R.P. 64, 181 Stich, H.F. 94 Gerin, M. 39 Svirchev, L.M. 181 He, X.Z. 106 Theriault, G. 166 Hoar Zahm, S. 132 Threlfall, W.J. 181 Howe, G.R. 1 Wigle, D. 1 Kan, D. 181 Wong, M. 181 Krewski, D. 1 * The address of the principal author is given on the first page of each contribution. 1 Page on which contribution begins. Role of Epidemiology in Health Risk Assessment D. Krewski!, D. Wigle!, D.B. Clayson!, and G.R. Howeh 1 Health Protection Branch, Health and Welfare Canada, Ottawa, Ontario KIA OL2, Canada 2 National Cancer Institute of Canada, University of Toronto, Toronto, Ontario, M5S 118, Canada Introduction The process of health risk assessment has been the subject of systematic study in recent years. For example, interdisciplinary groups established by the U.S. National Research Council (NRC 1983) and the World Health Organization (WHO 1985) have developed models for risk assessment and risk management. These investigations have provided clear frameworks within which questions pertaining to health risks may be addressed. This has proven useful in distin guishing between the scientific and extra scientific dimensions of risk assessment, and in ensuring that all relevant factors are given proper attention in risk management decisions. Most of our information on environmental and occupational health risks is based on epidemiological investigation of people exposed to the risk factor of interest or on toxicological experiments conducted on nonhuman test systems. Although epidemiological studies have the advantage of providing inform ation directly on humans, they are subject to certain limitations. Epidemiolo gical studies are only possible after human exposure has occurred, and after the health effect of interest has had sufficient time to develop. With chronic diseases such as cancer, this induction period may span one or two generations. There may also be technical problems in exposure assessment and adjustment for the presence of confounding risk factors in observational studies of human populations. In contrast, toxicological studies afford the opportunity to determine possible adverse effects before human exposure is permitted. Similarly, because experi ments are conducted in a carefully controlled laboratory environment, the toxicologist can attain a degree of precision denied to the epidemiologist who must work with a heterogeneous human population exposed to a wide variety of risk factors influencing health status. However, because toxicological studies are conducted in nonhuman test systems, this precision is achieved under penalty of * The authors would like to thank Terry Chernis for bibliographic assistance In preparing this article. Recent Results in Cancer Research. Vol. 120 © Springer-Verlag Berlin-Heidelberg 1990 2 D. Krewski et al. introducing the need for uncertain extrapolations from the test system em ployed to humans. In this paper, we examine the complementary roles of epidemiology and toxicology in human health risk assessment. We begin with a description of a model for risk assessment and risk management which will serve to place the role of epidemiology in health risk assessment in perspective. We then provide a brief description of the different epidemiological protocols which have been used in health risk assessment. Case studies of active and passive smoking as well as saccharin and formaldehyde are presented to illustrate the complement ary roles of epidemiology and toxicology in health risk assessment. The reso lution of discrepancies between epidemiological and toxicological findings is treated next, including the use of weight of evidence approaches for combining information from both sources to arrive at an overall assessment of risk. Finally, we present a summary of our main findings. Risk Assessment and Risk Management A number of recent studies have examined the process of health risk assessment and risk management. For example, several national and international organ izations have proposed specific models to describe this process (see Krewski and Birkwood 1987, for a recent detailed review). These models exhibit many similarities and serve to clarify the important elements of risk assessment and risk management, including the role of epidemiology in this process (Krewski 1987). Building on these foundations, a Working Group on Risk Assessment and Risk Management (1988) developed a model for use within the Environmental Health Directorate of Health and Welfare Canada (Fig. 1). The first two steps involve the identification of a specific environmental hazard and the estimation of the corresponding levels of risk. These two steps combined involve the analysis of epidemiological and toxicological data, and together comprise what may be termed risk analysis. Following risk analysis, the risk manager may consider a variety of options for the protection of human health. These risk management strategies may be broadly classified as regulatory, economic, advisory or technological (Krewski and Birkwood 1988). While direct regulation continues to be widely used, nonregulatory alternatives have received increasing attention in recent years. For example, economic options rely on economic incentives and disincentives to reduce environmental pollution. Advisory approaches rely on the provision of advice to promote risk avoidance. Formal regulations may also be avoided if new technological developments are used for pollution abatement. These op tions are not mutually exclusive, as demonstrated by the use of a combination of all four approaches in attempting to control acidic deposition in Canada (Burnett et al. 1988). The selection of a suitable risk management strategy is complex, and is frequently charged with social and political overtones (Somers 1983, 1984). As