Principles of toxicology - PDF Free Download (2023)


PRINCIPLES OF TOXICOLOGY Environmental and Industrial Applications SECOND EDITION

Edited by

Phillip L. Williams, Ph.D. Associate Professor Department of Environmental Health Science University of Georgia Athens, Georgia

Robert C. James, Ph.D. President, TERRA, Inc. Tallahassee, Florida Associate Scientist, Interdisciplinary Toxicology Center for Environmental and Human Toxicology University of Florida Gainesville, Florida

Stephen M. Roberts, Ph.D. Professor and Program Director Center for Environmental and Human Toxicology University of Florida Gainesville, Florida

A Wiley-Interscience Publication JOHN WILEY & SONS, INC. New York






This book is printed on acid-free paper. Copyright © 2000 by John Wiley & Sons, Inc. All rights reserved. Published simultaneously in Canada. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4744. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 605 Third Avenue, New York, NY 10158-0012, (212) 850-6011, fax (212) 850-6008, E-Mail: [emailprotected] For ordering and customer service, call 1-800-CALL-WILEY. Library of Congress Cataloging in Publication Data: Principles of toxicology: environmental and industrial applications / edited by Phillip L. Williams, Robert C. James, Stephen M. Roberts.—2nd ed. p. cm. Update and expansion on a previous text entitled: Industrial toxicology: safety and health applications in the workplace. Includes bibliographical references and index. ISBN 0-471-29321-0 (cloth: alk. paper) 1. Toxicology. 2. Industrial toxicology. 3. Environmental toxicology. I. Williams, Phillip L., 1952- II. James, Robert C., 1947- III. Roberts, Stephen M., 1950RA1211 .P746 2000 615.9Y02—dc21 Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1


CONTRIBUTORS LOUIS ADAMS, PH.D. Professor, Department of Medicine, University of Cincinnati, Cincinnati, Ohio JUDY A. BEAN, PH.D., Director, Biostatistics Program, Children’s Hospital, Cincinnati, Ohio CHISTOPHER J. BORGERT, PH.D., President and Principal Scientist, Appied Pharmacology and Toxicology, Inc.; Assistant Scientist, Department of Physiological Sciences, University of Florida College of Veterinary Medicine, Alachua, Florida JANICE K. BRITT, PH.D., Senior Toxicologist, TERRA, Inc., Tallahassee, Florida ROBERT A. BUDINSKY, JR., PH.D., Senior Toxicologist, ATRA, Inc., Tallahassee, Florida CHAM E. DALLAS, PH.D., Associate Professor and Director, Interdisciplinary Toxicology Program, University of Georgia, Athens, Georgia ROBERT P. DEMOTT, PH.D., Chemical Risk Group Manager, GeoSyntec Consultants, Inc., Tampa, Florida STEVEN G. DONKIN, PH.D., Senior Scientist, Sciences International, Inc., Alexandria, Virginia LORA E. FLEMING, M.D., PH.D., MPH, Associate Professor, Department of Epidemiology and Public Health, University of Miami, Miami, Florida MICHAEL R. FRANKLIN, PH.D., Interim Chair and Professor, Department of Pharmacology and Toxicology, University of Utah, Salt Lake City, Utah HOWARD FRUMKIN, M.D., DR.P.H., Chair and Associate Professor, Department of Environmental and Occupational Health, The Rollins School of Public Health, Emory University, Atlanta, Georgia EDWARD I. GALAID, M.D., MPH, Clinical Assistant Professor, Department of Environmental and Occupational Health, The Rollins School of Public Health, Emory University, Atlanta JAY GANDY, PH.D., Senior Toxicologist, Center for Toxicology and Environmental Health, Little Rock, Arkansas FREDRIC GERR, M.D., Associate Professor, Department of Environmental and Occupational Health, The Rollins School of Public Health, Emory University, Atlanta, Georgia PHILLIP T. GOAD, PH.D., President, Center for Toxicology and Environmental Health, Little Rock, Arkansas CHRISTINE HALMES, PH.D., Toxicologist, TERRA, Inc., Denver, Colorado DAVID E. JACOBS, PH.D., Director, Office of Lead Hazard Control, U.S. Department of Housing and Urban Development, Washington, D.C. ROBERT C. JAMES, PH.D., President, TERRA, Inc., Tallahassee, Florida; Associate Scientist, Interdisciplinary Toxicology, Center for Environmental and Human Toxicology, University of Florida, Gainesville, Florida WILLIAM R. KERN, PH.D., Professor, Department of Pharmacology and Therapeutics, University of Florida, Gainesville, Florida v



PAUL J. MIDDENDORF, PH.D., Principal Research Scientist, Georgia Tech Research Institute, Atlanta, Georgia GLENN C. MILLNER, PH.D., Vice President, Center for Toxicology and Environmental Health, Little Rock, Arkansas ALAN C. NYE, PH.D., Vice President, Center for Toxicology and Environmental Health, Little Rock, Arkansas ELLEN J. O’FLAHERTY, PH.D., Professor, Department of Environmental Health, University of Cincinnati, Cincinnati, Ohio DANNY L. OHLSON, PH.D., Toxicologist, Hazardous Substances and Waste Management Research, Tallahassee, Florida STEPHEN M. ROBERTS, PH.D., Professor and Program Director, Center for Environmental and Human Toxicology, University of Florida, Gainesville, Florida WILLIAM R. SALMINEN, PH.D., Consulting Toxicologist, Toxicology Division, Exxon Biomedical Sciences, Inc., East Millstone, New Jersey CHRISTOPER J. SARANKO, PH.D., Post Doctoral Fellow, Center for Environmental and Human Toxicology, University of Florida, Gainesville, Florida CHRISTOPER M. TEAF, PH.D., President, Hazardous Substances and Waste Management Research, Tallahassee, Florida; Associate Director, Center for Biochemical and Toxicological Research and Hazardous Waste Management, Florida State University, Tallahassee, Florida D. ALAN WARREN, PH.D., Toxicologist, TERRA, Inc., Tallahassee, Florida PHILLIP L. WILLIAMS, PH.D., Associate Professor, Department of Environmental Health Science, University of Georgia, Athens, Georgia GAROLD S. YOST, PH.D., Professor, Department of Pharmacology and Toxicology, University of Utah, Salt Lake City, Utah


xv xvii



1 General Principles of Toxicology


Robert C. James, Stephen M. Roberts, and Phillip L. Williams 1.1 1.2 1.3 1.4 1.5 1.6 1.7

Basic Definitions and Terminology 3 What Toxicologists Study 5 The Importance of Dose and the Dose–Response Relationship 7 How Dose–Response Data Can Be Used 17 Avoiding Incorrect Conclusions from Dose–Response Data 19 Factors Influencing Dose–Response Curves 21 Descriptive Toxicology: Testing Adverse Effects of Chemicals and Generating Dose–Response Data 26 1.8 Extrapolation of Animal Test Data to Human Exposure 28 1.9 Summary 32 References and Suggested Reading 32 2 Absorption, Distribution, and Elimination of Toxic Agents


Ellen J. O’Flaherty 2.1 2.2 2.3 2.4 2.5

Toxicology and the Safety and Health Professions 35 Transfer across Membrane Barriers 37 Absorption 41 Disposition: Distribution and Elimination 45 Summary 53 References and Suggested Reading 54

3 Biotransformation: A Balance between Bioactivation and Detoxification


Michael R. Franklin and Garold S. Yost 3.1 Sites of Biotransformation 62 3.2 Biotransformation Reactions 65 3.3 Summary 85 Suggested Reading 86 vii



4 Hematotoxicity: Chemically Induced Toxicity of the Blood


Robert A. Budinsky Jr. 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 4.15 4.16 4.17

Hematotoxicity: Basic Concepts and Background 87 Basic Hematopoiesis: The Formation of Blood Cells and their Differentiation 88 The Myeloid Series: Erythrocytes, Platelets, Granulocytes (Neutrophils), Macrophages, Eosinophils, and Basophils 91 The Lymphoid Series: Lymphocytes (B and T Cells) 94 Direct Toxicological Effects on the RBC: Impairment of Oxygen Transport and Destruction of the Red Blood Cell 95 Chemicals that Impair Oxygen Transport 97 Inorganic Nitrates/Nitrites and Chlorate Salts 99 Methemoglobin Leading to Hemolytic Anemia: Aromatic Amines and Aromatic Nitro Compounds 100 Autoimmune Hemolytic Anemia 101 Bone Marrow Suppression and Leukemias and Lymphomas 102 Chemical Leukemogenesis 104 Toxicities that Indirectly Involve the Red Blood Cell 105 Cyanide (CN) Poisoning 105 Hydrogen Sulfide (H2S) Poisoning 105 Antidotes for Hydrogen Sulfide and Cyanide Poisoning 107 Miscellaneous Toxicities Expressed in the Blood 108 Summary 108 References and Suggested Reading 108

5 Hepatotoxicity: Toxic Effects on the Liver


Stephen M. Roberts, Robert C. James, and Michael R. Franklin 5.1 The Physiologic and Morphologic Bases of Liver Injury 111 5.2 Types of Liver Injury 116 5.3 Evaluation of Liver Injury 124 References and Suggested Reading 127 6 Nephrotoxicity: Toxic Responses of the Kidney


Paul J. Middendorf and Phillip L. Williams 6.1 6.2 6.3 6.4

Basic Kidney Structures and Functions 129 Functional Measurements to Evaluate Kidney Injury 135 Adverse Effects of Chemicals on the Kidney 137 Summary 142 References and Suggested Reading 143

7 Neurotoxicity: Toxic Responses of the Nervous System Steven G. Donkin and Phillip L. Williams 7.1 Mechanisms of Neuronal Transmission 146 7.2 Agents that Act on the Neuron 149



7.3 7.4 7.5 7.6 7.7


Agents that Act on the Synapse 151 Interactions of Industrial Chemical with Other Substances 151 General Population Exposure to Environmental Neurotoxicants 152 Evaluation of Injury to the Nervous System 152 Summary 154 References and Suggested Reading 155

8 Dermal and Ocular Toxicology: Toxic Effects of the Skin and Eyes


William R. Salminen and Stephen M. Roberts 8.1 8.2 8.3 8.4

Skin Histology 157 Functions 158 Contact Dermatitis 160 Summary 167 References and Suggested Reading 168

9 Pulmonotoxicity: Toxic Effects in the Lung


Cham E. Dallas 9.1 Lung Anatomy and Physiology 169 9.2 Mechanisms of Industrially Related Pulmonary Diseases 181 9.3 Summary 185 References and Suggested Reading 186 10 Immunotoxicity: Toxic Effects on the Immune System


Stephen M. Roberts and Louis Adams 10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8

Overview of Immunotoxicity 189 Biology of the Immune Response 189 Types of Immune Reactions and Disorders 194 Clinical Tests for Detecting Immunotoxicity 195 Tests for Detecting Immunotoxicity in Animal Models 197 Specific Chemicals that Adversely Affect the Immune System 199 Multiple-Chemical Sensitivity 203 Summary 205 References and Suggested Reading 205



11 Reproductive Toxicology


Robert P. DeMott and Christopher J. Borgert 11.1 11.2 11.3 11.4 11.5

Male Reproductive Toxicology 210 Female Reproductive Toxicology 218 Developmental Toxicology 224 Current Research Concerns 232 Summary 236 References and Suggested Reading 236



12 Mutagenesis and Genetic Toxicology


Christopher M. Teaf and Paul J. Middendorf 12.1 12.2 12.3 12.4 12.5 12.6

Induction and Potential Consequences of Genetic Change 239 Genetic Fundamentals and Evaluation of Genetic Change 241 Nonmammalian Mutagenicity Tests 251 Mammalian Mutagenicity Tests 253 Occupational Significance of Mutagens 257 Summary 261 References and Suggested Reading 263

13 Chemical Carcinogenesis


Robert C. James and Christopher J. Saranko 13.1 13.3 13.4 13.5 13.6 13.7 13.8 13.9

The Terminology of Cancer 266 Carcinogenesis by Chemicals 268 Molecular Aspects of Carcinogenesis 280 Testing Chemicals for Carcinogenic Activity 289 Interpretation Issues Raised by Conditions of the Test Procedure 292 Empirical Measures of Reliability of the Extrapolation 299 Occupational Carcinogens 301 Cancer and Our Environment: Factors that Modulate Our Risks to Occupational Hazards 304 13.10 Cancer Trends and Their Impact on Evaluation of Cancer Causation 319 13.11 Summary 321 References and Suggested Reading 323

14 Properties and Effects of Metals


Steven G. Donkin, Danny L. Ohlson, and Christopher M. Teaf 14.1 14.2 14.3 14.4 14.5 14.6 14.7

Classification of Metals 325 Speciation of Metals 327 Pharmacokinetics of Metals 328 Toxicity of Metals 331 Sources of Metal Exposure 334 Toxicology of Selected Metals 336 Summary 343 References and Suggested Reading 343

15 Properties and Effects of Pesticides Janice K. Britt 15.1 Organophosphate and Carbamate Insecticides 346 15.2 Organochlorine Insecticides 352 15.3 Insecticides of Biological Origin 353



15.4 15.5 15.6 15.7 15.8


Herbicides 356 Fungicides 358 Rodenticides 360 Fumigants 361 Summary 362 References and Suggested Reading 363

16 Properties and Effects of Organic Solvents


Christopher M. Teaf 16.1 16.2 16.3 16.4 16.5 16.6 16.7 16.8 16.9 16.10 16.11 16.12 16.13 16.14 16.15 16.16 16.17 16.18 16.19

Exposure Potential 367 Basic Principles 368 Toxic Properties of Representative Aliphatic Organic Solvents 377 Toxic Properties of Representative Alicyclic Solvents 378 Toxic Properties of Representative Aromatic Hydrocarbon Solvents 379 Toxic Properties of Representative Alcohols 382 Toxic Properties of Representative Phenols 384 Toxic Properties of Representative Aldehydes 385 Toxic Properties of Representative Ketones 388 Toxic Properties of Representative Carboxylic Acids 389 Toxic Properties of Representative Esters 390 Toxic Properties of Representative Ethers 390 Toxic Properties of Representative Halogenated Alkanes 391 Toxic Properties of Representative Nitrogen-Substituted Solvents 398 Toxic Properties of Representative Aliphatic and Aromatic Nitro Compounds 402 Toxic Properties of Representative Nitriles (Alkyl Cyanides) 404 Toxic Properties of the Pyridine Series 405 Sulfur-Substituted Solvents 405 Summary 407 References and Suggested Reading 407

17 Properties and Effects of Natural Toxins and Venoms William R. Kem 17.1 17.2 17.3 17.4 17.5 17.6 17.7 17.8 17.9

Poisons, Toxins, and Venoms 409 Molecular and Functional Diversity of Natural Toxins and Venoms 410 Natural Roles of Toxins and Venoms 411 Major Sites and Mechanisms of Toxic Action 411 Toxins in Unicellular Organisms 415 Toxins of Higher Plants 417 Animal Venoms and Toxins 423 Toxin and Venom Therapy 430 Summary 432 Acknowledgments 432 References and Suggested Reading 432






18 Risk Assessment


Robert C. James, D. Alan Warren, Christine Halmes, and Stephen M. Roberts 18.1 18.3 18.4 18.5 18.6 18.7 18.8 18.9 18.10

Risk Assessment Basics 437 Exposure Assessment: Exposure Pathways and Resulting Dosages 445 Dose–Response Assessment 449 Risk Characterization 460 Probabilistic Versus Deterministic Risk Assessments 462 Evaluating Risk from Chemical Mixtures 464 Comparative Risk Analysis 468 Risk Communication 472 Summary 474 References and Suggested Reading 475

19 Example of Risk Assessment Applications


Alan C. Nye, Glenn C. Millner, Jay Gandy, and Phillip T. Goad 19.1. 19.2. 19.3. 19.4. 19.5. 19.6. 19.7.

Tiered Approach to Risk Assessment 479 Risk Assessment Examples 480 Lead Exposure and Women of Child-bearing Age 481 Petroleum Hydrocarbons: Assessing Exposure and Risk to Mixtures 483 Risk Assessment for Arsenic 486 Reevaluation of the Carcinogenic Risks of Inhaled Antimony Trioxide 490 Summary 496 References and Suggested Reading 497

20 Occupational and Environmental Health


Fredric Gerr, Edward Galaid, and Howard Frumkin 20.1 20.2 20.3 20.4 20.5 20.6 20.7

Definition and Scope of the Problem 499 Characteristics of Occupational Illness 502 Goals of Occupational and Environmental Medicine 502 Human Resources Important to Occupational Health Practice 503 Activities of the Occupational Health Provider 503 Ethical Considerations 507 Summary and Conclusion 508 References and Suggested Reading 509

21 Epidemiologic Issues in Occupational and Environmental Health Lora E. Fleming and Judy A. Bean 21.1 21.2 21.3 21.4 21.5

A Brief History of Epidemiology 511 Epidemiologic Causation 512 Types of Epidemiologic Studies: Advantages and Disadvantages 513 Exposure Issues 514 Disease and Human Health Effects Issues 515



21.6 21.7 21.8 21.9 21.10 21.11


Population Issues 516 Measurement of Disease or Exposure Frequency 516 Measurement of Association Or Risk 517 Bias 519 Other Issues 520 Summary 520 References and Suggested Reading 520

22 Controlling Occupational and Environmental Health Hazards


Paul J. Middendorf and David E. Jacobs 22.1 22.2 22.3 22.4 22.5

Background and Historical Perspective 523 Exposure Limits 524 Program Management 530 Case Studies 541 Summary 552 References and Suggested Reading 553





PREFACE Purpose of This Book Principles of Toxicology: Environmental and Industrial Applications presents compactly and efficiently the scientific basis to toxicology as it applies to the workplace and the environment. The book covers the diverse chemical hazards encountered in the modern work and natural environment and provides a practical understanding of these hazards for those concerned with protecting the health of humans and ecosystems. Intended Audience This book represents an update and expansion on a previous, very successful text entitled Industrial Toxicology: Satety and Health Applications in the Workplace. It retains the emphasis on applied aspects of toxicology, while extending its scope beyond the industrial setting to include environmental toxicology. The book was written for those health professionals who need toxicological information and assistance beyond that of an introductory text in general toxicology, yet more practical than that in advanced scientific works on toxicology. In particular, we have in mind industrial hygienists, occupational physicians, safety engineers, environmental health practitioners, occupational health nurses, safety directors, and environmental scientists. Organization of the Book This volume consists of three parts. Part I establishes the scientific basis to toxicology, which is then applied through the rest of the book. This part discusses concepts such as absorption, distribution, and elimination of toxic agents from the body. Chapters 4–10 discuss the effects of toxic agents on specific physiological organs or systems, including the blood, liver, kidneys, nerves, skin, lungs, and the immune system. Part II addresses specific areas of concern in the occupational and environmental—both toxic agents and their manifestations. Chapters 11–13 examine areas of great research interest—reproductive toxicology, mutagenesis, and carcinogenesis. Chapters 14–17 examine toxic effects of metals, pesticides, organic solvents, and natural toxins and venoms. Part III is devoted to specific applications of the toxicological principles from both the environmental and occupational settings. Chapters 18 and 19 cover risk assessment and provide specific case studies that allow the reader to visualize the application of risk assessment process. Chapters 20 and 21 discuss occupational medicine and epidemiologic issues. The final chapter is devoted to hazard control. Features The following features from Principles of Toxicology: Environmental and Industrial Applications will be especially useful to our readers:

• The book is compact and practical, and the information is structured for easy use by the health professional in both industry and government. xv



• The approach is scientific, but applied, rather than theoretical. In this it differs from more • • • • •

general works in toxicology, which fail to emphasize the information pertinent to the industrial environment. The book consistently stresses evaluation and control of toxic hazards. Numerous illustrations and figures clarify and summarize key points. Case histories and examples demonstrate the application of toxicological principles. Chapters include annotated bibliographies to provide the reader with additional useful information. A comprehensive glossary of toxicological terms is included. Phillip L. Williams Robert C. James Stephen M. Roberts

ACKNOWLEDGMENTS A text of this undertaking on the broad topic of toxicology would not be possible except for the contributions made by each of the authors in their field(s) of speciality. We especially appreciate the contributors patience during the many years it took to complete this revision. In addition, such an undertaking would not have been possible without the support provided by each of our employers— The University of Georgia, TERRA, Inc., and The University of Florida. We also owe a thank you to Valerie Rocchi for her administrative assistance throughout the effort and to Dr. Kelly McDonald for her editorial assistance. Phillip L. Williams Robert C. James Stephen M. Roberts



PART I Conceptual Aspects

Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.

1 General Principles of Toxicology GENERAL PRINCIPLES OF TOXICOLOGY


The intent of this chapter is to provide a concise description of the basic principles of toxicology and to illustrate how these principles are used to make reasonable judgments about the potential health hazards and the risks associated with chemical exposures. This chapter explains

• Some basic definitions and terminology • What toxicologists study, the scientific disciplines they draw upon, and specialized areas of • • • •

interest within toxicology Descriptive toxicology and the use of animal studies as the primary basis for hazard identification, the importance of dose, and the generation of dose–response relationships How dose–response data might be used to assess safety or risk Factors that might alter a chemical’s toxicity or the dose–response relationship The basic methods for extrapolating dose–response data when developing exposure guidelines of public health interest

1.1 BASIC DEFINITIONS AND TERMINOLOGY The literal meaning of the term toxicology is “ the study of poisons.” The root word toxic entered the English language around 1655 from the Late Latin word toxicus (which meant poisonous), itself derived from toxikón, an ancient Greek term for poisons into which arrows were dipped. The early history of toxicology focused on the understanding and uses of different poisons, and even today most people tend to think of poisons as a deadly potion that when ingested causes almost immediate harm or death. As toxicology has evolved into a modern science, however, it has expanded to encompass all forms of adverse health effects that substances might produce, not just acutely harmful or lethal effects. The following definitions reflect this expanded scope of the science of toxicology: Toxic—having the characteristic of producing an undesirable or adverse health effect. Toxicity—any toxic (adverse) effect that a chemical or physical agent might produce within a living organism. Toxicology—the science that deals with the study of the adverse effects (toxicities) chemicals or physical agents may produce in living organisms under specific conditions of exposure. It is a science that attempts to qualitatively identify all the hazards (i.e., organ toxicities) associated with a substance, as well as to quantitatively determine the exposure conditions under which those hazards/toxicities are induced. Toxicology is the science that experimentally investigates the occurrence, nature, incidence, mechanism, and risk factors for the adverse effects of toxic substances.

Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




As these definitions indicate, the toxic responses that form the study of toxicology span a broad biologic and physiologic spectrum. Effects of interest may range from something relatively minor such as irritation or tearing, to a more serious response like acute and reversible liver or kidney damage, to an even more serious and permanent disability such as cirrhosis of the liver or liver cancer. Given this broad range of potentially adverse effects to consider, it is perhaps useful for those unfamiliar with toxicology to define some additional terms, listed in order of relevance to topics that might be discussed in Chapters 2–22 of this book. Exposure—to cause an adverse effect, a toxicant must first come in contact with an organism. The means by which an organism comes in contact with the substance is the route of exposure (e.g., in the air, water, soil, food, medication) for that chemical. Dose—the total amount of a toxicant administered to an organism at specific time intervals. The quantity can be further defined in terms of quantity per unit body weight or per body surface area. Internal/absorbed dose—the actual quantity of a toxicant that is absorbed into the organism and distributed systemically throughout the body. Delivered/effective/target organ dose—the amount of toxicant reaching the organ (known as the target organ) that is adversely affected by the toxicant. Acute exposure—exposure over a brief period of time (generally less than 24 h). Often it is considered to be a single exposure (or dose) but may consist of repeated exposures within a short time period. Subacute exposure—resembles acute exposure except that the exposure duration is greater, from several days to one month. Subchronic exposure—exposures repeated or spread over an intermediate time range. For animal testing, this time range is generally considered to be 1–3 months. Chronic exposure—exposures (either repeated or continuous) over a long (greater than 3 months) period of time. With animal testing this exposure often continues for the majority of the experimental animal’s life, and within occupational settings it is generally considered to be for a number of years. Acute toxicity—an adverse or undesirable effect that is manifested within a relatively short time interval ranging from almost immediately to within several days following exposure (or dosing). An example would be chemical asphyxiation from exposure to a high concentration of carbon monoxide (CO). Chronic toxicity—a permanent or lasting adverse effect that is manifested after exposure to a toxicant. An example would be the development of silicosis following a long-term exposure to silica in workplaces such as foundries. Local toxicity—an adverse or undesirable effect that is manifested at the toxicant’s site of contact with the organism. Examples include an acid’s ability to cause burning of the eyes, upper respiratory tract irritation, and skin burns. Systemic toxicity—an adverse or undesirable effect that can be seen throughout the organism or in an organ with selective vulnerability distant from the point of entry of the toxicant (i.e., toxicant requires absorption and distribution within the organism to produce the toxic effect). Examples would be adverse effects on the kidney or central nervous system resulting from the chronic ingestion of mercury. Reversible toxicity—an adverse or undesirable effect that can be reversed once exposure is stopped. Reversibility of toxicity depends on a number of factors, including the extent of exposure (time and amount of toxicant) and the ability of the affected tissue to repair or regenerate. An example includes hepatic toxicity from acute acetaminophen exposure and liver regeneration.



Delayed or latent toxicity—an adverse or undesirable effect appearing long after the initiation and/or cessation of exposure to the toxicant. An example is cervical cancer during adulthood resulting from in utero exposure to diethylstilbestrol (DES). Allergic reaction—a reaction to a toxicant caused by an altered state of the normal immune response. The outcome of the exposure can be immediate (anaphylaxis) or delayed (cell-mediated). Idiosyncratic reaction—a response to a toxicant occurring at exposure levels much lower than those generally required to cause the same effect in most individuals within the population. This response is genetically determined, and a good example would be sensitivity to nitrates due to deficiency in NADH (reduced-form nicotinamide adenine dinucleotide phosphate)– methemoglobin reductase. Mechanism of toxicity—the necessary biologic interactions by which a toxicant exerts its toxic effect on an organism. An example is carbon monoxide (CO) asphyxiation due to the binding of CO to hemoglobin, thus preventing the transport of oxygen within the blood. Toxicant—any substance that causes a harmful (or adverse) effect when in contact with a living organism at a sufficiently high concentration. Toxin—any toxicant produced by an organism (floral or faunal, including bacteria); that is, naturally produced toxicants. An example would be the pyrethrins, which are natural pesticides produced by pyrethrum flowers (i.e., certain chrysanthemums) that serve as the model for the man made insecticide class pyrethroids. Hazard—the qualitative nature of the adverse or undesirable effect (i.e., the type of adverse effect) resulting from exposure to a particular toxicant or physical agent. For example, asphyxiation is the hazard from acute exposures to carbon monoxide (CO). Safety—the measure or mathematical probability that a specific exposure situation or dose will not produce a toxic effect. Risk—the measure or probability that a specific exposure situation or dose will produce a toxic effect. Risk assessment—the process by which the potential (or probability of) adverse health effects of exposure are characterized.

1.2 WHAT TOXICOLOGISTS STUDY Toxicology has become a science that builds on and uses knowledge developed in other related medical sciences, such as physiology, biochemistry, pathology, pharmacology, medicine, and epidemiology, to name only a few. Given its broad and diverse nature, toxicology is also a science where a number of areas of specialization have evolved as a result of the different applications of toxicological information that exist within society today. It might be argued, however, that the professional activities of all toxicologists fall into three main areas of endeavor: descriptive toxicology, research/mechanistic toxicology, and applied toxicology. Descriptive toxicologists are scientists whose work focuses on the toxicity testing of chemicals. This work is done primarily at commercial and governmental toxicity testing laboratories, and the studies performed at these facilities are designed to generate basic toxicity information that can be used to identify the various organ toxicities (hazards) that the test agent is capable of inducing under a wide range of exposure conditions. A thorough “ descriptive toxicological” analysis would identify all possible acute and chronic toxicities, including the genotoxic, reproductive, teratogenic (developmental), and carcinogenic potential of the test agent. It would also identify important metabolites of the chemical that are generated as the body attempts to break down and eliminate the chemical, as well as analyze the manner in which the chemical is absorbed into the body, distributed throughout the body and accumulated by various tissues and organs, and then ultimately excreted from the body. Hopefully,



appropriate dose–response test data are generated for those toxicities of greatest concern during the completion of the descriptive studies so that the relative safety of any given exposure or dose level that humans might typically encounter can be determined. Basic research or mechanistic toxicologists are scientists who study the chemical or agent in depth for the purpose of gaining an understanding of how the chemical or agent initiates those biochemical or physiological changes within the cell or tissue that result in the toxicity (adverse effect). They identify the critical biological processes within the organism that must be affected by the chemical to produce the toxic properties that are ultimately observed. Or, to state it another way, the goal of mechanistic studies is to understand the specific biological reactions (i.e., the adverse chain of events) within the affected organism that ultimately result in the toxicity under investigation. These experiments may be performed at the molecular, biochemical, cellular, or tissue level of the affected organism, and thus incorporate and apply the knowledge of a number of many other related scientific disciplines within the biological and medical sciences (e.g., physiology, biochemistry, genetics, molecular biology). Mechanistic studies ultimately are the bridge of knowledge that connects functional observations made during descriptive toxicological studies to the extrapolations of dose–response information that is used as the basis of risk assessment and exposure guideline development (e.g., occupational health guidelines or governmental regulations) by applied toxicologists. Applied toxicologists are scientists concerned with the use of chemicals in a “ real world” or nonlaboratory setting. For example, one goal of applied toxicologists is to control the use of the chemical in a manner that limits the probable human exposure level to one in which the dose any individual might receive is a safe one. Toxicologists who work in this area of toxicology, whether they work for a state or federal agency, a company, or as consultants, use descriptive and mechanistic toxicity studies to develop some identifiable measure of the safe dose of the chemical. The process whereby this safe dose or level of exposure is derived is generally referred to as the area of risk assessment. Within applied toxicology a number of subspecialties occur. These are: forensic toxicology, clinical toxicology, environmental toxicology, and occupational toxicology. Forensic toxicology is that unique combination of analytical chemistry, pharmacology, and toxicology concerned with the medical and legal aspects of drugs and poisons; it is concerned with the determination of which chemicals are present and responsible in exposure situations of abuse, overdose, poisoning, and death that become of interest to the police, medical examiners, and coroners. Clinical toxicology specializes in ways to treat poisoned individuals and focuses on determining and understanding the toxic effects of medicines and simple over-the-counter (nonprescription) drugs. Environmental toxicology is the subdiscipline concerned with those chemical exposure situations found in our general living environment. These exposures may stem from the agricultural application of chemicals (e.g., pesticides, growth regulators, fertilizers), the release of chemicals during modern-day living (e.g., chemicals released by household products), regulated and unintentional industrial discharges into air or waterways (e.g., spills, stack emissions, NPDES discharges, etc.), and various nonpoint emission sources (e.g., the combustion byproducts of cars). This specialty largely focuses on those chemical exposures referred to as environmental contamination or pollution. Within this area there may be even further subspecialization (e.g., ecotoxicology, aquatic toxicology, mammalian toxicology, avian toxicology). Occupational toxicology is the subdiscipline concerned with the chemical exposures and diseases found in the workplace. Regardless of the specialization within toxicology, or the types of toxicities of major interest to the toxicologist, essentially every toxicologist performs one or both of the two basic functions of toxicology, which are to (1) examine the nature of the adverse effects produced by a chemical or physical agent (hazard identification function) and (2) assess the probability of these toxicities occurring under specific conditions of exposure (risk assessment function). Ultimately, the goal and basic purpose of toxicology is to understand the toxic properties of a chemical so that these adverse effects can be prevented by the development of appropriate handling or exposure guidelines.



1.3 THE IMPORTANCE OF DOSE AND THE DOSE–RESPONSE RELATIONSHIP It is probably safe to say that among lay individuals there exists considerable confusion between the terms poisonous and toxic. If asked, most lay individuals would probably define a toxic substance using the same definition that one would apply to highly poisonous chemicals, that is, chemicals capable of producing a serious injury or death quickly and at very low doses. However, this is not a particularly useful definition because all chemicals may induce some type of adverse effect at some dose, so all chemicals may be described as toxic. As we have defined toxicants (toxic chemicals) as agents capable of producing an adverse effect in a biological system, a reasonable question for one to ask becomes “ Which group of chemicals do we consider to be toxic?” or “ Which chemicals do we consider safe?” The short answer to both questions, of course, is all chemicals; for even relatively safe chemicals can become toxic if the dose is high enough, and even potent, highly toxic chemicals may be used safely if exposure is kept low enough. As toxicology evolved from the study of just those substances or practices that were poisonous, dangerous, or unsafe, and instead became a more general study of the adverse effects of all chemicals, the conditions under which chemicals express toxicity became as important as, if not more important than, the kind of adverse effect produced. The importance of understanding the dose at which a chemical becomes toxic (harmful) was recognized centuries ago by Paracelsus (1493–1541), who essentially stated this concept as “ All substances are poisons; there is none which is not a poison. The right dose differentiates a poison and a remedy.” In a sense this statement serves to emphasize the second function of toxicology, or risk assessment, as it indicates that concern for a substance’s toxicity is a function of one’s exposure to it. Thus, the evaluation of those circumstances and conditions under which an adverse effect can be produced is key to considering whether the exposure is safe or hazardous. All chemicals are toxic at some dose and may produce harm if the exposure is sufficient, but all chemicals produce their harm (toxicities) under prescribed conditions of dose or usage. Consequently, another way of viewing all chemicals is that provided by Emil Mrak, who said “ There are no harmless substances, only harmless ways of using substances.” These two statements serve to remind us that describing a chemical exposure as being either harmless or hazardous is a function of the magnitude of the exposure (dose), not the types of toxicities that a chemical might be capable of producing at some dose. For example, vitamins, which we consciously take to improve our health and well-being, continue to rank as a major cause of accidental poisoning among children, and essentially all the types of toxicities that we associate with the term “ hazardous chemicals” may be produced by many of the prescription medicines in use today. To help illustrate this point, and to begin to emphasize the fact that the dose makes the poison, the reader is invited to take the following pop quiz. First, cross-match the doses listed in column A of Table 1.1, doses that produce lethality in 50 percent of the animals (LD50), to the correct chemical listed in column B. The chemicals listed in column B are a collection of food additives, medicines, drugs of abuse, poisons, pesticides, and hazardous substances for which the correct LD50 is listed somewhere in column A. To perform this cross-matching, first photocopy Table 1.1 and simply mark the ranking of the dose (i.e., the number corresponding next to the dose in column A) you believe correctly corresponds to the chemical it has been measured for in column B. [Note: The doses are listed in descending order, and the chemicals have been listed alphabetically. So, the three chemicals you believe to be the safest, should have the three largest doses (you should rank them as 1, 2, and 3), and the more unsafe or dangerous you perceive the chemical to be, the higher the numerical ranking you should give it. After testing yourself with the chemicals listed in Tables 1.1, review the correct answers in tables found at the end of this chapter.] According to the ranking scheme that you selected for these chemicals, were the least potent chemicals common table salt, vitamin K (which is required for normal blood clotting times), the iron supplement dosage added to vitamins for individuals that might be slightly anemic, or a common pain relief medication you can buy at a local drugstore? What were the three most potently toxic chemicals (most dangerous at the lowest single dose) in your opinion? Were they natural or synthetic (humanmade) chemicals? How toxic did you rate the nicotine that provides the stimulant properties of tobacco products? How did the potency ranking of prescription medicines like the sedative phenobarbital or


GENERAL PRINCIPLES OF TOXICOLOGY TABLE 1.1 Cross-Matching Exercise: Comparative Acutely Lethal Doses The chemicals listed in this table are not correctly matched with their acute median lethal doses (LD50’s). Rearrange the list so that they correctly match. The correct order can be found in the answer table at the end of the chapter. A B N 1 2 3 4 5 6 7 8 9 10 11 12 13

LD50 (mg/kg) 15,000 10,000 4,000 1,500 1,375 900 150 142 2 1 0.5 0.001 0.00001

Toxic Chemical

Correct Order

Alcohol (ethanol) Arrow poison (curare) Dioxin or 2,3,7,8-TCDD (PCBs)—an electrical insulation fluid Food poison (botulinum toxin) Iron supplement (ferrous sulfate) Morphine Nicotine Insecticide (malathion) Rat poison (strychnine) Sedative/sleep aid (phenobarbital) Tylenol (acetaminophen) Table salt (sodium chloride)

____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________

the pain killer morphine compare to the acutely lethal potency of a poison such as strychnine or the pesticide malathion? Now take the allowable workplace chronic exposure levels for the following chemicals—aspirin, gasoline, iodine, several different organic solvents, and vegetable oil mists—and again rank these substances going from the highest to lowest allowable workplace air concentration (listed in Table 1.2). Remember that the lower (numerically) the allowable air concentration, the more potently toxic the substance is per unit of exposure. Review the correct answers in the table found at the end of this chapter. Defining Dose and Response Because all chemicals are toxic at some dose, what judgments determine their use? To answer this, one must first understand the use of the dose–response relationship because this provides the basis for estimating the safe exposure level for a chemical. A dose–response relationship is said to exist when changes in dose produce consistent, nonrandom changes in effect, either in the magnitude of effect or in the percent of individuals responding at a particular level of effect. For example, the number of animals dying increases as the dose of strychnine is increased, or with therapeutic agents the number of patients recovering from an infection increases as the dosage is increased. In other instances, the severity of the response seen in each animal increases with an increase in dose once the threshold for toxicity has been exceeded. The Basic Components of Tests Generating Dose–Response Data The design of any toxicity test essentially incorporates the following five basic components: 1. The selection of a test organism 2. The selection of a response to measure (and the method for measuring that response) 3. An exposure period



TABLE 1.2 Cross-Matching Exercise: Occupational Exposure Limits—Aspirin and Vegetable Oil Versus Industrial Solvents The chemicals listed in this table are not correctly matched with their allowable workplace exposure levels. Rearrange the list so that they correctly match. The correct order can be found in the answer table at the end of the chapter. N 1 2 3 4 5 6 7 8 9 10 11

Allowable Workplace Exposure Level Chemical (use) (mg/m3) 0.1 5 10 55 170 188 269 590 890 1590 1910

Aspirin (pain reliever) Gasoline (fuel) Iodine (antiseptic) Naphtha (rubber solvent) Perchloroethylene (dry-cleaning fluid) Tetrahydrofuran (organic solvent) Trichloroethylene (solvent/degreaser) 1,1,1-Trichloroethane (solvent/degreaser) 1,1,2-Trichloroethane (solvent/degreaser) Toluene (organic solvent) Vegetable oil mists (cooking oil)

Correct Order ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________ ____________

4. The test duration (observation period) 5. A series of doses to test Possible test organisms range from isolated cellular material or selected strains of bacteria through higher-order plants and animals. The response or biological endpoint can range from subtle changes in organism physiology or behavior to death of the organism, and exposure periods may vary from a few hours to several years. Clearly, tests are sought (1) for which the response is not subjective and can be consistently determined, (2) that are conclusive even when the exposure period is relatively short, and (3) (for predicting effects in humans) for which the test species responds in a manner that mimics or relates to the likely human response. However, some tests are selected because they yield indirect measurements or special kinds of responses that are useful because they correlate well with another response of interest; for example, the determination of mutagenic potential is often used as one measure of a chemical’s carcinogenic potential. Fortunately or unfortunately, each of the five basic components of a toxicity test protocol may contribute to the uniqueness of the dose–response curve that is generated. In other words, as one changes the species, dose, toxicity of interest, dosage rate, or duration of exposure, the dose–response relationship may change significantly. So, the less comparable the animal test conditions are to the exposure situation you wish to extrapolate to, the greater the potential uncertainty that will exist in the extrapolation you are attempting to make. For example, as can be seen in Table 1.3, the organ toxicity observed in the mouse and the severity of that toxic response change with the air concentration of chloroform to which the animals are exposed. Both of these characteristics of the response—organ type and severity—also change as one changes the species being tested from the mouse to the rat. In the mouse the liver is apparently the most sensitive organ to chloroform-induced systemic toxicity; therefore, selecting an air concentration of 3 ppm to prevent liver toxicity would also eliminate the possibility of kidney or respiratory toxicity. If the concentration of chloroform being tested is increased to 100 ppm, severe liver injury is observed, but still no injury occurs in the kidneys or respiratory tract of the mouse. If test data existed only for the renal and respiratory systems, an exposure level of 100 ppm might be selected as a no-effect level with the assumption that an exposure limit at this concentration would provide complete safety for the mouse. In this case the assumption would be incorrect, and this allowable exposure level would produce an adverse exposure condition for the mouse in the form of severe liver injury. Note also that a safe exposure level for kidney toxicity in the mouse, 100 ppm, would not prevent kidney injury in a closely related species like the rat. This illustrates the problem in assuming that two


GENERAL PRINCIPLES OF TOXICOLOGY TABLE 1.3 Chloroform Toxicity: Inhalation Studies Species Mouse Mouse Mouse Mouse Mouse Mouse Rat Rat Rat Rat Rat Rat

Toxicity of Interest No effect—liver Mild liver damage Severe liver damage No effect—kidneys Mild kidney injury No effect—respiratory No effect—respiratory Nasal injury No effect—kidneys Mild kidney injury No effect—liver Mild liver damage

Duration of Exposure 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days 6 h/day for 7 days

Exposure/Dose (ppm) 3 10 100 100 300 300 3 10 10 30 100 300

Source: Adapted from ATSDR (1996), Toxicant Profile for Chloroform.

similar rodent species like the mouse and rat have very similar dose–response curves and the same relative organ sensitivities to chloroform. For example, an investigator assuming both species have the same dose–response relationships might, after identifying liver toxicity as the most sensitive target organ in the mouse, use only clinical tests for liver toxicity as the biomarker for safe concentrations in the rat. Following this logic, the investigator might erroneously conclude that chloroform concentrations of 100 ppm were completely protective for this species (because no liver toxicity was apparent), although this level would be capable of producing nasal and kidney injury. This simple illustration emphasizes two points. First, it emphasizes the fact that dose–response relationships are sensitive to, and dependent on, the conditions under which the toxicity test was performed. Second, given the variety of the test conditions that might be tested or considered and the variety of dose–response curves that might ultimately be generated with each new test system, the uncertainty inherent in any extrapolation of animal data for the purpose of setting safe exposure limits for humans is clearly dependent on the breadth of toxicity studies performed and the number of different species tested in those studies. This underscores the need for a toxicologist, when attempting to apply animal data for risk assessment purposes, to seek test data where the response is not subjective, has been consistently determined, and has been measured in a species that is known to, or can reasonably be expected to, respond qualitatively and quantitatively the way humans do. Because the dose–response relationship may vary depending on the components of the test, it is, of course, best to rely on human data that have been generated for the same exposure conditions of interest. Unfortunately, such data are rarely available. The human data that are most typically available are generated from human populations in some occupational or clinical setting in which the exposure was believed at least initially, to be safe. The exceptions, of course, are those infrequent, unintended poisonings or environmental releases. This means that the toxicologist usually must attempt to extrapolate data from as many as four or five different categories of toxicity testing (dose–response) information for the safety evaluation of a particular chemical. These categories are: occupational epidemiology (mortality and morbidity) studies, clinical exposure studies, accidental acute poisonings, chronic environmental epidemiology studies, basic animal toxicology tests, and the less traditional alternative testing data (e.g., invertebrates, in vitro data). Each type or category of toxicology study has its own advantages and disadvantages when used to assess the potential human hazard or safety of a particular chemical. These have been summarized in Table 1.4, which lists some of the advantages and disadvantages of toxicity data by category: Part a—occupational epidemiology (human) studies



TABLE 1.4 Some Advantages and Disadvantages of Toxicity Data by Category Advantages

Disadvantages a. Occupational Epidemiology (Human) Studies

May have relevant exposure conditions for the intended use of the chemical As these exposure levels are usually far higher than those found in the general environment, these studies generally allow for a realistic extrapolation of a safe level for environmental exposures

Exposures (especially past exposures) may have been poorly documented Difficult to properly control; many potential confounding influences (lifestyle, concurrent diseases, genetic, etc.) are inherent in most work populations; these potential confounders are often difficult to identify The chance to study the interactive effects of other Post facto—not necessarily designed to be protective chemicals that might be present; again at high doses of health relative to most environmental situations Avoid uncertainties inherent in extrapolating toxicities The increase in disease incidence may have to be large and dose–response relationships across species or the measured response severe to be able to demonstrate the existence of the effect being monitored (e.g., cancer) The full range of human susceptibility (sensitivity) The full range of human sensitivity for the toxicity of may be measurable if sufficiently large and diverse interest may not be measurable because some populations can be examined potentially sensitive populations (young, elderly, infirm) are not represented May help identify gender, race, or genetically Effects must be confirmed by multiple studies as controlled differences in responses heterogeneous populations are examined, and confounders cannot always be excluded The potential to study human effects is inherent in Often costly and time-consuming; cost/benefit may be almost all industrial uses of chemicals; thus, a large low if confounders or other factors limit the range of number of different possible exposure/chemical exposures, toxicities, confounders, or population regimens are available for study variations that might occur with the chemical’s toxicity b. Clinical (Human) Exposure Studies The toxicities identified and the dose–response relationship measured are reported for the most relevant species to study (humans) Typically the components of these studies are better defined and controlled than occupational epidemiology studies The chance to study the interactive effects of other chemicals The dose–response relationship is measured in humans; exposure conditions may be altered during the exposure interval in response to the presence or lack of an effect making NOAELs or LOAELs easier to obtain Better than occupational studies for detecting relatively subtle effects; greater chance to control for the many confounding factors that might be found in occupational studies Allows the investigator to test for and identify possible confounders or potential treatments Allows one to test specific subpopulations of interest May help identify gender, race or genetically controlled differences in responses May be the best method for allowing initial human exposure to the chemical, particularly if medical monitoring is a prominent feature of the study

The most sensitive group (e.g., young, elderly, infirm) may often be inappropriate for study May be costly to perform

Usually limited to shorter exposure intervals than occupational epidemiologic studies Only NOAELs are targeted for study; these studies are primarily limited to examining safe exposure levels or effects of minimal severity; more serious effects caused by the chemical cannot intentionally be examined by this type of study Chronic effects are generally not identifiable by this type of study

Requires study participant compliance May require confirmation by another study May raise ethical questions about intentionally exposing humans to toxicants —




TABLE 1.4 Continued Advantages

Disadvantages c. Environmentally Exposed Epidemiologic Studies

The toxicities identified and the dose–response relationship measured are reported for the most relevant species to study (humans)

Exposures to the chemical are typically low relative to other types of human exposure to the chemical in question, or to chemicals causing related toxicities (e.g., exposure to other environmental carcinogens); thus, attributing the effects observed in a large population may be difficult if many confounding risk factors are present and uncontrolled for in the exposed population Exposure conditions are relevant to understanding or The exposure of interest may be so low that it is preventing significant environmentally caused nontoxic and only acting as a surrogate indicator for health effects from occurring another risk factor that is present but not identified by the study The chance to study the effects of interactive chemicals The number of chemicals with interactive effects may may be possible be numerous and their exposure heavy relative to the chemical of interest; this will confound interpretations of the data The full range of human susceptibility may be present The full range of human susceptibility may not be present, depending upon the study population May allow one to test specific subpopulations of The full complement of relevant environmental interest for differences in thresholds, response rates, exposure associated with the population are not and other important features of the dose–response necessarily identified or considered relationship May help identify gender, race or genetically Large populations may be so heterogeneous in their controlled differences in responses makeup that when compared to control responses, differences in confounders, gender, age, race. etc., may weaken the ability to discriminate real disease associations with chemical exposure from other causes of the disease d. Acute Accidental Poisonings Exposure conditions are realistic for this particular safety extrapolation These studies often provide a temporal description indicating how the disease will develop in an exposed individual Inexpensive relative to other types of human studies

If the exposure is accidental, or related to a suicide, accurate exposure information may be lacking and difficult to determine The knowledge gained from these studies may be of limited relevance to other human exposure situations

Confounding factors affecting the magnitude of the response may be difficult to identify as exposure conditions will not be recreated to identify modifying factors Identifies the target organs affected by high, acute Acute toxicities may not mimic those seen with exposures; these organs may become candidate chronic exposure; this may mislead efforts to targets for chronic toxicity studies characterize the effects seen under chronic exposure situations Requires very few individuals to perform these studies These studies are typically case reports or a small case series, and so measures of individual variations in response may be difficult to estimate The clinical response requires no planning as the These chance observations develop without information gathering typically consists of warning, a feature that prevents the responding to and treating the organ injuries present development of a systematic study by interested as they develop scientists who are knowledgeable about the chemical (continued)



TABLE 1.4 Continued Advantages

Disadvantages e. Animal Toxicity Tests

Easily manipulated and controlled

Best ability to measure subtle responses

Widest range of potential toxicities to study Chance to identify and elucidate mechanisms of toxicity that allow for more accurate risk extrapolations to be made using all five categories of toxicity test data

Cheaper to perform than full-scale epidemiology studies

No risk of producing adverse human health effects during the study

Test species response is of uncertain human relevance; thus, the predictive value is lower than that of human studies Species responses may vary significantly both qualitatively and quantitatively; thus, a number of different species should be tested Exposures levels may not be relevant to (they may far exceed) the human exposure level Selecting the best animal species to study, i.e., the species with the most accurate surrogate responses, is always unknown and is difficult to determine a priori (without a certain amount of human test data); thus, animal data poses somewhat of a catch-22 situation, i.e., you are testing animals to predict human responses to the chemical but must know the human response to that chemical to accurately select the proper animal test species May be a poor measure of the variability inherent to human exposures because animal studies are so well controlled for genetics, doses, observation periods, etc. The reproducibility of the animal response may create a false sense of precision when attempting human extrapolations

Source: Adapted from Beck et al. (1989).

f. Alternatives to Traditional Animal Testing Type of Toxicity Test Structure–activity relationships (SARs) In vitro testing

Alternative animal testing (nonmammalian and nonavian species)



Does not require the use of any Many toxicants with very similar experimental animals chemical properties have very different toxicities Quick to perform Reduces the number of experimental Cannot fully approximate the animals needed complexities that take place in whole organisms (i.e., absorption, Allows for better control of the toxicant distribution, biotransformation, and concentration at the target site elimination) Allows for the study of isolated functions such as nerve-muscle interaction and release of neurotransmitter Easier to control for host factors such as age dependency, nutritional status, and concurrent disease Possible to use human tissue Less expensive and quicker (due to Since the animal is far removed from shorter lifespans) than using higher humans, the effect of a toxicant can animals be very different from that found with higher animals Since a whole organisms is used allows for absorption, distribution, biotransformation, and elimination of the toxicant



Part b—clinical (human) exposure studies Part c—environmentally exposed epidemiology studies Part d—acute accidental poisonings Part e—animal toxicity tests Part f—alternative animal test systems

Frequency-Response and Cumulative-Response Graphs Not only does response to a chemical vary among different species; response also varies within a group of test subjects of the same species. Experience has shown that typically this intraspecies variation follows a normal (Gaussian) distribution when a plot is made relating the frequency of response of the organisms and the magnitude of the response for a given dose (see Figure 1.1a). Well-established statistical techniques exist for this distribution and reveal that two-thirds of the test population will exhibit a response within one standard deviation of the mean response, while approximately 95 and 99 percent, respectively, lie within two and three standard deviations of the mean. Thus, after testing a relatively small number of animals at a specific dose, statistical techniques can be used to define the most probable response (the mean) of that animal species to that dose and the likely range of responses one would see if all animals were tested at that dose (about one or two standard deviations about the



Figure 1.1 (a) When the response of test animals is plotted for a given dose, we see that some may show a minimal effect while others are more affected by the same dose. Plotting the percent of animals showing a particular magnitude of response gives a bell-shaped curve about the mean response. One standard deviation in either direction from the mean should encompass the range of responses for about two-thirds (67 percent) of the animals. Two standard deviations in both directions encompasses 95 percent of the animals. (b) The probable response for a test animal can therefore be easily predicted by testing n animals at a dose. By plotting the average of the n values as a point bracketed by one standard deviation, the probable response of an animal should fall within the area bracketed about the mean at least two-thirds of the time. (c) By plotting the cumulative dose-response (the probable responses for various doses), we generate a curve that is representative of the probable response for any given dose. (d) By plotting the cumulative dose–response curve, using the logarithm of the dose, we transform the hyperbolic shape of the curve to a sigmoid curve. This curve is nearly linear over a large portion of the curve, and it is easier to see or estimate values from this curve.



mean). Typically, a frequency–response curve for each dose of interest is not used to illustrate the dose–response relationship; instead, cumulative dose–response curves are generally used because they depict the summation of several frequency–response curves over a range of different dosages. Graphically, the separate results for each dose are depicted as a point (the average response) with bars extending above and below it to exhibit one standard deviation greater and less than this average response (see Figure 1.1b). A further refinement is then made by plotting the cumulative response in relation to the logarithm of the dose, to yield plots that are typically linear for most responses between 0 and 100 percent, and it is from this curve that several basic features of the dose–response relationship can be most readily identified (see Figures 1.1c,d). In Figure 1.2, a cumulative dose–response curve is featured with a dotted line falling through the highest dose that produces no response in the test animals. Because this dose, and all doses lower than it, fail to produce a toxic response, each of these doses might be referred to as no-observable-effect levels (NOELs), which are useful to identify because they represent safe doses of the chemical. The highest of these NOELs is commonly referred to as the “ threshold” dose, which may be simply defined as the dose below which no toxicity is observed (or occurs). For all doses that are larger than the threshold dose, the response increases with an increase in the dose until the dose is high enough to produce a 100 percent response rate (i.e., all subjects respond). All doses larger than the lowest dose producing a 100 percent response will also produce a 100 percent response and so the curve becomes flat as increasing dose no longer changes the response rate. For therapeutic effects, this region of the dose–response curve is typically the region physicians seek when they prescribe medicines. Because physicians are seeking a beneficial (therapeutic) effect, typically they would select a dose in this region



Figure 1.1 Continued



Figure 1.2 The no effect region is the range of doses that falls below the threshold dose. The threshold is the highest dose which elicits no effect (or the dose below which a response is not observed).

Figure 1.3 (a) The dose-response curve can have a variety of shapes, including line 1, which is linear; line 2, which is sublinear; and line 3, which is supralinear. (b) U-shaped curve representing a dose-response relationship for a chemical with beneficial, as well as adverse effects. At very low doses, a beneficial effect occurs, which is lost with increasing dose. Even higher doses produce a toxic effect. Other variations on the shape of the dose-response curve are possible, depending upon how toxic and beneficial effects are portrayed.



that is just large enough so that individual variations in response to the dose would still result in a 100 percent response, so as to ensure the efficacy of the drug. In contrast, a toxicologist is generally seeking those doses that produce no response because the effect induced by the chemical is an undesirable one. Thus, toxicologists seek the threshold dose and no effect region of the dose–response curve. Before discussing other ways in which dose–response data can be used to assess safety, it will be useful to briefly discuss the various shapes a dose–response curve might take. Although the schematic shapes illustrated in the Figures 1.1 and 1.2 are the most common shapes, the dose–response curve could have either a supralinear or sublinear shape to it (see Figure 1.3). In Figure 1.3, the normal linear sigmoid curve is illustrated by line 1, line 2 is an example of a sublinear relationship, and line 3 depicts a supralinear relationship. In addition, some chemicals, while toxic at high doses, produce beneficial effects at low doses. Graphical presentation of this somewhat more complicated dose-response relationship results in a so-called U-shaped curve (Figure 1.3). The phenomenon of low dose stimulation (e.g., of growth, reproduction, survival, or longevity) and high dose inhibition is termed hormesis, and the most obvious examples of chemicals that exhibit this phenomenon are vitamins and essential nutrients. There are other agents that display hormesis for which the benefit of low doses is less intuitive. For example, a number of studies on animals and humans have suggested that low doses of ionizing radiation decrease cancer incidence and mortality while high doses lead to increased cancer risk. There is some evidence that hormesis may be applicable to a variety of types of chemical toxicants, but a careful assessment of the extent to which this represents a generalized phenomenon has been hampered by the limited availability of response data below the toxic range for most chemicals.

1.4 HOW DOSE–RESPONSE DATA CAN BE USED Dosages are often described as lethal doses (LD), where the response being measured is mortality; toxic doses (TD), where the response is a serious adverse effect other than lethality; and sentinel doses (SD), where the response being measured is a non- or minimally-adverse effect. Sentinel effects (e.g., minor irritation, headaches, drowsiness) serve as a warning that greater exposure may result in more serious effects. Construction of the cumulative dose–response curve enables one to identify doses that affect a specific percent of the exposed population. For example, the LD50 is the dosage lethal to 50 percent of the test organisms (see Figure 1.4), or one may choose to identify a less hazardous dose, such as LD10 or LD01. Dose–response data allow the toxicologist to make several useful comparisons or calculations. As Figure 1.4 shows, comparisons of the LD50 doses of toxicants A, B, and C indicate the potency (toxicity relative to the dose used) of each chemical. Knowing this difference in potency may allow comparisons among chemicals to determine which is the least toxic per unit of dose (least potent), and therefore the safest of the chemicals for a given dose. This type of comparison may be particularly informative when there is familiarity with at least one of the substances being compared. In this way, the relative human risk or safety of a specific exposure may be approximated by comparing the relative potency of the unknown chemical to the familiar one, and in this manner one may approximate a safe exposure level for humans to the new chemical. For toxic effects, it is typically assumed that humans are as sensitive to the toxicity as the test species. Given this assumption, the test dose producing the response of interest [in units of milligrams per kilogram of body weight (mg/kg)], when multiplied by the average human weight (about 70 kg for a man and 60 kg for a woman), will give an approximation of the toxic human dose. A relative ranking system developed years ago uses this approach to categorize the acute toxicity of a chemical, and is shown in Table 1.5. Using this ranking system, an industrial hygienist might get some idea of the acute danger posed by a workplace exposure. Similarly, if chronic toxicity is of greatest concern, that is, if the toxicity occurring at the lowest average daily dose is chronic in nature, combining a measure of this toxic dose (e.g., TD50) and appropriate safety factors might generate an acceptable workplace air concentration for the chemical. Often the dose–response curve for a relatively minor acute toxicity such as odor, tearing, or irritation involves lower doses than more severe toxicities such



Figure 1.4 By plotting the cumulative dose–response curves (log dose), one can identify those doses of a toxicant or toxicants that affect a given percentage of the exposed population. Comparing the values of LD50A to LD50B or LD50C ranks the toxicants according to relative potency for the response monitored.

as coma or liver injury, and much lower doses than fatal exposures. This situation is shown in Figure 1.5, and it can be easily seen that understanding the relationship of the three dose–response curves might allow the use of sentinel effects (represented in Figure 1.5 by the SD curve) to prevent overexposure and the occurrence of more serious toxicities. The difference in dose between the toxicity curve and a sentinel effect represents the margin of safety. Typically, the margin of safety is calculated from data like that shown in Figure 1.5, by dividing TD50 by the SD50. The higher the margin of safety, the safer the chemical is to use (i.e., greater room for error). However, one may also want to use a more protective definition of the margin of safety (for example, TD10/SD50 or TD01/SD100) depending on the circumstances of the substance’s use and the ease of identifying and monitoring either the sentinel response or the seriousness of the toxicity produced. Changing the definition to include a higher percentile of the sentinel dose–response curve (e.g., the SD100) and correspondingly lower percentile of the toxic dose–response curve (e.g., the TD10 or the TD01) forces the margin of safety to be protective for the vast majority of a population.

TABLE 1.5 A Relative Ranking System for Categorization of the Acute Toxicity of a Chemical Probable Oral Lethal Dose for Humans Toxicity Rating or Class 1. 2. 3. 4. 5. 6.

Practically nontoxic Slightly toxic Moderately toxic Very toxic Extremely toxic Supertoxic

Dose (mg/kg) > 15,000 5000–15,000 50–5000 50–500 5–50 <5

For Average Adult > 1 quart 1 pint to 1 quart 1 ounce to 1 pint 1 teaspoonful to 1 ounce 7 drops to 1 teaspoonful < 7 drops

Source: Reproduced with permission of the American Industrial Hygiene Association Journal.



Figure 1.5 By plotting or comparing several dose–response curves for a toxicant, one can see the relationship which exists for several responses the chemical might produce. For example, the sentinel response (SD curve) might represent a relatively safe acute toxicity, such as odor or minor irritation to the eyes or nose. The toxic response (TD curve) might represent a serious toxicity, such as organ injury or coma. The lethal response (LD curve), of course, represents the doses producing death. Thus, finding symptoms of minor toxicity in a few people at sentinel response (SD10) would be sufficient warning to prevent a serious or hazardous exposure from occurring.

Margin of safety =

TD50 SD50

Or redefine it as =

TD01 SD100

Finally, the use of dose–response curves allows for the estimation of the threshold dose or exposure (see Figure 1.2). The threshold is the lowest point on the dose–response curve, or that dose below which an effect by a given agent is not detectable. Thus, all doses, or exposures producing doses, less than the threshold dose should represent safe doses and exposures. As explained in more detail later in this chapter, the safety of extrapolating from the threshold dose is enhanced by dividing it by uncertainty factors, a procedure that is equivalent to selecting a lower dose from the no-effect region of the dose–response curve shown in Figure 1.2.

1.5 AVOIDING INCORRECT CONCLUSIONS FROM DOSE–RESPONSE DATA While the dose–response relationship can be determined for each adverse health effect of a toxicant, one must be cognizant of certain limitations when using these data: 1. If only single values from the dose–response curves are available, it must be kept in mind that those values will not provide any information about the shape of the curve. So, while toxicant A in Figure 1.6 would appear to be more toxic than toxicant B chemical at higher doses, this is not true at lower doses. Toxicant B has a lower threshold and actually begins to cause adverse effects at lower doses than toxicant A. Once someone is exposed to a toxicant, the shape of the dose–response curve may be as important as the dose at which toxicity first begins (the threshold dose). Actually, in this regard toxicant A is of greater concern, not necessarily because of its lower LD50 and LD100 but rather



Figure 1.6 The shape of the dose–response curve is important. By finding the LD50 values for toxicants A and B from a table, one would erroneously assume that A is (always) more toxic than B. The figure demonstrates that this is not true at low doses.

because of its steeper dose–response curve. Once individuals become overexposed (exceed the threshold dose), the increase in response occurs with much smaller increases in dose, and more persons are affected with subsequent increases in dose. In other words, once the toxic level is reached, the margin of error for substance A decreases more rapidly than for substance B, because each incremental increase in exposure greatly increases the percent of individuals affected. 2. Acute toxicity, which is often generated in tests because of the savings in time and expense, may not accurately reflect chronic toxicity dose–response relationships. The type of adverse response generated by a substance may differ significantly as the exposure duration increases in time. Chronic toxicities are often not the same as acute adverse responses. For example, both toluene and benzene cause depression of the central nervous system, and for this acute effect toluene is the more potently toxic of the two compounds. However, benzene is of greater concern to those with chronic, long-term exposure, because it is carcinogenic while toluene is not. 3. There is usually little information for guidance in deciding what animal data will best mimic the human response. For example, a question that often arises initially in the study of a chemical is the following: Is the test species less sensitive or more sensitive than humans? As shown in Table 1.6, the dose of chloroform that is lethal to 50 percent of the test animals (i.e., the LD50) varies depending on the species and strain of animal tested. Estimation of the fatal human dose based on the animal results shown in Table 1.6 would overstate the toxicity of chloroform when using the rabbit or CD-1 mouse data, and underestimate the toxicity of chloroform if projecting lethality using data from the two remaining mouse strains or the two rat strains tested. Unfortunately, there are anatomical, physiological, and biochemical differences among animal species. These differences may confound the animal to human extrapolation by increasing the uncertainty and concern we have for the accuracy of the extrapolation being made. For example, some laboratory animals possess certain anatomical features that humans lack, such as the Zymbal gland and a forestomach. So, when a chemical produces organ toxicity or cancer within these structures, the relevance to humans is unknown. Similarly, male rats produce a protein known as α-2-microglobulin, which has been shown to interact with the metabolites of certain chemicals in a manner that results in repeated cellular injury within the kidney. This reaction is believed to be responsible for the kidney tumors seen in the male rat after chronic exposure to these chemicals. Because this unique protein from these animals does not occur to any appreciable extent in female rats or in mice, kidney tumors are not



TABLE 1.6 Oral LD50 Data for Chloroform Species Rabbit (Dutch Belted) Mouse (CD-1) Human Rat (Sprague–Dawley) Mouse (Swiss) Mouse (ICR-Swiss) Rat (Wistar)

LD50 (mg/kg/day) 100a 250 602 908 1100 1400b 2180

Source: Adapted from ATSDR (1996), Toxicant Profile for Chloroform. a Based on 13 days of dosing. b Female mice.

seen in female rats or male and female mice. From these important sex and species differences, regulatory agencies have concluded the male rat kidney tumors are of limited relevance to humans, a species which is also deficient in α-2-microglobulin. Finally, certain animal strains are uniquely sensitive to certain types of cancer. For example, a large proportion of B6C3F1 mice develop liver tumors before they die, and this sensitivity appears to be due in part to the fact that the H-ras oncogene in this mouse strain is hypomethylated, allowing this oncogene to be expressed more easily, especially during recurrent hepatocellular injury. Similarly, 100 percent of strain A mice typically develop lung tumors before these animals die, and so a chemical that promotes the early development of lung tumors in this strain of mice may not produce any lung tumors in other strains. To summarize, then, there are a number of important species differences that may cause changes in (1) basal metabolic rates; (2) anatomy and structure; (3) physiology and cellular biochemistry; (4) the distribution of chemicals to certain tissues and pharmacokinetics of the chemical in the animal; (5) the metabolism, bioactivation, and detoxification of the chemical; and (6) ultimately the cellular, tissue, or organ response to actions of the chemical at the biochemical, cellular, tissue, or organ level. This problem of species-specific responses to chemicals creates somewhat of a paradox in toxicological research. We use animals as models to study the toxicities of many chemicals; yet, the proper selection of the animal to serve as the test system ideally requires prior knowledge of which animal species most closely resembles humans with respect to the chemical interaction of interest. Thus, the toxicologist is almost always faced with a dilemma. The goal of the toxicologist’s study is the prediction of chemical effects on humans by using animal studies. However, selection of the right animal for that study requires a prior knowledge of the fate and effects of the chemical in humans (the goal), as well as its fate and effects in various animals. Thus, once data are generated in a test species, there are always inherent limitations to extrapolating the observed effects to humans. This is especially problematic when, as sometimes happens, one of the species tested is susceptible to a very undesirable effect, such as cancer or birth defects, yet several other species show no such effects. In that situation, determining or choosing which species represents the human response most accurately has, of course, a great impact on the estimated risk.


Organism-Related Factors Characteristics of the test species or the human population may alter the dose–response curve or limit its usefulness. The following variables should be considered when extrapolating toxicity data:



Route of Exposure The exposure pathway by which a substance comes in contact with the body determines how much of it enters (rate and extent of absorption) and which organs are initially exposed to the largest concentration of the substance. For example, the water and lipid solubility characteristics of a chemical affect its absorption across the lungs (after inhalation), the skin (after dermal application), or the gastrointestinal (GI) tract (after oral ingestion), and the effect differs for each organ. The rate and site of absorption (organ) also may in turn determine the rate of metabolism and excretion of the chemical. So, changing the route of exposure may alter the dose required to produce toxicity. It may also alter the organ toxicity that is observed. For example, the organ with generally the greatest capacity for the metabolism and breakdown of chemicals is the liver. Therefore, a chemical may be more or less toxic per unit of dosage when the chemical is given orally or peritoneally, routes of administration that ensure the chemical absorbed into the bloodstream passes through the liver before it perfuses other organs within the animal. If the capacity of the liver to metabolize the chemical within the bloodstream is great, this leads to what is referred to as a first-pass effect, in which the liver metabolizes a large proportion of the chemical as it is absorbed and before it can be distributed to other tissues. If the metabolism of this chemical is strictly a detoxification process, then the toxic potency of the chemical (i.e., toxicity observed per unit of dose administered) may be reduced relative to its potency when administered by other routes (e.g., intravenously). On the other hand, if the metabolism of that dose generates toxic, reactive metabolites, then a greater toxic potency may be observed when the chemical is given orally relative to inhalation, dermal, or intramuscular administrations of the chemical. (See also discussion in Chapters 2 and 3.) As an illustration that the route of exposure may or may not affect the toxic potency of the chemical, Table 1.7 lists LD50 data for various routes of exposure for three different chemicals. All of these chemicals were administered to the same test species so that differences relating to the route of exposure may be compared. As this table shows, in some instances the potency changes very little with a change in the route of administration (e.g., potency is similar for the pesticide DFP for all routes except dermal), in other instances—DDT, for example—the potency decreases 10-fold when changing the route of administration from intravenous to oral, and another 10-fold when moving from oral to dermal. Sex Gender characteristics may affect the toxicity of some substances. Women have a larger percent of fat in their total body weight than men, and women also have different susceptibilities to reproduction system disorders and teratogenic effects. Some cancers and disease states are sex-linked. Large sex-linked differences are also present in animal data. One well-known pathway for sex-related differences occurs in rodents where the male animals of many rodent strains have a significantly greater capacity for the liver metabolism and breakdown of chemicals (they have more cytochrome P450; see Chapter 3). This greater capacity for oxidative metabolism can cause the male animals of certain rodent strains to be more or less susceptible to toxicity from a chemical depending on whether oxidative

TABLE 1.7 Effect of Route of Administration on Response (LD50)a Route of Administration Oral Subcutaneous Intramuscular Intraperitoneal Intravenous Intraocular Dermal

Methadoneb 90 48 — 33 10 — —

Source: Adapted from Handbook of Toxicology, 1956, Vol. I. a All doses are in units of mg/kg. b Rat. c


Strychnineb 16.2 3 4 1.4 1.1

DDTb 420 1500 — 100 40 — 3000

DFPc 4 1 0.75 1 0.3 1.15 117



metabolism represents a bioactivation or detoxification pathway for the chemical at the dose it is administered. For example, in the rat, strychnine is less toxic to male rats when administered orally because their greater liver metabolism allows them to break down and clear more of this poison before it reaches the systemic circulation. This allows them to survive a dose that is lethal to their female counterparts. Alternatively, this greater capacity for oxidative metabolism renders male rodents more susceptible to the liver toxicity and carcinogenicity of a number of chemicals that are bioactivated to a toxic, reactive intermediate during oxidative metabolism. Age Older people have differences in their musculature and metabolism, which change the disposition of chemicals within the body and therefore the levels required to induce toxicity. At the other end of the spectrum, children have higher respiration rates and different organ susceptibilities [generally they are less sensitive to central nervous system (CNS) stimulants and more sensitive to CNS depressants], differences in the metabolism and elimination of chemicals, and many other biological characteristics that distinguish them from adults in the consideration of risks or chemical hazards. For example, the acute LD50 dose of chloroform is 446 mg/kg in 14-day-old Sprague–Dawley rats, but this dose increases to 1188 mg/kg in the adult animal. Effects of Chemical Interaction (Synergism, Potentiation, and Antagonism) Mixtures represent a challenge because the response of one chemical might be altered by the presence of another chemical in the mixture. A synergistic reaction between two chemicals occurs when both chemicals produce the toxicity of interest, and when combined, the presence of both chemicals causes a greater-than-additive effect in the anticipated response. Potentiation describes that situation when a chemical that does not produce a specific toxicity nevertheless increases the toxicity caused by another chemical when both are present. Antagonists are chemicals that diminish another chemical’s measured effect. Table 1.8 provides simple mathematical illustrations of how the effect of one or two chemicals changes if their combination causes synergism, potentiation, additivity or antagonism, and gives a well-known example of a chemical combination that produces each type of interaction. Modes of Chemical Interaction Chemical interactions can be increased or decreased in one of four ways 1. Functional—both chemicals affect the same physiologic function. 2. Chemical—a chemical interaction between the two compounds affects the toxicity of one of the chemicals. 3. Dispositional—the absorption, metabolism, distribution, or excretion of one of the chemicals is altered by the second chemical. 4. Receptor-mediated—when two chemicals bind to the same tissue receptor, the second chemical, which differs in activity, competes for the receptor and thereby alters the effect produced by the first chemical.

TABLE 1.8 Mathematical Representations of Chemical Interactions Effect Additive Synergistic Potentiation Antagonism

Relative Toxicity (hypothetical) 2+3=5 2 + 3 = 20 2 + 0 = 10 6 + 6 = 8 or 5 + (–5) = 0 or 10 + 0 = 2

Example Organophosphate pesticides Cigarette smoking + asbestos Alcohol + carbon tetrachloride Toluene + benzene or caffeine + alcohol or BAL + mercury



TABLE 1.9 Chemical Interactions with Ethanol Agent Aspirin Barbiturates

Benzene Caffeine

Carbon disulfide

Chloral hydrate

Ethylene glycol


Toxic Interaction

Mode: Mechanism

Increased gastritis

Functional—both agents irritate the GI tract Increased barbiturate toxicity Functional/Dispositional—both agents are CNS depressants; altered pharmacokinetics and pharmacodynamics of the barbiturates Increased benzene-induced hematotoxicity Dispositional—enhanced benzene bioactivation to toxic metabolites Caffeine antagonizes the CNS depressant Functional—both agents affect the CNS, effects of ethanol but one is a stimulant and one is a depressant Dispositional—increased CS2 Enhanced CS2 toxicity bioactivation and retention in critical tissues Increased CNS sedative effects of chloral Functional/dispositional—both agents are hydrate CNS depressants; ethanol also alters the metabolism of chloral hydrate, leading to greater trichloroethanol accumulation Decreased ethylene glycol toxicity Dispositional—ethanol inhibits the metabolism of ethylene glycol to its toxic metabolites Increase in formation of extrahepatic Dispositional—ethanol alters the tissue tumors induced by nitrosamines distribution of nitrosamines by inhibiting hepatic metabolism

Source: Adapted from Calabrese (1991).

To help illustrate the ways in which chemical interactions are increased (additive, potentiation, synergism) or decreased (antagonism), Tables 1.9 and 1.10, adapted from a textbook on chemical interactions by Edward Calabrese, are provided. Table 1.9 summarizes a few of the chemical interactions identified for drinking alcohol (ethanol) and other chemical agents that might be found in home or occupational environments. Like alcohol, smoking may also alter the effects of other chemicals, and the incidence of some minor drug-induced side effects have been reported to be lower in individuals who smoke. For example,

TABLE 1.10 Aquatic Toxicity Interactions between Ammonia and Other Chemicals Chemicals Ammonia + cyanide Ammonia + sulfide Ammonia + copper

Toxic Endpoint

Ammonia + phenol

96-h LC50 24-h LC50 48-h LC50 48-h LC25 48-hr LC10 24-h LC50

Ammonia + phenol + zinc

48-h LC50

Source: Adapted from Calabrese (1991).

Ratio of Chemical EC50s 1 1 1 1 1 1 1 1 1 1

: : : : : : : : : :

1 2.2 1 1 1 0.1 0.7 1 : 0.5 7 : 1 1 : 6

Interaction Additive Antagonism Additive Synergism Synergism Antagonism Additive Additive Synergism Antagonism



smoking seems to diminish the effectiveness of propoxyphene (Darvon) to relieve pain, and it lowers the CNS depressant effects of sedatives from the benzodiazepine and barbiturate families. Smoking also increases certain metabolic pathways in the liver and so enhances the metabolism of a number of drugs. Examples of drugs whose metabolism is increased by smoking include antipyrine, imipramine, nicotine, pentazocine, and theophylline. Table 1.10 summarizes a few of the chemical interactions that have been reported in aquatic toxicity studies. Note that when the same chemicals are present but the ratio of components present in the mixture is changed, the type of interaction observed may change. Genetic Makeup We are not all born physiologically equal, and this provides both advantages and disadvantages. For example, people deficient in glucose-6-phosphate dehydrogenase (G6PD deficiency) are more susceptible than others to the hemolysis of blood by aspirin or certain antibiotics, and people who are genetically slow acetylators are more susceptible to neuropathy and hepatotoxicity from isoniazid. Table 1.11 lists some of the genetic differences that have been identified in humans and some of the agents that may trigger an abnormal response in an affected individual. Health Status In addition to the genetic status, the general well-being of an individual, specifically, their immunologic status, nutritional status, hormonal status, and the absence or presence of concurrent diseases, are features that may alter the dose–response relationship.

Chemical-Specific Factors We have seen that a number of factors inherent in the organism may affect the predicted response; certain chemical and physical factors associated with the form of the chemical or the exposure conditions also may influence toxic potency (i.e., toxicity per unit of dose) of a chemical. Chemical Composition The physical (particle size, liquid or solid, etc.) and chemical (volatility, solubility, etc.) properties of the toxic substance may affect its absorption or alter the probability of TABLE 1.11 Pharmacogenetic Differences in Humans Condition Acatalasia Atypical cholinesterase Acetylation deficiency

Enzyme Affected Catalase—red blood cells Plasma cholinesterase Isoniazid acetylase

Acetophenetidin-induced Cytochrome P450 methemaglobinemia Polymorphic hydroxylation Cytochrome P450 of debrisoquine Polymorphic hydroxylation CYP 2C19 of mephenytoin Glucose-6-phosphate Glucose-6-phosphate dehydrogenase deficiency dehydrogenase

Source: Adapted from Vesell (1987).

Some Chemicals Provoking Abnormal Responses Hydrogen peroxide Succinyl choline Isoniazid, sulfamethazine, procainamide, dapsone, hydralazine Acetophenetidin Encainide, metoprolol, debrisoquine, perphenazine Mephenytoin Hemolytic anemia: aspirin, acetanilide, aminosalicylic acid, antipyrine, aminopyrine, chloroquine, dapsone, dimercaprol, Gantrasin, methylene blue, naphthalene, nitrofurantoin, probenecid, pamaquin, primaquine, phenacetin, phenylhydrazine, potassium perchlorate, quinacrine, quinine, quinidine, sulfanilamide, sulfapyridine, sulfacetamide, trinitrotoluene



exposure. For example, the lead pigments that were used in paints decades ago were not an inhalation hazard when applied because they were encapsulated in the paints. However, as the paint aged, peeled, and chipped, the lead became a hazard when the paint chips were ingested by small children. Similarly, the hazards of certain dusts can be reduced in the workplace with the use of water to keep finely granulated solids clumped together. Exposure Conditions The conditions under which exposure occurs may affect the applied dose of the toxicant, and as a result, the amount of chemical that becomes absorbed. For example, chemicals bound to soils may be absorbed through the skin poorly compared to absorption when a neat solution is applied because the chemical may have affinity for, and be bound by, the organic materials in soil. Concentration, type of exposure (dermal, oral, inhalation, etc.), exposure medium (soil, water, air, food, surfaces, etc.), and duration (acute or chronic) are all factors associated with the exposure conditions that might alter the applied or absorbed dose.

1.7 DESCRIPTIVE TOXICOLOGY: TESTING ADVERSE EFFECTS OF CHEMICALS AND GENERATING DOSE–RESPONSE DATA Since the dose–response relationship aids both basic tasks of toxicologists—namely, identifying the hazards associated with a toxicant and assessing the conditions of its usage—it is appropriate to summarize toxicity testing, or descriptive toxicology. While a number of tests may be used to assess toxic responses, each toxicity test rests on two assumptions: 1. The Hazard Is Qualitatively the Same. The effects produced by the toxicant in the laboratory test are assumed to be the same effects that the chemical will produce in humans. Therefore, the test species or organisms are useful surrogates for identifying the hazards (qualitative toxicities) in humans. 2. The Hazard Is Quantitatively the Same. The dose producing toxicity in animal tests is assumed to be the same as the dose required to produce toxicity in humans. Therefore, animal dose–response data provide a reliable surrogate for evaluating the risks associated with different doses or exposure levels in humans. Which tests or testing scheme to follow depends on the use of the chemical and the likelihood of human exposure. In general, part or all of the following scheme might be required in a descriptive toxicology testing program. Level 1: Testing for acute exposure a. Plot dose–response curves for lethality and possible organ injuries. b. Test eyes and skin for irritation. c. Make a first screen for mutagenic activity. Level 2: Testing for subchronic exposure a. Plot dose–response curves (for 90-day exposure) in two species; the test should use the expected human route of exposure. b. Test organ toxicity; note mortality, body weight changes, hematology, and clinical chemistry; make microscopic examinations for tissue injury. c. Conduct a second screen for mutagenic activity. d. Test for reproductive problems and birth defects (teratology). e. Examine the pharmacokinetics of the test species: the absorption, distribution, metabolism, and elimination of chemicals from the body. f. Conduct behavioral tests. g. Test for synergism, potentiation, and antagonism.



Level 3: Test for chronic exposure a. Conduct mammalian mutagenicity tests. b. Conduct a 2-year carcinogenesis test in rodents. c. Examine pharmacokinetics in humans. d. Conduct human clinical trials. e. Compile the epidemiologic data of acute and chronic exposure. Establishing the safety and hazard of a chemical is a costly and time-consuming effort. For example, the rodent bioassay for carcinogenic potential requires 2–3 years to obtain results, at a cost of between $3,000,000–$7,000,000 and when completed the results, if positive, may in the end severely limit or prohibit the use of the chemical in question. Thus, this final test may entail additional costs if now a replacement chemical must be sought that does not have significant carcinogenic activity. Figure 1.7 outlines the approximate time required to test and develop the safety of chemicals assumed to have widespread human impact.

Figure 1.7 A timeline showing the approximate time that it might take to test a chemical having a broad exposure to the human population. The bars represent the approximate time required to complete the tests and suggest when testing might be initiated and completed.



1.8 EXTRAPOLATION OF ANIMAL TEST DATA TO HUMAN EXPOSURE Several models can be used to extrapolate the human risks from chemical exposure on the basis of toxicity tests in animals. The model chosen is primarily determined by the health hazard of most concern. In the past, however, two basic methods for extrapolation were used. The first type consisted of extrapolating the human risk directly from either the threshold dose or some no-observable-effectlevel (NOEL) dose. This method was applied to most toxicities or health hazards (except cancer), since thresholds were assumed to be present for all of these health hazards. The second type of model was generally used to assess the risk associated with carcinogens. Since the regulatory approach to carcinogens has been to assume that no identifiable threshold exists for this type of toxicity, any exposure was assumed to involve some quantifiable amount of risk. This concept dictated that the mathematical models used to extrapolate to exposures far below the dosages that induce observable responses in the test animal population involve some form of linear extrapolation at low doses. For noncancer-causing toxicants (those with threshold toxicity), the models for extrapolating risk are relatively simple and similar to the methods that have been suggested or used by the National Academy of Sciences (NAS) and various governmental agencies such as the Food and Drug Administration (FDA) or the Environmental Protection Agency (USEPA). These models derive a safe dosage by dividing the threshold (or NOEL/LOEL) by uncertainty factors. The purpose of adding these uncertainty factors is to ensure that the allowable human dose is one that falls within the no-effect region of the human dose–response curve. Basically, this type of calculation assumes that humans are as sensitive as the test species used; so, the amount of a chemical ingested by the test animal that gives no toxic response is considered the safe upper limit of exposure for humans (especially after inclusion of appropriate safety factors).

Calculating Safety for Threshold Toxicities: The Safe Human Dose Approach The calculation of a safe human dose essentially makes an extrapolation on the basis of the size differential between humans and the test species. Usually this is a straightforward body-weight extrapolation, but a surface area scalar for dose could also be used. The calculation is similar to the following:

SHD = where

NOAEL = (mg / kg per day) × 70 kg = N mg/day UF

NOAEL = threshold dose or some other no-observable-adverse-effect-level selected from the no-effect region of the dose–response curve SHD = safe human dose UF = the total uncertainty factor, which depends on the nature and reliability of the animal data used for the extrapolation N = number of milligrams consumed per day (Note: In this example we are extrapolating for an average adult male, and so we have assumed a 70 kg body weight.) Typically, the uncertainty factor used varies from 10 to 10,000 and is dependant on the confidence placed in the animal database as well as whether there are human data to substantiate the reliability of the animal no-effect levels that have been reported. Of course, the number calculated should use chronic exposure data if chronic exposures are expected. This type of model calculates one value, the expected safe human dosage, which regulatory agencies have referred to as either the acceptable daily intake (ADI) or the reference dose (RfD). Exposures which produce human doses that are at or below these safe human dosages (ADIs or RfDs) are considered safe.



Routes of Exposure and the SHD (Safe Human Dose) Once the safe human dose has been estimated, it may be necessary to convert the dose into a concentration of the chemical in a specific environmental medium (air, water, food, soil, etc.) that corresponds to a safe exposure level for that particular route of exposure. That is, while some dose (in mg/day) may be the total safe daily intake for a chemical, the allowable exposure level of that chemical will differ depending on the route of exposure and the environmental medium in which it is found. Exposure by Inhalation Inhalation is usually a major route for occupational exposures and safe levels are determined by the comparison of airborne concentrations to established standards. For converting the safe daily dose into a safe air concentration, the following formula may be used: Dosage = where

(α)(BR)(C)(t) = N mg/kg BW

α = percent of the chemical absorbed by the lungs (if not known, considered to be 100 percent) BR = breathing rate of the individual (which, for a normal worker, can be estimated as 2 h of heavy breathing at 1.47 m3/h or as 6 h of moderate breathing at 0.98 m3/h), depending on the size and physical activity of the individual C = concentration of the chemical in the air (mg/m3) t = time of exposure in hours (usually considered to be 8 h) BW = body weight in kilograms (usually considered to be 70 kg for men and 60 kg for women)

Thus, using the animal data, the preceding formula can be converted to calculate the safe air concentration if the SHD is known: C=

SHD = N mg/m3 (α)(BR)(t)

[Note: SHD = (threshold dosage divided by the uncertainty factor) × BW.] This type of calculation can be used in two important ways:

• To predict a safe occupational airborne concentration for a chemical when there are no established airborne standards

• To compare an established occupational airborne standard (such as the TLV®—the threshold limit value established by the ACGIH—or an OSHA standard) to newly derived animal toxicity data For many environmental exposures it may be assumed that α = 100 percent, and for adults that daily inhalation volume, equal to (BR)(t), is 20–30 m3 for a 24-h period. To calculate an environmental air concentration for a chemical, the safe human daily dose (in units of mg/day) is divided by this total inhalation volume (in units of m3/day). So, the acceptable air concentration (C) mg/m3 = SHD ÷ 20 m3/day (or 30 m3/day). Should it be desirable to express the safe air concentration in parts of toxicant per million parts of air, the value of C [where the air concentration is in units of milligrams per cubic meter of air (mg/m3)] may be converted to a ppm level by the following relationship:

ppm =

C (mg / m3) × 24.5 MW



where MW is the molecular weight of the chemical (g/mol) and 24.5 is the amount (liters) of vapor per mole of contaminant at 25 °C and 760 mm Hg. Example Calculations—Pentachlorophenol Occupational Exposure Guidelines Pentachlorophenol (PCP), a general-purpose biocide, will be used as an example of how to derive various occupational and environmental exposure guideline extrapolations from an estimate of the safe human dosage. A literature review of the noncarcinogenic effects of PCP has shown that the toxicological effect of greatest concern is its teratogenic and fetotoxic effects in test animals. The PCP NOAEL for these effects has been reported to be as great as 5.8 mg/kg daily. Using the formulas shown above, an occupational exposure limit could be calculated as follows: 5.8 mg / kg daily (60 kg) OEL =

100 1.0[(0.98 m3 / h) 6 h / day + (1.47 m3 / h) 2 h / day] OEL =

3.48 mg / day

8.82 m3 / day

= 0.39 mg / m3

where OEL = occupational exposure limit. In this example, an uncertainty factor (UF) of 100 was chosen because there is extensive animal testing data for PCP; a BW of 60 kg was chosen since this type of OEL would be used to protect pregnant women; an α value of 100 percent was chosen because the amount of PCP that may be absorbed through the lungs is not known so this assumption is the most conservative; and the BR value is a standard estimate of the amount of air breathed daily by a worker performing moderately strenuous activity. This calculated level could then become a guideline for evaluating the occupational exposure of females to PCP. In using this approach, dermal exposure was not considered, but it is expected that the proper precautions (e.g., personal protective equipment and strict personal hygiene) could be used to limit these exposure pathways. Another approach to the data would be to rearrange the formula to enable one to compare established OELs to animal toxicity data. Again, using PCP as an example, the ACGIH TLV and OSHA PEL for PCP is an 8-h time-weighted average (TWA) exposure of 0.5 mg/m3. This value can be compared to the animal daily NOAEL of 5.8 mg/kg by solving the following: Calculated daily dose =

[1.0(0.98 m3 / h) 6 h / day + (1.47 m3 / h)2 h / day)]0.5 mg / m3 60 kg

= 0.0735 mg / kg The calculated daily dose of 0.0735 mg/kg can then be compared with the safe human dosage (SHD) of 0.058 mg/kg per day, which was generated by dividing the NOEL dosage of 5.8 mg/kg by a total uncertainty factor (UF) of 100. As one can see, if the present ACGIH TLV® and OSHA PEL are reached or exceeded, an occupationally exposed female may receive a dosage rate that exceeds the calculated or estimated SHD. Additionally, workers handling PCP would likely have some dermal exposure that will add to the daily dose calculation presented here causing the total female worker dosage to be even higher. Environmental Exposure Guidelines A similar approach can be used to set an acceptable ambient-air level (AAAL) or an environmental exposure guideline for other sources of exposure, such as water consumption and ingestion of foodstuffs.



Here again, it may be assumed that α = 100 percent, and that (BR) × (t) is 20 m3 for a 24-h period (the USEPA has recommended this value). Since environmental exposures include a more diverse population than the workplace (e.g., the old, the sick, the young), one might choose to use a UF larger than 100, one possibly as high as 1000. Thus, for a constant daily exposure the formula reduces to AAAL =

SHD 20 m3 / day

= N mg / m3

Again using PCP as an example, the following calculation can be made:


5.8 mg / kg per day 60 kg 1000 20 m3 / day

where AAAL = 1.7 × 10–2 mg/m3 per day, or 17 µg/m3. This value could be used as an acceptable 24-h concentration of PCP in the ambient air. Another approach to establishing an AAAL is to use the estimated permissible concentration (EPC). This approach uses an established OEL and applies two factors: one to take into account the potential increased exposure time for environmental exposures (i.e., 24 h per day for 7 days per week versus 8 h per day for 5 days per week); and an increased UF for the differences in populations between the workplace and the general community. The EPC can be calculated as follows: EPC =




OEL 420

The value of 100 is used as an UF and the 4.2 value is used simply for the increased exposure time of 168 h per week (24 h per day for 7 days per week) versus a 40-h workweek (i.e., 168/40 =4.2). Using the PCP example, the following can be calculated: AAAL =

0.5 mg / m3 = 1.19 × 10−3 mg / m3, or 1.2 µg/m3 420

Both of these approaches could be used for environmental exposures, but the first approach is preferable, assuming that the NOEL data for the most significant adverse effect (in this case, that occurring at the lowest dose) of a chemical are known. For water consumption, one might adopt a 1000-fold UF and assume the average individual ingests 2 L of water per day. In this scenario, the safe water concentration for PCP becomes 174 µg/L or 174 ppb [(5.8 mg/kg per day × 60 kg) ÷ 1,000 = 348 µg/day of PCP is the SHD, which when divided by the water ingestion rate of 2 L/day of water becomes 174 µg/L.)] If the route of environmental exposure to PCP were via the ingestion of food, then the level of PCP considered safe for a particular food item would be dependent on how much of the item is consumed each day. For this example let us assume that the fish ingestion rate is 20.1 g/day for the average fish consumer and 63 g/day for the high-end consumer (assuming this represents the 95th percentile). A safe fish concentration for PCP could be calculated as follows: 1. For the average ingestion rate: 5.8 mg / kg daily 60 kg 1000 Fish concentration = 20.1 g / day Fish concentr ation (for aver age consumption r ate) = 0.0173 mg/g = 17.3 µg/g or 17.3 ppm.



2. For the 95th percentile ingestion rate: 5.8 mg / kg per day 60 kg 1000 Fish concentration = 63 g / day Fish concentration (95th) = 0.0055 mg/g = 5.5 µg/g or 5.5 ppm. For the intake of fruits and vegetables, if we assume a daily mean consumption rate of 5.28 g/kg and a 95th percentile daily consumption rate of 22.44 g/kg, once again, using the PCP as an example, a safe vegetable and fruit concentration mean consumption rate could be calculated as follows: Safe fruit–vegetable concentration = 0.001 mg/g = 1 µg/g = 1 ppm To calculate the safe exposure levels for those individuals consuming fruits and vegetables at the 95th percentile consumption rate we simply divide by 22.44 g/kg per day rather than 5.28 g/kg per day and the safe fruit–vegetable concentration becomes 0.00025 mg/g = 0.25 µg/g = 0.25 ppm.

1.9 SUMMARY Toxicology is a broad scientific field that utilizes basic knowledge of many other scientific disciplines.

• A toxicologist must understand these disciplines in order to discover and examine the variety of adverse effects produced by any toxicant.

• A toxicologist must utilize an understanding of each particular toxicant’s adverse effects, •

and the dose–response curves for these toxicities, to develop either antidotal therapies or guidelines for risk prediction and prevention. A toxicologist uses dose–response relationships as a basic means of identifying the potency and toxicities that determine a chemical’s relative hazards. Ultimately the dose–response curve for the toxicity of greatest concern is used to develop exposure guidelines for the human populations exposed to the chemical. These exposure levels may be dependent on the route of exposure and the perceived sensitivity of the population exposed.

Many types of toxicity tests and different factors can affect the outcome of a test or create uncertainty about its extrapolation to a heterogeneous human population.

• Often the inherent toxicity of a compound cannot be altered; in such cases the only way to lower the risk is to lower the exposure.

• Likewise, when unknown compounds are suspected of posing a hazard, or when our confidence in the estimate of their toxicity is poor, the only way to limit the risk and its liability is to limit exposure.

REFERENCES AND SUGGESTED READING Ballantyne, B., T. C. Marrs, and P. Turner. “ Fundamentals of toxicology,” in General and Applied Toxicology, B. Ballantyne, T. Marrs, and P. Turner, eds., M. Stockton Press, New York, 1993, pp. 3–38. Ballantine, B., “ Exposure-dose-response relationships,” in Hazardous Materials Toxicology: Clinical Principles of Environmental Health, J. B. Sullivan and G. R. Krieger, eds., Williams & Wilkins, Baltimore, 1992, pp. 24–30.



Ballantine, B., and J. B. Sullivan, “ Basic principles of toxicology,” in Hazardous Materials Toxicology: Clinical Principles of Environmental Health, J. B. Sullivan and G. R. Krieger, eds., Williams & Wilkins, Baltimore, 1992, pp. 9–23. Beck, B. D., E. J. Calabrese, and P. D. Anderson, “ The use of toxicology in the regulatory process,” in Principles and Methods of Toxicology, 2nd ed., A. W. Hayes, ed., Raven Press, New York, 1989, pp. 1–28. Calabrese, E. J., in Multiple Chemical Interactions, E. J. Calabrese, ed., Lewis Publishers, Chelsea, MI, 1991, pp. 467–544, 585–600. Deschamps, J., and D. Morgan, “ Information resources for toxicology,” in General and Applied Toxicology, B. Ballantyne, T. Marrs, and P. Turner, eds., M. Stockton Press, New York, 1993, pp. 217–230. Eaton, D. L., and C. D. Klassen, "Principles of toxicology," in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 5th ed., C. D. Klassen, ed., McGraw-Hill, New York, 1996, pp 13–34. Gallo, M. A., “ History and scope of toxicology,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 5th ed., C. D. Klassen, ed., McGraw-Hill, New York, 1996, pp. 3–12. Koeman, J. H., “ Toxicology: History and scope of the field,” in Toxicology: Principles and Applications, R. J. M. Niesink, J. deVries, and M. A. Hollinger, eds., CRC Press, New York, 1996, pp. 2–15. Musch, A., “ Exposure: Qualitative and quantitative aspects,” in Toxicology: Principles and Applications, R. J. M. Niesink, J. deVries, and M. A. Hollinger, eds., CRC Press, New York, 1996, pp. 16–39. Ottobani, M. A., “ How chemicals cause harm,” in The Dose Makes the Poison, M. A. Ottobani, ed., Van Nostrand-Rheinhold, New York, 1991, pp. 19–28. Ottobani, M. A., “ Toxiology—a brief history,” in The Dose Makes the Poison, M. A. Ottobani, ed., Van Nostrand-Rheinhold, New York, 1991, pp. 29–38. Ottobani, M. A., “ Factors that influence toxicity: How much—how often,” in The Dose Makes the Poison, M. A. Ottobani, ed., Van Nostrand-Rheinhold, New York, 1991, pp. 39–54. Ottobani, M. A., “ Other factors that influence toxicity,” in The Dose Makes the Poison, M. A. Ottobani, ed., Van Nostrand-Rheinhold, New York, 1991, pp. 55–68. Rhodes, C., M. Thomas, and J. Athis, “ Principles of testing for acute effects,” in General and Applied Toxicology, B. Ballantyne, T. Marrs, and P. Turner, eds., M. Stockton Press, New York, 1993, pp. 49–88. Sullivan, J. B., and G. R. Krieger, “ Introduction to hazardous material toxicology,” in Hazardous Materials Toxicology: Clinical Principles of Environmental Health, J. B. Sullivan and G. R. Krieger, eds., Williams & Wilkins, Baltimore, 1992, pp. 2–8. Vesell, E. S., “ Pharmacogenetic differences between humans and laboratory animals: Implications for modeling,” in Human Risk Assessment: The Role of the Animal Selection and Extrapolation, V. M. Roloff, ed., Taylor & Francis, 1987, pp. 229–238. Williams, C. A., H. D. Jones, R. W. Freeman, M. J. Wernke, P. L. Williams, S. M. Roberts, and R. C. James, “ The EPC approach to estimating safety from exposure to environmental chemicals,” Regul. Pharmacol. Toxicol. 20: 259–280 (1994). Williams, P. L., “ Pentachlorophenol: An assessment of the occupational hazard,” Am. Ind. Hyg. Assoc. J. 43: 799–810 (1982).


GENERAL PRINCIPLES OF TOXICOLOGY Answers to Table 1.1 Comparative Acutely Lethal Doses Actual Ranking No. 1 2 3 4 5 6 7 8 9 10 11 12 13

LD50 (mg/kg) 15,000 10,000 4,000 1,500 1,375 900 150 142 2 1 0.5 0.001 0.00001

Toxic Chemical PCBs Alcohol (ethanol) Table salt—sodium chloride Ferrous sulfate—an iron supplement Malathion—pesticide Morphine Phenobarbital—a sedative Tylenol (acetaminophen) Strychnine—a rat poison Nicotine Curare—an arrow poison 2,3,7,8-TCDD (dioxin) Botulinum toxin (food poison)

Adapted from Loomis’s Essentials of Toxicology, Fourth Edition, T.A. Loomis and A.W. Hayes, Academic Press, San Diego, CA, 1996.

Answers to Table 1.2 Occupational Exposure Limits: Aspirin and Vegetable Oil Versus Industrial Solvents N 1 2 3 4 5 6 7 8 9 10 11

Allowable Workplace Exposure Level (mg/m3) Chemical (use) 0.1 5 10 55 188 170 269 590 890 1590 1910

Iodine Aspirin Vegetable oil mists (cooking oil) 1,1,2-Trichloroethane (solvent/degreaser) Perchloroethylene (dry-cleaning fluid) Toluene (organic solvent) Trichloroethylene (solvent/degreaser) Tetrahydrofuran (organic solvent) Gasoline (fuel) Naphtha (rubber solvent) 1,1,1-Trichloroethane (solvent/degreaser)

Source: American Conference of Government Industrial Hygienists (ACGIH), 1996.

2 Absorption, Distribution, and Elimination of Toxic Agents ABSORPTION, DISTRIBUTION, AND ELIMINATION OF TOXIC AGENTS


This chapter explains fundamental principles of toxicology, and discusses

• The broad principles that govern transfer of molecules across membranes • The factors that influence absorption of foreign compounds from the GI tract and the lung, and across the skin

• Simple kinetic models that describe disposition (distribution and elimination) • Mechanisms of elimination (biotransformation and excretion) 2.1 TOXICOLOGY AND THE SAFETY AND HEALTH PROFESSIONS Occupational health specialists, including toxicologists, rely upon human and animal data to determine safe exposure levels. If effects observed in workers can be reproduced in a laboratory animal, it becomes possible to investigate the mechanisms that might reasonably be expected to produce such effects. On the other hand, shedding light on the mechanism by which a designated effect is produced in a test animal species may make it easier to find ways to prevent such effects from occurring in humans. Such an understanding may also help to identify subtle or delayed effects that have not been observed in workers, but to which health professionals should be alerted. The uncertainties associated with converting test results in small animals to predictions relevant to humans are frequently stressed. Quantitative differences among species do exist. Occasionally, these differences can be so great that they obscure fundamental similarities. Physiological and biochemical attributes characteristic of a particular species can shift patterns of absorption, distribution, metabolism, excretion, or effect in significant ways. Nonetheless, the basic principles that control these processes are the same for all mammalian species. These principles will be surveyed in this chapter.

Toxic Agents and the Body Figure 2.1 will be referred to several times during this chapter. It is a schematic overview of the behavior of a foreign compound as it enters the body, is distributed into tissues, exerts an effect, and is eliminated. A toxicant is absorbed into the body and then into the blood. From the blood, it is simultaneously eliminated and distributed to various tissues, including the target tissue. The target tissue is the tissue on which the toxicant exerts its effect.

Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.










Tissue Dose Tissue Interaction


Figure 2.1 An overview of the absorption and disposition of a foreign compound. From the blood, the chemical is both eliminated and distributed to the target tissue, where it exerts its effect.

Generally, a toxicant must be considered absorbed in order to have an effect, but this is not always true. Some toxicants are locally toxic or irritating. For example, acid can cause serious damage to the skin even though it is not absorbed through the skin. Although a distinction is made in Figure 2.1 between the target tissue and the central compartment that includes the blood, in some instances the blood itself represents the target tissue. Carbon monoxide, for example, combines with hemoglobin to form carboxyhemoglobin, whose presence in the blood reduces the availability of oxygen to the tissues. Hemolytic agents such as arsine are also active in the blood compartment, and blood is their target tissue. But most often the target tissue is a tissue other than the blood.

Significance of the Target Tissue The target tissue or target organ is not necessarily the tissue in which the toxicant is most highly concentrated. For example, over 90 percent of the lead in the adult human body is in the skeleton, but lead exerts its effects on the kidney, the central and peripheral nervous systems, and the hematopoietic system. It is well known that chlorinated hydrocarbons tend to become concentrated in body fat stores, but they are not known to exert any effects in these tissues. Whether distribution and/or storage processes such as these are actually protective—that is, whether they act to lower the concentration of toxicant at its site of action—is not always clear. There is experimental support for the idea that certain highly localized and specialized sequestration mechanisms, such as incorporation of lead into intranuclear lead inclusion bodies or binding of cadmium to the tissue protein metallothionein, do indeed function as protective mechanisms. Whatever the case with regard to their function, however, the existence of sequestration mechanisms for many compounds means that the bulk movement of a toxicant through the body, or its kinetic behavior as reflected in plasma and tissue concentrations, must be interpreted with care. The concentration or amount of the biologically active form of the toxicant at sites in the target tissue controls the action—the dynamic behavior—of the toxicant.



2.2 TRANSFER ACROSS MEMBRANE BARRIERS Every compound that reaches the systemic circulation and has not been intravenously injected has had to cross membrane barriers. Therefore, the first topic to be considered is the membrane itself and what enables a toxicant to cross it. All membranes are similar in structure. They consist of a phospholipid bilayer, toward the interior of which are positioned the long hydrocarbon or fatty acid tails of the phospholipids, and toward the outside of which are the more polar and hydrophilic portions of the phospholipid molecules. The fatty acid tails align themselves in the interior of the membrane in a formation that is relatively fluid at body temperatures. The polar portions of the phospholipid molecules maintain a relatively rigid outer structure. Proteins embedded throughout the lipid bilayer have specific functions that will be considered later. Molecules can traverse membranes by three principal mechanisms:

• Passive diffusion • Facilitated diffusion • Active transport Passive Diffusion Passive transfer does not involve the participation of any membrane proteins. Two factors determine the rate at which passive diffusion takes place across a membrane: (1) the difference between the concentrations of the chemical on the two sides of the membrane and (2) the ease with which a molecule of the chemical can move through the lipophilic interior of the membrane. Three major factors affect ease of passage: lipid solubility, or lipophilicity; molecular size; and degree of ionization. The Partition Coefficient The lipid solubility of a compound is frequently expressed by its partition coefficient. The partition coefficient is defined as the concentration of the chemical in an organic phase divided by its concentration in water at equilibrium between the two phases. The organic phase is often chloroform, hexane or heptane, or octanol. The partition coefficient is determined by shaking the chemical with water and the organic solvent, and measuring the concentration of the chemical in each phase when equilibrium has been reached. Although the partition coefficient does not have much meaning as an absolute value, it is very useful as an expression of the relative lipophilicities of a series of compounds. It is the rank order that is meaningful in most cases. For example, it has been shown that the partition coefficients of the nonionized forms of several series of representative drugs can be correlated with their rates of transfer across a number of biological membrane systems—from intestinal lumen into blood, from plasma into brain and into cerebrospinal fluid, and from lung into blood. In general, as lipophilicity increases, the partition coefficient increases, and so does ease of movement through the membrane (Table 2.1). Molecular Size The second important feature of a molecule determining ease of movement across a membrane is molecular size. As the cylindrical radius of the molecule increases, with lipophilicity remaining approximately constant, rate of movement across the membrane decreases. This is because the transfer of larger molecules is slowed by frictional resistance and, depending on the structure of the molecule, may also be slowed by steric hindrance. Figure 2.2 illustrates the dependence of the permeability coefficient/partition coefficient ratio on molecular size in a series of lipophilic amides. The ratio would be constant if molecular size were not important. In this set of amides, both molecular size and steric hindrance (the branched-chain forms) are factors in slowing the diffusion of the larger molecules. Very small molecules, in contrast, may move across the membrane more rapidly than would be predicted on the basis of their partition coefficients alone. Small molecules are likely to be more


ABSORPTION, DISTRIBUTION, AND ELIMINATION OF TOXIC AGENTS TABLE 2.1 Partition Coefficients and Rates of Transfer of Selected Drugs from Plasma into Cerebrospinal Fluid of Dogs Heptane–Water

Drug Thiopental Aniline Aminopyrine Pentobarbital Antipyrine Barbital N-Acetyl-4-aminoantipyrine Sulfaguanidine

Partition Coefficient of Nonionized

Half-Life of

Form of Drug

Transfer Process (min)

3.3 1.1 0.21 0.05 0.005 0.002 0.001 < 0.001

1.4 1.7 2.8 4.1 5.8 27 58 230

Source: Adapted from Brodie et al. (1960), Table 2.

water-soluble than their larger homologs. If this is the case, they may be able to move through membrane pores. Pores are features of all membranes. Their size varies with the nature and function of the membrane. Cell membranes will not allow passage of water-soluble molecules larger than about 4 × 10–4 µm in diameter, while blood capillary walls allow passage of water-soluble molecules up to about 30 × 10–4 µm in diameter. Even within this size range of large water-soluble compounds, the rate of transcapillary movement is inversely proportional to molecular radius. Note that the cutoff of 30 × 10–4 µm excludes plasma proteins, so that they are retained within the plasma fluid volume. Degree of Ionization The third important feature of the molecule determining ease of movement through membranes is its degree of ionization. Electrolytes are ionized at the pH values of body fluids. With the exception of very small ionized molecules that can pass through membrane pores, only the nonionized forms of most electrolytes are able to cross membranes. The ionized forms are generally too large to pass through the aqueous pores, and are insufficiently lipophilic to be transferred by passive diffusion. The rate of diffusion therefore will depend not only on the amount of an electrolyte present in the nonionized form but also on the ease with which the nonionized form of the molecule can cross the membrane, that is, on its molecular size and lipophilicity. All ionizable acids and bases have a pKa value related to the dissociation constant. The dissociation constant is always expressed for either acids or bases as an acid dissociation constant, Ka: For acids:

Ka =

(H+)(A−) (HA)

For bases:

Ka =

(H+)(B) (HB+)



Figure 2.2 Dependence of the ratio of permeability coefficient to partition coefficient on the cylindrical radius of the diffusing molecule for diffusion of a series of lipophilic amides across the human red cell membrane. The ordinate scale is in relative, not absolute, values. (P—propionamide; B—butyramide; IB—isobutyramide; V— valeramide; IV—isovaleramide). The partition coefficients are P—0.01; B,IB—0.05; and V,IV—0.175. [Data from Sha’afi et al. (1971).]

The pKa is the negative logarithm of the acid dissociation constant. If, for example, the acid dissociation constant Ka is 10–3, then the pKa is 3. The degree of ionization in body fluids depends on the pH of the medium as well as on the pKa of the acid or base. This relationship can be expressed by the Henderson–Hasselbalch equations: For acids: pKa − pH = log

(nonionized form) (HA) = log (ionized form) (A−)

pKa − pH = log

(ionized form) (HB+) = log (B) (nonionized form)

For bases:

When the pH is equal to the pKa, half of the acid or base is present in the ionized form and half in the nonionized form. At pH values less than the pKa, acids are less completely ionized. At pH greater than the pKa, bases are less completely ionized. Another way of stating these relationships is that at a given



pH, acids having a large pKa (weak acids) are not as fully ionized as strong acids, while bases having a large pKa (strong bases) are more fully ionized (associated with a hydrogen ion) than weak bases. Schanker and his co-workers (Hogben et al., 1959) studied the intestinal absorption of a series of acids and bases in the rat, finding that the percentage of the drug absorbed depended on its pKa and on whether it was an acid or a base. Weak acids and bases, largely nonionized at the intestinal surface pH of 5.3, were readily absorbed, while strong acids (pKa < 3) and strong bases (pKa > 7) were not. When the pH of the intestinal contents was increased from about 4 to about 8 by dissolving the test compounds in strongly buffered solutions rather than in water, the percentage absorption of the acids was decreased and that of the bases was increased.

Facilitated Diffusion The second type of passage across a membrane is facilitated diffusion. Facilitated diffusion requires the participation of a carrier protein molecule. The carrier proteins are a subgroup of the proteins embedded in the membrane lipid bilayer. Because the number of carrier molecules is limited, facilitated diffusion has a maximum rate. It can also be inhibited selectively, competitively by agents that compete for binding sites on the carrier protein because they are structurally similar to the diffusing material, or noncompetitively by agents that affect the carrier in some other way. If diffusion equilibrium is reached, concentrations on both sides of the membrane are equal. In other words, with regard to the fraction of the substance that is absorbed, facilitated diffusion gives the same net results as passive diffusion, but facilitated diffusion is faster. Several carrier protein systems have been studied in reasonable detail. Although many of the proteins incorporated into membranes appear at only one membrane face, all membrane carrier proteins that have been studied are known to span the membrane and to surface at both faces. It is probable that carrier proteins undergo a conformational change, associated with the binding of the diffusing molecule to the carrier, that facilitates their transport across the membrane. The requirement that the diffusing molecule must be able to bind to the carrier protein confers a certain degree of specificity on this mechanism. As might be expected, facilitated mechanisms have evolved to transport essential nutrients across membrane barriers. For example, the transport of essential nutrients such as sugars and amino acids into red blood cells and into the central nervous system takes place by facilitated diffusion.

Active Transport Active transport is the third general process by which molecules can traverse cell membranes. In addition to its requirement for a carrier molecule, active transport requires controlled energy input. It maintains transport against a concentration gradient so that, when equilibrium is reached, the concentrations on the two sides of the membrane are not equal. Adenosine triphosphate (ATP) is the source of the energy required to maintain this concentration gradient. Active transport processes are critical to conservation and regulation of the body’s supply of essential nutrients. These are important functions of the kidney and the liver. Another function of these organs is the excretion of toxic chemicals and their metabolites. Accordingly, there are at least two active transport processes in the kidney for secretion into the urine (one for organic acids and one for organic bases) and at least four in the liver for elimination into the bile (one for acids, one for bases, one for neutral compounds, and one for metals). Other specialized active transport processes are found in the placenta and the intestine, and in the kidney for reabsorption of essential nutrients. Of course, other processes may operate simultaneously with active transport processes. Facilitated diffusion can occur along with active transport, and passive diffusion will take place in the presence of a concentration gradient whenever physical factors are sufficiently favorable.



Specialized Transport Processes Certain specialized processes also transfer molecules across membrane barriers. Phagocytosis takes place in the alveoli of the lung and in the reticuloendothelial system of the liver and spleen. In phagocytosis, the cell membrane surrounds a particle to form a vesicle that detaches itself and moves into the cell interior. For macromolecules such as proteins, phagocytosis is particularly important. It is thought to be biologically significant in interiorizing certain kinds of specialized enzyme systems. Phagocytosis is also important because it is the mechanism used by the alveolar macrophage to scavenge particulates in the alveoli of the lung. A similar process, pinocytosis, is responsible for the cellular internalization of liquids.

2.3 ABSORPTION For most practical purposes, we consider absorption to be absorption into the systemic circulation. Figure 2.3 is a schematic diagram drawn to show that the lungs and the lumen of the gastrointestinal (GI) tract, along with its contents, are exterior to the body. Toxicants may be absorbed from the GI tract, from the lung, or through the skin. In experimental studies, toxicants may also be injected. Injections are commonly given intravenously, intraperitoneally, subcutaneously, or intramuscularly.

Gastrointestinal Tract The GI tract is a very important route for absorption of toxicants. Toxicants may be present in food or drinking water or, if they have been inhaled but are in the form of relatively large particles, they may

Figure 2.3 Schematic diagram of the entry of a chemical into the human body.



have been collected in the nasopharyngeal area and swallowed. Essentially only one cell thick, the epithelial wall of the GI tract is specialized not only for absorption but also for elimination. Absorption from the GI tract is strongly site-dependent, since the pH varies from the very acidic range of about 1–3 in the stomach (depending on the amount and quality of the food and when it was eaten) to around 5–8 in the small intestine and colon (depending on location, food, and intestinal microflora). The intestinal contents can therefore be neutral or even slightly basic. Absorption of Organic Acids and Bases Application of the Henderson-Hasselbalch equation to organic acids, which have pKa values of 3–5, suggests that they should be relatively well absorbed from the acidic pH of the stomach. Salicylic acid is shown as an example in Figure 2.4. Its pKa is about 2. The efficiency of its transfer across the gastric mucosa is dependent on the concentration gradient of the nonionized form across the mucosa as well as on the physical features of salicylic acid that control its rate of diffusion. As Figure 2.4 shows, in the stomach there are 100 nonionized molecules of salicylic acid for every salicylate ion. On the plasma side of the mucosal cell, however, there is relatively little salicylic acid; salicylate ion is overwhelmingly the dominant species. These calculations were carried out for steady-state conditions. In fact, once salicylic acid molecules have entered the plasma, they will be both ionized to a large extent and carried away from the absorption site by the plasma flow. These factors should combine to promote efficient absorption of organic acids from the stomach. However, organic acids are actually absorbed only moderately well in the stomach, perhaps because of its relatively small absorbing surface. Organic bases, in contrast, are largely ionized at the pH of the stomach contents, and so are much more efficiently absorbed from the intestine. Determinants of GI Absorption A number of other factors are important in determining whether, and how rapidly, a compound will be absorbed from the GI tract. The physical factors, such as lipid solubility and molecular size, which determine the rate of diffusion of nonionized species, have already been discussed. Diffusion is also favored by the presence of villi and microvilli in the intestine. These greatly increase the surface area available for diffusion. Thus, even though absorption may not be particularly efficient per unit surface area, the very large total surface area helps to promote intestinal absorption. Facilitated and active transport mechanisms present in the GI tract provide specialized transport for essential nutrients and electrolytes, including sugars, amino acids, sodium, and calcium. A toxicant that mimics the molecular size, configuration, and charge distribution of an essential nutrient sufficiently well may be transported by the carrier process already in place for absorption of that nutrient. Known examples of such mimicry are rare. 5-Fluorouracil has been shown to be absorbed by a pyrimidine transport mechanism. Interaction among metal ions with respect to their use of common transport mechanisms has also been documented. Other factors affect absorption from the GI tract. Compounds that are chemically unstable at the acid pH of the stomach will not even reach the intestine to be absorbed there. Other compounds are

Figure 2.4 Partitioning of salicylic acid across the gastric mucosa. The numbers in parentheses are the numbers of molecules of the ionized or nonionized species present on either side of the membrane. (Reproduced with permission from O’Flaherty, 1981, Figure 2.9.)



susceptible to alteration by the actions of intestinal microflora, which are important for digestion of plant materials resistant to the action of mammalian enzymes. Enzyme systems of the intestinal wall and/or the liver may metabolize chemicals before they reach the systemic circulation, the intestinal and/or hepatic first-pass effect, which can result in significant reduction in bioavailability. For example, compared to its 100 percent availability on intravenous injection, the systemic bioavailability in rats of buprenorphine, an opiate analgesic, was found to be 49 percent when the drug was given intrahepatoportally and 10 percent when it was given intraduodenally. It can be calculated that after 80 percent of the intraduodenal dose had been inactivated in the intestine, half of the surviving 20 percent was inactivated in the liver. Another determinant of gastrointestinal absorption is the rate at which foodstuffs pass through the GI tract. If the rate of passage is slowed, the length of time during which the compound is available for absorption is increased. Absorption also tends to increase during short periods of fasting but may fall off after a lengthy fast, probably consequent to a decrease in intestinal blood flow. Other important influences on absorption include the chemical and physical characteristics of the compound, its solubility under the conditions present in the GI tract, and its interactions with other compounds. Age and nutritional status of the individual may also affect absorption from the GI tract.

Skin The second major pathway for absorption is the skin. The skin is a very effective barrier to absorption, primarily because of the outermost keratinized layer of thick-walled epidermal cells, the stratum corneum, which in general is not very permeable to toxicants, although its permeability varies from location to location. Compared with the total thickness of the epidermis and dermis together, the thickness of the stratum corneum is relatively slight, but this barrier is rate-limiting in the process of absorption through the skin. There may be slight absorption through sweat glands or hair follicles, but these structures represent a very small percentage of the total surface area and are not ordinarily important in the process of dermal absorption. All toxicants that penetrate the skin appear to do so by passive diffusion. Lipophilic chemicals are much better absorbed through the skin than are hydrophilic chemicals, and the ease with which a compound penetrates the skin is correlated with its partition coefficient. Dermal absorption can be increased in various ways. An increase in capillary blood flow, as in response to the demand of a warm environment for efficient heat loss, is associated with increased percutaneous absorption. Abrasion, which damages or removes the stratum corneum, greatly increases the permeability of the damaged area. The skin is normally partially hydrated; an increase in the degree of hydration increases permeability and promotes absorption. Certain solvents, such as dimethyl sulfoxide (DMSO), also increase skin permeability and facilitate absorption of toxicants. Lipophilic drugs that would suffer extensive first-pass metabolism if given orally can be administered dermally. The glyceryl trinitrate patch used in treatment and prevention of angina is a good example. Certain toxicants can produce systemic injury by percutaneous absorption. Hydrocarbon solvents, such as hexane, can produce a peripheral neurotoxicity, and carbon tetrachloride can produce liver injury. Organophosphate insecticides such as parathion and malathion have caused toxicity and deaths in industrial and field workers after absorption through the skin.

Lung The third major site of toxicant absorption is the lung. In occupation-linked toxicology, the lung is a very important route of uptake. Gases and vapors such as carbon monoxide, sulfur dioxide, and volatile hydrocarbons are absorbed through the lung, and liquid or particulate aerosols, such as sulfuric acid aerosols or silica dust, are also deposited and/or absorbed in the lung. With solid and liquid particulates, the site of deposition is critical to the degree of absorption of a compound.



Solid and Liquid Particulates The lung can be thought of as consisting of three basic regions: the nasopharyngeal region, the tracheobronchiolar region, and the distal or alveolar region. Particles that are roughly 5 µm or greater in diameter are generally deposited in the nasopharyngeal region. If they are deposited very close to the surface, they can be sneezed out, blown out, or wiped away. If they are deposited slightly farther back, they may be picked up by the mucus-blanketed cilia lining the lung in this region (the mucociliary “ escalator” ) and moved back up into the nasopharyngeal region, where they may be swallowed and absorbed in the GI tract in accordance with their solubility and absorption characteristics. Particles that fall into the size range of 2–5 µm generally reach the tracheobronchial region before they impact the lung surface. Most of these particles are also cleared by the mucociliary escalator back up to the nasopharyngeal region, where they are either eliminated directly or swallowed and absorbed or excreted in the GI tract. Particles smaller than 1 µm in diameter may reach the alveolar regions of the lung. Absorption in the lung, if it takes place at all, will most likely take place in the alveolar region, although there may be some absorption in the tracheobronchiolar region, particularly if the material is soluble in the mucus. Size is probably the most important single characteristic determining the efficiency of particulate absorption in the lung. Size determines the region of the lung in which the aerosol is likely to be deposited. Even within the range of very small particles that reach the alveolar region and may be absorbed there, size is inversely proportional to the magnitude of particle deposition. Figure 2.5 shows the dependence of lead deposition in the human lung on the size of the lead particles in an artificially generated lead sesquioxide aerosol that was inhaled by the subjects. Size is expressed as diffusion mean equivalent diameter (DMED), a measure of mean particle diameter. The amount deposited was calculated as the difference between the amount of lead that entered the lung and the amount that the subject exhaled. Thus, the lung regions in which the particles were actually deposited were not identified. However, the DMEDs for all three aerosols were less than 1 µm. For a standard subject, with a breathing cycle of about 4 s, the lung deposition of lead varied from about 24 percent for particles with a DMED of 0.09 µm to 68 percent for very small particles with a DMED of

Figure 2.5 Deposition in the human lung of lead particles of various sizes: • DMED, 0.02 µm; " DMED, 0.04 µm; H DMED, 0.09 µm. Data from Chamberlain et al., 1978, Table 5.2.



0.02 µm. Note that deposition also depends on the breathing cycle. Slow, deep breathing was associated with greater percentage deposition of lead in each particle size range. After being deposited in the alveolar region, particulates may be dissolved and absorbed into the bloodstream, reaching the systemic circulation directly. If they are not readily soluble, they may be phagocytized by alveolar macrophages and then either transferred to the lymphatic system, where they may remain for a considerable time period, or moved together with the macrophage to the mucociliary escalator for clearance by that route. They may also occasionally remain in the alveolus for an extended period of time. Absorption of particulates tends to be slower than absorption of gases and vapors, and appears to be controlled primarily by the solubility of the particulate. For example, the systemic bioavailability of chromium(VI) and nickel salts from the lung has been shown to parallel their solubility. The absorption of water-soluble chemicals in the lung, and even in the nasal cavity, can be quite high. For example, aspirin was found to be 100 percent bioavailable from the rat nasal cavity but only 59 percent bioavailable when given orally. Nicotine was fully absorbed from intratracheal, bronchial, and distal sites in the dog lung, although absorption was not equally rapid from all three sites. Gases and Vapors Absorption of gases and vapors in the lung depends on their solubility in the blood perfusing the lung. Very soluble compounds will be almost completely cleared from inhaled air and transferred to pulmonary blood in a single respiration. For such compounds, increasing the rate of pulmonary blood flow makes very little difference in absorption rate. The only way to increase absorption is to increase the rate of respiration; that is, to increase ventilation. Absorption of these compounds is said to be ventilation-limited. If they are also lipid-soluble, they will find their way rapidly to the lipid depots of the body. Chloroform is a good example of such a compound. It is very highly lipid-soluble, and is readily cleared from inspired air. As the blood circulates through the body, the chloroform is transferred to fat, so that the blood is also effectively cleared and during its next pass through the lung is able to pick up more chloroform. The absorption of chloroform is ventilation-limited. For poorly soluble gases, the capacity for absorption is rather limited. Little of the compound will be transferred to pulmonary blood in a single respiration. Often these are not compounds that are readily cleared from the blood. If this is the case, the blood will become saturated quickly, and the only way to increase absorption then is to increase the rate of pulmonary blood flow. Such compounds are said to be flow-limited in their absorption characteristics. Of course, there is a range of transition between these two extremes of pulmonary absorption behavior.

2.4 DISPOSITION: DISTRIBUTION AND ELIMINATION Unlike absorption, disposition consists not just of one kind of process but, rather, of a number of different kinds of processes taking place simultaneously. Disposition includes both distribution and elimination, which occur in parallel in almost all cases (Figure 2.1). Elimination is also made up of two kinds of processes, excretion and biotransformation, which usually take place simultaneously. Distribution and elimination are often considered independently of each other. While it is convenient to separate them for discussion, it is important to remember that they are taking place at the same time. If a substance is effectively excreted, it will not be distributed into peripheral tissues to any great extent. On the other hand, wide distribution of the compound may impede its excretion. Kinetic Models Before discussing some of the specific mechanisms for distribution, excretion, and biotransformation, it is useful to consider some simple kinetic disposition models. Rates of distribution are related to kinetic distribution constants. Rates of metabolism, or biotransformation, and of excretion are related to kinetic constants of elimination. It is possible to integrate all the essential information about



distribution, biotransformation, and excretion of a chemical into a single kinetic model. Such models can be used to help formulate predictions about the kinetic behavior of a toxicant under different exposure conditions and, if used carefully and in conjunction with effect data, can sometimes also help to suggest a toxicant’s site of action or mechanism of action. Kinetic models are finding increasing use in the interpretation of human exposure and response experience through an understanding of the dose–distribution/action–effect sequence gained from animal studies. In this connection, they are essential to the growing field of human health risk assessment. Kinetic models can be assigned to two general groups: classical models and physiologically based models. The classical descriptive pharmacokinetic models were developed largely by and for the pharmaceutical industry. Their applications are often to situations in which at least some human data are available. Physiologically based models have been intensively developed and used in recent years by the toxicology community. These are based on actual anatomic and physiologic characteristics of the species and on physicochemical and metabolic characteristics of the chemical, and so lend themselves to the cross-species extrapolations with which the toxicologist must frequently deal. Classical Kinetic Models The simplest classical kinetic model is the one-compartment open model (Figure 2.6), in which the compound is assumed to have been introduced instantaneously into the body, distributed instantaneously and homogeneously, and eliminated at a rate that is at all times directly proportional to the amount left in the body, that is, a first-order rate. The constant of proportionality between the rate of elimination and the amount in the body is the elimination rate constant (ke), which has dimensions of reciprocal time, or time–1. For this model, the logarithm of concentration in the blood is a linear function of time, as shown in Figure 2.7. The elimination rate constant is the absolute value of the slope of this line. The half-life is the time required for half the compound to be cleared from the plasma or, in this simple model, from the body. The nature of the half-life is such that it can be measured at any point on a concentration–time curve. Thus, it is a constant value, independent of dose, that characterizes the kinetic behavior of the chemical. The half-life is inversely proportional to the elimination rate constant: t1 / 2 = ln2 / ke. Another useful concept derived from classical kinetics is the clearance. In first-order kinetics, clearance is defined as the rate constant times the volume of distribution. Therefore, clearance has dimensions of volume per unit time, or flow rate. It represents a volume of fluid cleared of the chemical per unit time by metabolism or excretion. Few chemicals obey simple first-order one-compartment kinetics. Most compounds require at least a two-compartment model (Figure 2.8). In the two-compartment model, the chemical is assumed to enter the first or central compartment, which includes the blood; to be distributed instantaneously and homogeneously throughout this compartment; and then to be subject to the parallel processes of elimination and distribution to a second or peripheral compartment from which the chemical can return to the central compartment. Concentration in the central compartment declines smoothly as a function of time. Concentration in the peripheral compartment rises, peaks, and subsequently declines (Figure 2.9). Assuming first-order kinetics, a half-life may also be calculated for a compound whose kinetic behavior fits a two-compartment model or, for that matter, a model with any number of compartments.

Figure 2.6 The linear one-compartment open model. C(t) is the concentration, which is a function of time; and ke is the elimination rate constant. (Reproduced with permission from O’Flaherty, 1981, Figure 2.12.)



Figure 2.7 Plot of the logarithm of the concentration versus time for the linear one-compartment open model. C0 is the concentration at time t = 0, assuming instantaneous distribution. (Reproduced with permission from O’Flaherty, 1981, Figure 2.15a.)

Calculated from the terminal slope of a plot of the natural logarithm of the concentration in the central compartment as a function of time, this half-life is designated the biological half-life. It is the parameter most frequently used to characterize the in vivo kinetic behavior of an exogenous compound. Other features of chemical kinetic behavior or of mode of administration may be incorporated into the model as appropriate. For example, there may be more than one peripheral tissue compartment, as in Figure 2.1; or absorption, which is never truly instantaneous even for intravenous injection, may be first-order instead. An oral exposure, in which the rate of absorption is usually considered to be directly proportional to the amount remaining available in the GI tract, is an example of first-order uptake. The important group of models that incorporate non-first-order kinetics should also be mentioned. Absorption and distribution are conventionally considered to be passive, first-order processes unless observation dictates otherwise. However, elimination often is not first-order. Frequently this is because excretion or metabolism is saturable, or capacity-limited, due to a limitation on the maximum number of active transport sites in organs of excretion or the maximum number of active sites on metabolizing enzymes. When all active elimination sites are occupied, the elimination process is said to be saturated. Kinetically it is a zero-order process, operating at a constant maximum rate independent of the amount or concentration of the chemical in the body. At very low concentrations at which relatively few elimination sites are occupied, capacity-limited kinetics reduces to pseudo-first-order kinetics. Capacity-limited kinetics is often referred to as Michaelis–Menten kinetics, after the authors of an early paper analyzing and interpreting this type of kinetic behavior. Classical kinetic models incorporating Michaelis–Menten elimination have been developed.



Figure 2.8 The linear two-compartment open model, where C1 and C2 are the concentrations in the central and peripheral compartments, respectively, and k12 and k21 are the rate constants for transfer between the two compartments. (Reproduced with permission from O’Flaherty, 1981, Figure 2.22.)

Most industrial or environmental exposures are not acute. Acute exposures do occur, but chronic exposures are much more frequent in both industrial and environmental settings. When exposure is approximately constant and continuous over a long period of time (e.g., if a contaminant is widely dispersed in ambient air), a steady state or “ plateau” level will eventually be reached in all tissues. As long as elimination processes remain first-order (typical, e.g., of excretion by glomerular filtration in

Figure 2.9 Plot of the logarithm of the concentration versus time for the linear two-compartment open model, showing ln C as a function of time for the central (C1) and peripheral (C2) compartments. (Reproduced with permission from O’Flaherty, 1981, Figure 3-24b.)



the kidney, or of loss of a volatile chemical in expired air), this steady state should be directly proportional to both the magnitude of exposure and the biological half-life. If exposure were truly constant, the plateau level would be constant also. More commonly, exposure is intermittent, in which case blood concentrations at steady state will cycle in a way that reflects the absorption and elimination characteristics of the compound as well as the exposure pattern (Figure 2.10). However, on a larger timescale this cycling will take place about a constant mean that is predictable from the equivalent constant exposure rate and the biological half-life. This is one of the reasons why biological half-life is such an important attribute. Together with exposure rate, it determines mean steady-state blood level irrespective of whether exposure is continuous or intermittent. However, the individual exposed to large amounts of a substance at wide intervals will experience greater peak concentrations in blood and tissues following each new exposure than will an individual exposed to the same total amount as frequent small exposures. If the large peak concentrations are associated with toxicity or with saturation of elimination processes, then it becomes important to consider the pattern of administration as well as the equivalent mean exposure rate. Physiologically Based Kinetic Models Physiologically based kinetic (PBK) models are simplified but anatomically and physiologically reasonable models of the body. Tissues are selected or grouped according to their perfusion (blood flow) characteristics and whether they are sites of absorption or elimination (by excretion or metabolism). The model design process is facilitated by reference to compilations of anatomic and physiologic data, including tissue and organ perfusion rates, that are now widely available. Within this general structural framework, the kinetic behavior of the selected chemical is modeled. A key question is how the chemical is taken up into tissues. When flow-limited kinetics are assumed, the chemical is presumed to be in equilibrium between each tissue group and the venous blood leaving

__ Figure 2.10 The relationship between average concentration C(n), calculated for repetitive administation, and the time course of concentration change during continuous administration of a hypothetical compound. Cmax and Cmin are the maximum and minimum concentrations in each time interval between doses, assuming instantaneous distribution of each successive dose. (Reproduced with permission from O’Flaherty, 1981. Figure 5-4.)



the tissue. This equilibrium will vary from tissue to tissue and may also vary from species to species. Simple partitioning phenomena, such as into body lipid stores, can be described by defining partition coefficients, whose values can be determined experimentally at steady state in vivo or in vial equilibration experiments in vitro. More complex partitioning, such as capacity-limited binding of a metal to specific binding sites in tissues, must be defined appropriately. Estimates of dissociation constants may be required. Diffusion-limited kinetics can also be accommodated within the framework of PBK models. In diffusion-limited kinetics, the process of transfer across the membrane separating tissue from blood is the rate-limiting step in tissue uptake. The distinction between flow-limited and diffusion-limited tissue-uptake kinetics is roughly analogous to the distinction between ventilation-limited and flowlimited absorption in the lung. The metabolism of the compound must also be known. Metabolic parameters are more likely than anatomic or physiologic parameters to be species-specific or even tissue-specific. The differences may be quantitative or qualitative. Capacity-limited metabolism, absorption, and/or excretion can be incorporated into PBK models as needed. Figure 2.11 is a schematic diagram of a PBK model that might be designed for a volatile lipophilic chemical. Arrows designate the direction of blood flow, with arterial blood entering the organs and tissue groups and mixed venous blood returning to the lung to be reoxygenated. Organs of entry (lung, liver), excretion (kidney, intestine, lung), and metabolism (liver), and tissue of accumulation (fat) for this chemical class are explicitly included in the model. Other tissues are lumped into well-perfused and poorly perfused groups. Note that uptake into the liver is considered to take place both by way of the portal vein coming from the intestine and by way of the hepatic artery. An enterohepatic recycling


Excretion Lung


Well-perfused Tissues

Poorly-perfused Tissues Metabolism Liver

Intestine Excretion Kidney Excretion Figure 2.11 Schematic diagram of a physiologically-based model of the kinetic behavior of a volatile chemical compound.



between liver and intestine is also included in the model. These features of the model are choices made by the model developer, and reflect the known physicochemical behavior of the agent whose kinetics are being modeled. Models for other chemicals will be quite different. A model for a nonvolatile chemical would not include an explicit lung compartment, while models for bone-seeking elements like lead and uranium include bone as a distinct tissue. In a sense, classical and PBK models work in opposite directions. In classical descriptive kinetics, model compartments having no necessary relationship to actual tissue volumes and clearances having no necessary relationship to tissue blood flow are inferred from a set of concentration data. In contrast, the PBK model is constructed from basic anatomic, physiologic, physicochemical, and metabolic building blocks. It is then used to simulate concentrations under a defined set of conditions, and its predictions are compared with observations. If the predictions are not accurate, some premise of the model is at fault. The need for model revision can afford insight into the processes that control the kinetic behavior of the chemical. A PBK model for dichloromethane (DCM) forms the basis of a current human health risk assessment. DCM is metabolized by two pathways, a capacity-limited oxidative pathway and firstorder conjugation with glutathione (for descriptions of these biotransformation processes, see Chapter 3). Either pathway was thought potentially capable of generating reactive intermediates involved in the tumorigenicity of DCM in mice. Andersen et al. (1987) demonstrated that tumorigenicity correlated well with the activity of the glutathione pathway, but not with the activity of the oxidative pathway. These investigators scaled a PBK model developed for DCM from mouse to human and from high dose to low dose in order to predict, based on studies carried out at high doses in mice, the risk associated with human environmental exposure to DCM. The mouse-to-human scaling of metabolism relied on experimentally-determined human metabolic parameter values. Their physiologic foundation and the inclusion of species-specific physiologic and metabolic mechanisms, when these are known, confer on PBK models a flexibility that allows their use for route-to-route, dose-to-dose, and species-to-species extrapolations such as this one, for which classical models would be wholly inappropriate. Biotransformation Biotransformation is one of the two general elimination mechanisms. Biotransformation reactions in general can be divided into two classes: phase I and phase II reactions. Phase I reactions are catabolic or breakdown reactions (oxidation, reduction, and hydrolysis) that generate or free up a polar functional group. They produce metabolites that may be excreted directly or may become substrates for phase II reactions. Phase II reactions, which are often coordinated with phase I activity, are synthetic reactions in which an additional molecule is covalently bound to the parent or the metabolite, which usually results in a more water-soluble conjugate. Biotransformation reactions, and the factors that influence them, are discussed in detail in Chapter 3. Excretion Excretion takes place simultaneously with biotransformation and, of course, with distribution. The kidney is probably the single most important excretory organ in terms of the number of compounds excreted, but the liver and lung are of greater importance for certain classes of compounds. The lung is active in excretion of volatile compounds and gases. The liver, because it is a key biotransforming organ as well as an organ of excretion, is in a unique position with regard to the elimination of foreign chemicals. Excretion in the Kidney About 20 percent of all dissolved compounds of less than protein size are filtered by the kidney in the glomerular filtration process. Glomerular filtration is a passive process; it does not require energy input. Filtered compounds may be either excreted or reabsorbed. Passive reabsorption in the kidney, as elsewhere, is a diffusion process. It is governed by the usual principles.



Thus, lipid-soluble compounds are subject to reabsorption after having been filtered by the kidney. The degree of reabsorption of electrolytes will be strongly influenced by the pH of the urine, which determines the amount of the chemical present in a nonionized form. It is to be expected that some control could be exerted over the rate of excretion of weak acids and bases by adjusting urine pH. This type of treatment can be used very effectively in some cases. Alkalinization of the urine by administration of bicarbonate has been used to treat salicylic acid poisoning in humans. Alkalinization causes the weak acid to become more fully ionized; the ionized molecule is excreted in the urine rather than reabsorbed. There are also active secretory and reabsorptive processes in the renal tubules of the kidney. These processes are specialized to handle endogenous compounds; active reabsorption helps to conserve the essential nutrients, glucose and amino acids. These pathways can also be used by exogenous compounds, provided the compounds have the structural and electronic configurations required by the carrier molecules. The renal clearance represents a hypothetical plasma volume cleared of solute by the net action of all renal mechanisms during the specified period of time. A compound such as creatinine that is filtered but not secreted or reabsorbed is cleared in adult humans at a rate of about 125 mL/min. Compounds that are reabsorbed as well as filtered have clearances less than the creatinine clearance. Compounds that are actively secreted can have clearances as large as the renal plasma flow, about 600 mL/min. The presence of disease in the kidney can affect the half-life of a compound eliminated via the kidney, just as the presence of disease in the liver can affect the half-life of a compound that is largely biotransformed. Excretion in the Liver The liver is both the major metabolizing organ and a major excretory organ. Large fractions of many toxicants absorbed from the gastrointestinal tract are eliminated in the liver by metabolism or excretion before they can reach the systemic circulation, the hepatic first-pass effect. In addition, metabolites formed in the liver may be excreted into the bile before they themselves have had a chance to circulate. Although it does not excrete as many different compounds as the kidney does, the liver is in an advantageous position with regard to excretion, particularly of metabolites. There are at least three active systems for transport of organic compounds from liver into bile: one for acids, one for bases, and one for neutral compounds. Certain metals are also excreted into bile against a concentration gradient. These transport processes are efficient and can extract protein-bound as well as free chemicals. The characteristics that determine whether a compound will be excreted in the bile or in the urine include its molecular weight, charge, and charge distribution. In general, highly polar and larger compounds are more frequently found in the bile. The threshold molecular weight for biliary excretion is species-dependent. In the rat, compounds with molecular weights greater than about 350 can be excreted in the bile. Those having molecular weights greater than about 450 are excreted predominantly in the bile, while compounds with molecular weights between 350 and 450 are frequently found in both urine and bile. Once a compound has been excreted by the liver into the bile, and thereby into the intestinal tract, it can either be excreted in the feces or reabsorbed. Most frequently the excreted compound itself, being water-soluble, is not likely to be reabsorbed directly. However, glucuronidase enzymes of the intestinal microflora are capable of hydrolyzing glucuronides, releasing less polar compounds that may then be reabsorbed. The process is termed enterohepatic circulation. It can result in extended retention of compounds recycled in this manner. Techniques have been developed to interrupt the enterohepatic cycle by introducing an adsorbent that will bind the excreted chemical and carry it through the gastrointestinal tract. Certain factors influence the efficiency of liver excretion. Liver disease can reduce the excretory as well as the metabolic capacity of the liver. On the other hand, a number of drugs increase the rate of hepatic excretion by increasing bile flow rate. For example, phenobarbital produces an increase in bile flow that is not related to its ability to induce metabolizing enzymes. Whether the increased rate of bile flow will increase the rate of elimination of a compound that is both metabolized and excreted by the liver depends on whether the rate-limiting step is the enzyme-catalyzed biotransformation or



the transfer from liver to bile. If transfer from liver to bile is the rate-limiting step, enhancement of the rate of bile flow will enhance the rate of excretion.

Excretion in the Lung The third major organ of elimination is the lung, the key organ for the excretion of volatile chemical compounds. Pulmonary excretion, like pulmonary absorption, is by passive diffusion. For example, the rate of transfer of chloroform out of pulmonary blood is directly proportional to its concentration in the blood. Essentially, pulmonary excretion is the reverse of the uptake process, in that compounds with low solubility in the blood are perfusion-limited in their rate of excretion, whereas those with high solubility are ventilation-limited. Highly lipophilic chemicals that have accumulated in lipid depots may be present in expired air for a very long time after exposure.

Other Routes of Excretion Skin, hair, sweat, nails, and milk are other, usually minor routes of excretion. Hair can be a significant route of excretion for furred animals, and indeed the amount of a metal in hair, like the amount of a volatile compound in exhaled air, can be used as an index of exposure in both laboratory animals and humans. Hair is not quantitatively an important route of excretion in humans, however. Sweat and nails are only rarely of interest as routes of excretion, simply because loss by these routes is quantitatively so slight. Milk may be a major route of excretion for some compounds. Milk has a relatively high fat content, 3–5 percent or even higher, and therefore compounds that are lipophilic may be excreted in milk to a significant extent. Some of the toxicants known to be present in milk are the highly lipid-soluble chlorinated hydrocarbons: for example, the polychlorinated biphenyls (PCBs) and DDT. Certain heavy metals may also be excreted in milk. Lead is thought to be secreted into milk by the calcium transport process.

2.5 SUMMARY This chapter has conveyed some of the general biochemical and physiological principles that govern absorption, distribution, and elimination of toxic agents, in particular

• The importance of lipid solubility, molecular size, and degree of ionization to the rate at which a molecule moves through a membrane by passive transfer or diffusion.

• The characteristics of other transfer processes such as facilitated diffusion, active transport, phagocytosis, and pinocytosis.

• Absorption from the gastrointestinal tract with particular emphasis on the importance of pH as a determinant of absorption of ionizable organic acids and bases as well as on compoundspecific and host-related factors such as lipid solubility and molecular size, the presence of villi and microvilli in the intestine, the possibility that the compound can be absorbed by facilitated or active transport mechanisms, and the action of gastrointestinal enzymes or intestinal microflora.

• Factors determining the rate of diffusion across the skin. • Absorption of solid and liquid particulates and of gases and vapors in the lung. • Simple classical and physiologically based kinetic models describing disposition (distribution, metabolism, and excretion).

• Excretion from kidney, liver (including enterohepatic circulation), and lung, and by less general routes such as skin, hair, sweat, nails, or milk.



REFERENCES AND SUGGESTED READING Abernethy, D. R., and D. J. Greenblatt, “ Drug disposition in obese humans: An update,” Clin. Pharmacokinet. 11: 199–212 (1986). Andersen, M. E., H. J. Clewell, M. L. Gargas III, F. A. Smith, and R. H. Reitz, “ Physiologically-based pharmaco-kinetics and the risk assessment process for methylene chloride,” Toxicol. Appl. Pharmacol. 87: 185–205 (1987). Bragt, P. C., and E. A. van Dura, “ Toxicokinetics of hexavalent chromium in the rat after intratracheal administration of chromates of different solubilities,” Ann. Occup. Hyg. 27: 315–322 (1983). Brewster, D., M. J. Humphrey, and M. A. McLeavy, “ The systemic bioavailability of buprenorphine by various routes of administration,” J. Pharm. Pharmacol. 33: 500–506 (1981). Brodie, B. B., H. Kurz, and L. S. Shanker, “ The importance of dissociation constant and lipid-solubility in influencing the passage of drugs into the cerebrospinal fluid,” J. Pharmacol. Exp. Therap. 130: 20–25 (1960). Chamberlain, A. C., M. J. Heard, P. Little, D. Newton, A. C. Wells, and R. D. Wiffen. Investigations into Lead from Motor Vehicles, AERE. Publication N2R9198, Harwell, England, 1978. Crouthamel, W. G., J. T. Doluisio, R. E. Johnson, and L. Diamond, “ Effect of mesenteric blood flow on intestinal drug absorption,” J. Pharm. Sci. 59: 878–879 (1970). English, J. C., R. D. R. Parker, R. P. Sharma, and S. G. Oberg, “ Toxicokinetics of nickel in rats after intratracheal administration of a soluble and insoluble form,” Am. Ind. Hyg. Assoc. J. 42: 486–492 (1981). Gariépy, L., D. Fenyves, and J.-P. Villeneuve, “ Propranolol disposition in the rat: Variation in hepatic extraction with unbound drug fraction,” J. Pharm. Sci. 81: 255–258 (1992). Gregus, Z., and C. D. Klaassen, “ Disposition of metals in rats: A comparative study of fecal, urinary, and biliary excretion and tissue distribution of eighteen metals,” Toxicol. Appl. Pharmacol. 85: 24–38 (1986). Guidotti, G., “ The structure of membrane transport systems,” Trends Biochem. Sci. 1: 11–12 (1976). Hamilton, D. L., and M. W. Smith, “ Inhibition of intestinal calcium uptake by cadmium and the effect of a low calcium diet on cadmium retention,” Environ. Res. 15: 175–184 (1978). Herrmann, D. R., K. M. Olsen, and F. C. Hiller, “ Nicotine absorption after pulmonary instillation,” J. Pharm. Sci. 81: 1055–1058 (1992). Hirom, P. C., P. Millburn, and R. L. Smith, “ Bile and urine as complementary pathways for the excretion of foreign organic compounds,” Xenobiotica 6: 55–64 (1976). Hogben, C. A. M., D. J. Tocco, B. B. Brodie, and L. S. Shanker, “ On the mechanism of intestinal absorption of drugs,” J. Pharmacol. Exp. Therap. 125: 275–282 (1959). Hussain, A. A., K. Iseki, M. Kagoshima, L. W. Dittert, “ Absorption of acetylsalicylic acid from the rat nasal cavity,” J. Pharm. Sci. 81: 348–349 (1992). King, F. G., R. L. Dedrick, J. M. Collins, H. B. Matthews, and L. S. Birnbaum, “ Physiological model for the pharmacokinetics of 2,3,7,8-tetrachlorodibenzofuran in several species,” Toxicol. Appl. Pharmacol. 67: 390– 400 (1983). Lien, E. J., and G. L. Tong, “ Physicochemical properties and percutaneous absorption of drugs,” J. Soc. Cosmet. Chem. 24: 371–384 (1973). Nebert, D. W., A. Puga, and V. Vasiliou, “ Role of the Ah receptor and the dioxin-inducible [Ah] gene battery in toxicity, cancer, and signal transduction,” Ann. NY Acad. Sci. 685: 624–640 (1993). Nelson, D. R., T. Kamataki, D. J. Waxman, F. P. Guengerich, R. W. Estabrook, R. Feyereisen, F. J. Gonzalez, M. J. Coon, I. C. Gunsalus, O. Gotoh, K. Okuda, and D. W. Nebert, “ The P450 superfamily: Update on new sequences, gene mapping, accession numbers, early trivial names of enzymes, and nomenclature,” DNA Cell Biol. 12: 1–51 (1993). O’Flaherty, E. J., Toxicants and Drugs: Kinetics and Dynamics, Wiley, New York, 1981. O’Flaherty, E. J., “ Physiologically based models for bone-seeking elements. IV. Kinetics of lead disposition in humans,” Toxicol. Appl. Pharmacol. 118: 16–29 (1993). Rollins, D. E., and C. D. Klaassen, “ Biliary excretion of drugs in man,” Clin. Pharmacokinet. 4: 368–379 (1979). Schanker, L. S., and J. J. Jeffrey, “ Active transport of foreign pyrimidines across the intestinal epithelium,” Nature 190: 727–728 (1961).



Sha’afi, R. I., C. M. Gary-Bobo, and A. K. Solomon, “ Permeability of red cell membranes to small hydrophilic and lipophilic solutes,” J. Gen. Physiol. 58: 238–258 (1971). U.S. Environmental Protection Agency, Update to the Health Risk Assessment Document and Addendum for Dichloromethane: Pharmacokinetics, Mechanism of Action and Epidemiology, EPA 600/8-87/030A (1987). Wagner, J. G., “ Properties of the Michaelis-Menten equation and its integrated form which are useful in pharmacokinetics,” J. Pharmacokinet. Biopharmaceut. 1: 103–121 (1973). Williams, R. T., “ Interspecies scaling,” in T. Teorell, R. L. Dedrick, and P. G. Condliffe, eds., Pharmacology and Pharmacokinetics, Plenum, New York, 1974, Table IV, p. 108.

3 Biotransformation: A Balance between Bioactivation and Detoxification BIOTRANSFORMATION: A BALANCE BETWEEN BIOACTIVATION AND DETOXIFICATION


This chapter identifies the fundamental principles of foreign compound (xenobiotic) modification by the body and discusses

• • • • •

How xenobiotics enter, circulate, and leave the body The sites of metabolism of the xenobiotic within the body The chemistry and enzymology of xenobiotic metabolism The bioactivation as well as inactivation of xenobiotics during metabolism The variations in xenobiotic metabolism resulting from prior or concomitant exposure to xenobiotics and from physiological factors

The body is continuously exposed to chemicals, both naturally occurring and synthetic, which have little or no value in sustaining normal biochemistry and cell function. These chemical substances (xenobiotics) can be absorbed from the environment following inhalation, ingestion in food or water, or simple exposure to the skin (Figure 3.1). Biotransformation or metabolism of the chemicals allows the elimination of the absorbed chemicals to occur. Without this process, chemicals that were readily absorbed through lipid membranes because of a high octanol/water partition coefficients would fail to leave the body. They would be passively reabsorbed through the lipid membrane of the kidney tubule instead of remaining in, and passing out with, the urine (Figure 3.2). In addition, they would not be subject to active transport mechanisms capable of actively secreting many xenobiotic metabolites. Thus, an important objective of biotransformation is to promote the excretion of chemicals by the formation of water-soluble metabolites or products. Biotransformation can also alter the biological activity of chemicals, including endogenous chemicals released in the body, such as steroids and catecholamines, both by structural alteration and by enhancing their partition away from cellular compartments, membranes, and receptors. Thus biotransformation helps to both terminate the biological activity of chemicals and increase their ease of elimination. Biotransformation is defined as the chemical alteration of substances by reactions in the living organism. For convenience, the conversion of xenobiotics is divided into two phases: metabolic transformations (phase I reactions) and conjugation with natural body constituents (phase II reactions) (Figure 3.3). The reactions of both of these phases are predominantly enzyme-catalyzed. A xenobiotic does not necessarily undergo metabolism by a sequential combination of phase I followed by phase II reactions for successful elimination. It may undergo phase I metabolism alone, phase II alone, and occasionally, phase I reactions subsequent to phase II conjugations are encountered. An important objective of biotransformation is to promote the excretion of absorbed chemicals by the formation of water-soluble drug metabolites or products (p in Figure 3.1). Increased water solubility is derived primarily from the phase II reactions since most conjugates exist in the ionized state at physiological pH levels. This promotes excretion (e in Figure 3.1) by decreasing xenobiotic reabsorpPrinciples of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




Abbreviations: a = major absorption sites; e = excretion sites; f = filtration sites; m = major metabolism site; p = metabolic product; s = secretion sites; x = xenobiotic.

Figure 3.1 Diagram of major sites of xenobiotic absorption, metabolism, and excretion.

tion from the renal tubule following glomerular filtration or active secretion (f and s, respectively, in Figures 3.1 and 3.2) and from the gastrointestinal tract following biliary secretion. Biotransformation also decreases the entry of xenobiotics into cells of all organs and makes them more suitable for secretion by active transport mechanisms into the bile and urine. Active secretion requires both energy and a carrier protein and is capable of forcing molecules up a chemical gradient. Of the carrier molecules, those that recognize and transport organic acids have particular importance for drug conjugates since they can carry glucuronides, sulfate esters, and amino acid conjugates. While



Figure 3.2 The role of metabolism in increasing urinary excretion.

excretion of xenobiotics into the urine largely terminates the exposure of the body to the chemical, excretion in the bile may not always result in efficient drug elimination because enterohepatic recirculation may occur. This can result in the prolonged effects and persistence seen with some drugs and chemicals. Enterohepatic recirculation most often involves the secretion of xenobiotic conjugates in the bile and their hydrolysis by enzymes from the host or microorganisms in the gastrointestinal tract. This deconjugation releases the free xenobiotic, which is often sufficiently lipid soluble (high octanol/water partition coefficient), to be reabsorbed. The reabsorbed xenobiotic returns in the portal circulation to the liver where it is reconjugated, resecreted, and so on. The same reabsorption can also occur if an unmetabolized lipid soluble xenobiotic is secreted in the bile. As stated above, the conversion of xenobiotics is divided into the two phases of metabolic transformation and conjugation (Figure 3.3). The main chemical reactions involved in phase I or metabolic transformation, in approximate order of capacity or importance, are oxidation, hydrolysis, and reduction. Of the phase II or conjugation reactions, glucuronidations are generally the most prevalent in mammals, with the other conjugations having lesser overall capacity. All conjugation reactions, except with glutathione, involve the participation of energy-rich or activated cosubstrates. Conjugation with the cellular nucleophile, glutathione, is an especially important mechanism for the sequestering of electrophilic intermediates generated during phase I metabolism, and it can occur, albeit less efficiently, in the absence of enzyme. As mentioned above, with reference to the generation of electophilic metabolites, biotransformation can have a variety of effects on the biological reactivity of the xenobiotic. The chemical can be inactivated or detoxified, can be changed into a more toxic substance (bioactivated), or can be changed into other chemical entities having effects that differ both quantitatively and qualitatively from the parent compound (Table 3.1). Generally, phase II metabolites are inactive, but important exceptions exist. Phase I metabolites may or may not be inactive, and many are more reactive than the original xenobiotic. The greater reactivity can be viewed as an unfortunate necessary prerequisite to conjugation, which is the step contributing most to the facilitation of excretion (Figure 3.4).

60 Figure 3.3 Xenobiotic metabolism summary; reaction characteristics and flowchart.



TABLE 3.1 Pharmacologic Effects with Xenobiotic Metabolism Phase I

Phase II Active to Inactive

Amphetamine —P450→ phenylacetone Cocaine —esterase→ benzoylecgonine Hexobarbital —P450→ Phenytoin —P450→

Acetaminophen —UGT/ST→ Aflatoxin 2,3-epoxide —GST→ 8 glutathionyl-9 hydroxyaflatoxin Morphine —UGT→ morphine-3-glucuronide Testosterone —ST→ Active to Active

Acetylsalicylic acid —esterase→ salicylic acid Codeine —P450→ morphine Heroin —esterase→ morphine Primidone-P450→ phenobarbital

Morphine —UGT→ morphine-6-glucuronide Procainamide —AT → N-acetylprocainamide Thiobarbital —P450→ barbital Inactive to Active

Chloral hydrate —reductase→ trichloroethanol Prontosil —reductase→ sulfanilamide Sulindac —reductase→ sulfide Inactive to Toxic N-Hydroxyacetylaminofluorene —ST→ Acetaminophen —P450→ N-acetyl-p-benzoquinine imine Acetylhydrazine —P450→ acetylcarbonium ion N-Hydroxymethylaminoazobenzene —ST→ Aflatoxin —P450→ aflatoxin-8,9 epoxide Tetrachloroethylene —GST→ Malathion —P450→ malaoxon Tolmetin —UGT→ Nitrofurantoin —reductase→ hydroxylamine Benzo(a)pyrene 7,8-diol —P450→ benzo(a)pyrene 7,8-diol 9,10-epoxide Dimethylnitrosamine —P450→ methyldiazohydroxide

Figure 3.4 The balance of reactivity and excretability in xenobiotic metabolism.



3.1 SITES OF BIOTRANSFORMATION Xenobiotic metabolism occurs in all organs and tissues in the body. Because many of the chemicals metabolized can have deleterious effects on the body, xenobiotic metabolism can be considered a defense mechanism that hastens the elimination of a toxic chemical and thus terminates the exposure. When viewed as a defense mechanism, it is not surprising that the exposure is best terminated at the point of exposure. These are the so-called portals of entry (shown as sites of absorption [a] for xenobiotics [X] in Figure 3.1), and constitute mainly the skin, lung, and intestinal mucosa. While drug metabolizing enzymes are present in all these tissues (Table 3.2), and at relatively high activity in some, particularly intestine and lung, the liver is by far the most important tissue for xenobiotic metabolism (site [m] in Figure 3.1). Although it is not the first tissue of the body to be exposed to chemicals, the liver receives the entire chemical load absorbed from the gastrointestinal tract, which is the predominant portal of entry for most xenobiotics (Figure 3.1). The xenobiotic metabolizing enzymes are present in high concentrations and the organ itself has large bulk, approximately 5 percent of the total body weight. Xenobiotics absorbed from the lungs and skin can also quickly move to the liver for metabolism. Once in the liver, the highly vascular nature of the tissue and the intimate contact between blood and hepatocytes, which contain the xenobiotic metabolizing enzymes, allows for the rapid diffusion of chemicals in and metabolites out (Figure 3.5). Although not a portal of entry, the kidney is an organ where xenobiotics are likely to be concentrated during the excretion process, and this may be the reason for the relatively high level of xenobiotic metabolizing enzymes in this tissue. Although the data presented in Table 3.2 are from laboratory animals, there is little evidence to contraindicate the existence of a similar distribution pattern in humans. Within the liver, hepatocytes or parenchymal cells are the major site of drug biotransformation, and within these cells it is the endoplasmic reticulum, which occupies about 15 percent of the hepatocyte volume and contains 20 percent of the hepatocyte protein, which houses the bulk of the critical drug metabolizing enzyme activity. (The nonparenchymal cells, including endothelial and Kupffer cells, constitute 35 percent of liver cell number but only contribute 5–10 percent of liver mass. Their drug metabolizing enzyme activities are typically less than 20 percent of that in hepatocytes). When liver is carefully homogenized, fragments of the endoplasmic reticulum are converted to microsomes (an artifact of cell disruption). The drug-metabolizing enzymes located in the endoplasmic reticulum are often referred to as microsomal enzymes, and it is often stated that chemicals are metabolized by liver microsomes. Enriched microsomal fractions are usually obtained by differential sedimentation, either as a suspension with cytoplasm (10,000g supernatant) or as a sediment free of cytosol (105,000g precipitate) (Table 3.3). Many important xenobiotic metabolizing enzymes reside in the cytoplasm and microsomal fractions (Figures 3.3 and 3.6). Oxidations and glucuronidations are the most common reactions occurring in microsomes. The terminal oxidase responsible for many of the oxidations, cytochrome P450, represents about 5 percent of the microsomal protein under normal conditions; more if induction has occurred (see text below). Other flavoproteins necessary for cytochrome P450 function and epoxide hydrolase, an enzyme important in the further metabolism of epoxides formed by cytochrome P450–dependent oxidation, are also conveniently located in the endoplasmic reticulum (Figure 3.6). Microsomal metabolism in tissues other than liver is seldom quantitatively important in overall drug elimination, but local generation of active metabolites may be important in drug-induced tissue damage, carcinogenesis, and other effects. Enzymes located in the cytoplasm of the hepatocyte catalyze a wide variety of both phase I and phase II reactions. Dehydrogenases and esterases are examples of phase I enzymes found predominantly in the cytosol. The sulfotransferase and glutathione transferase enzymes also depicted in Figure 3.6 serve as examples of phase II enzymes that are similarly located.


Figure 3.5 Diagrammatic rendition of hepatic lobule blood flow.

TABLE 3.2 Drug-Metabolizing Enzyme Activitiesa in Various Organs Lung Rabbit Cytochrome P450 UDP-glucuronosyltransferase (p-nitrophenol)b Glutathione S-transferase (DCNB)b Rat Cytochrome P450 Ethoxyresorufin demethylase (P4501A) Erythromycin demethylase (P4503A) UDP-glucuronosyltransferase (p-nitrophenol)b Glutathione S-transferase (DCNB)b a

Intestine Mucosa


Kidney Cortex

0.4 0.4 5.3

0.34 — —

1.45 6.6 21.9

0.33 2.9 7.4


0.09 0.003 — 0.8 2.1

0.05 0.001 0.12 — —

0.84 0.034 0.47 4.4 76.4

0.12 0.001 0.06 3.3 3.8


All activities are expressed on a per milligram of protein basis (DCNB = 1,2-dichloro 4-nitrobenzene). Litterst CL, Mimnaugh EG, Reagan RL, Gram TE, Drug Metab. Disp. 3: 259 (1975).






TABLE 3.3 Preparation of Subcellular Fractions for Xenobiotic Metabolism Studies Step 1

2 3 4 5

Procedure Liver pieces homogenized in 4 volumes of 0.25 M sucrose in Potter Elvehjem glass–Teflon homogenizer Homogenate centrifuged at 2000g for 10 min

Result Tissue structure disrupted and hepatocytes sheared.

Unbroken cells, connective tissue, and nucleii sedimented 2000g supernatant centrifuged at 10,000g for 15 min Heavy mitochondria sedimented as pellet 10,000g supernatant centrifuged at 18,000g for 15 Light mitochondria sedimented as pellet min 18,000g supernatant centrifuged at 105,000g for 60 Microsomes sedimented as pellet leaving nonturbid min cytosol in 0.2 M sucrose supernatant

Without exception, the xenobiotic metabolizing enzymes occur in multiple forms (isozymes), often with differing substrate selectivities. The presence of specialized isozymes, which can more efficiently metabolize a specific range of chemicals, may enable those specific chemical challenges to be met more effectively. With differing substrate selectivities, often comes different sensitivity to inhibitors. The presence of multiple forms thus carries the advantage of not having all the metabolism of all compounds metabolized by that route or chemical reaction being subject to inhibitory influences at the same time. It has also been found that the synthesis of individual isozymes can be under different regulatory influences. The body can thus meet a chemical challenge with a finely tuned response to increase the production of only that enzyme best equipped to counter or neutralize the challenge.

Abbreviations (clockwise) are ST = sulfotransferase; PAPS = adenosine 3′-phosphate 5′-phosphosulfate; GST = glutathione S-transferase; GSH = glutathione; AlcDH = alcohol dehydrogenase; ES = esterase; FP1 = NADH cytochrome b5 reductase; b5 = cytochrome b5; P450 = cytochrome P450; mEH = microsomal epoxide hydrolase; FP2 = NADPH-cytochrome P450/c reductase; UGT = UDP-glucuronosyltransferase; UDPGA = uridine 5′-diphosphoglucuronic acid; FP3 = flavin-dependent monooxygenase.

Figure 3.6 Diagram of the subcellular localization and organization of major xenobiotic metabolizing enzymes and necessary cofactors.



3.2 BIOTRANSFORMATION REACTIONS There is multiple redundancy in metabolism. There may be more than one site of attack on a xenobiotic (e.g., amine and ester group of cocaine), there may be more than one metabolic reaction at a single site (e.g., sulfation and glucuronidation of the phenolic group of acetaminophen), and more than one enzyme/isozyme capable of catalyzing a single reaction at a single site. An example of the complexity of possible metabolism of a relatively simple hypothetical chemical is shown in Figure 3.7. From considerations in this chapter so far, it can be seen that the subcellular location of a metabolic reaction does not dictate the nature of the reaction. Both oxidations and hydolyses, albeit by different enzymes, occur in the cytoplasm and endoplasmic reticulum. Likewise, so do conjugations when considered collectively, but a specific form of conjugation may occur only in a single fraction (e.g., sulfation in the cytoplasm). The enzymes are therefore considered in the following paragraphs by the nature of the chemical reaction that they catalyze, and only for phase I oxidations is the subcellular location used as a convenient subdivision. Phase I; Oxidations Microsomal Microsomal oxidations are predominantly catalyzed by a group of enzymes called mixed-function oxidases or monooxygenases. The terminal oxidase is generally a hemoprotein called cytochrome P450 but can be a flavoprotein.

Figure 3.7 Possible metabolic conversions of a simple hypothetical xenobiotic.



Cytochrome P450 is a collective term for a group of related hemoproteins, all with a molecular weight (MW) around 50,000 daltons, which as will be seen later, differ in their substrate selectivity and in their ability to be induced and inhibited by drugs and chemicals (Table 3.4). Cytochrome P450–catalyzed oxidations are categorized by the nature of the atom that is oxidized (see Figure 3.8). Subsequent to the oxidation, the oxygen atom from molecular oxygen may be retained within the major fragment of the chemical or it may be eliminated by molecular rearrangement (e.g., O and N dealkylations). Whatever the atom oxidized, or the name given to the reaction, the cytochrome P450–mediated oxidation involves the same cyclic three-step series (Figure 3.9). Step 1. The xenobiotic [X] first binds to the cytochrome at a substrate binding site on the protein and alters the conformation sufficiently to enable the efficient transfer of electrons to the heme from NADPH via a nearby (see Figure 3.6) flavoprotein, NADPH cytochrome P450 reductase. (The activity of this FAD- and FMN-containing flavoprotein is often determined experimentally using exogenously added mitochondrial cytochrome c rather than microsomal cytochrome P450 as the electron acceptor and so is often identified as NADPH cytochrome c reductase). The conformational change can sometimes be seen in vitro (in the absence of electron transfer) as an alteration of the heme from a low-spin to a high-spin state, which results in a blue shift in the absorbance maximum of the hemoprotein. The gain at 390 nm and loss at 420 nm, when seen by difference spectroscopy, is termed a type I binding spectrum (not to be confused with phase I metabolism). Step 2. The reduction of the heme iron from its normal ferric state to the ferrous state allows a molecule of oxygen (O–O) to bind (the binding of CO rather than oxygen to ferrous cytochrome P450 in the in vitro situation provides a characteristic absorbance maximum around 450 nm, which gives this cytochrome its name). Step 3. The ternary complex of xenobiotic, cytochrome, and oxygen receives another electron, either through the same flavoprotein as before or through an alternative path involving a different flavoprotein in which the electron is first passed through cytochrome b5, another cytochrome present in the endoplasmic reticulum (see Figure 3.6). This alternate pathway for the second electron can also use NADH as the pyridine nucleotide electron donor. The addition of the second electron to the ternary complex results in a eventual splitting of the molecular oxygen, one atom of which oxidizes the chemical, the other atom picking up protons to form water, returning ferric cytochrome P450 to repeat the cycle. Flavoprotein-catalyzed oxidations differ from cytochrome P450–catalyzed oxidations in mechanism and in substrate selectivity. For the flavoproteins (a 65,000-dalton protein containing only FAD), the enzyme forms an activated oxygen complex (“ cocked gun” ) and the addition of a metabolizable chemical discharges this, in the process of becoming oxidized. The electrophilic oxygenated species attacks nucleophilic centers. A wide range of chemicals can thus be metabolized by this flavoprotein; the important feature for metabolism being a heteroatom (nitrogen, sulfur) presenting a lone pair of electrons (Table 3.5). Some compounds are metabolized both by flavin-containing monooxygenases and cytochrome P450 but to different products. An example is dimethylaniline, which is metabolized to the N-oxide by the flavoprotein and is N-demethylated by cytochrome P450. Nonmicrosomal Oxidations in other subcellular organelles can be catalyzed by flavoproteins (e.g., monoamine oxidase in mitochondria) or pyridine nucleotide linked dehydrogenases (e.g., alcohol and aldehyde dehydrogenases in cytoplasm). Dehydrogenase-catalyzed oxidations do not involve molecular oxygen. The oxidation of the chemicals or drugs occurs through electron transfer to a pyridine nucleotide, usually NAD+. Most of the dehydrogenases are cytoplasmic in location. The most noteworthy of this class of enzymes in humans is the dehydrogenase responsible for the metabolism of ethanol. In contrast to the major



1,2 1,2 6,7


2A 2B











6 1








Forms Present in

TABLE 3.4 Important Cytochrome P450sa



9,10,11, 12,13

4,5 9,10,13




liver, EH

liver, EH

liver, GI

liver, EH




ethoxyresorufin, PCB, PAH, TCDD, 3MC, 7,8-benzoflavone, phenacetin deE, caffeine 3 isosafrole, omeprazole ellipticine, furafylline deM, benzopyrene, aflatoxin, cooked-food heterocyclic amines, NBI PAH (2A1) pentoxyresorufin, PB benzphetamine PB tolbutamide PB, RIF sulfaphenazole sex, maturation, S-mephenytoin 4-OH, tranylcypromine debrisoquine 4-OH, RIF quinidine sparteine, dextromethorphan, bufuralol 1′-OH ethanol, disulfiram chlozoxazone, ethanol ketones, pyridine, dimethylnitrosamine, isoniazid acetaminophen, CCl4, 4-nitrophenol-OH 3-methylindole naphthalene

Substrates/ Reactions


1,2,3,8 1











Abbreviations: deE = deethylation deM = demethylation DEHP = di(2-ethylhexyl)phthalate DEX = dexamethasone EH = extrahepatic GI = gastrointestinal tract Kid = kidney MI = metabolic-intermediate 3MC = 3-methylcholanthrene NBI = N-benzylimidazole OH = hydroxylation PAH = polycyclic aromatic hydrocarbons PCB = polychlorinated biphenyls PCN = pregeneolone 16α carbonitrile PB = phenobarbital RIF = rifampicin TAO = troleandomycin TCDD = 2,3,7,8-tetrachlorodibenzo-p-dioxin






Forms Present in

TABLE 3.4 Important Cytochrome P450sa





liver, kidney


valproic acid

erythromycin N deM, TAO MI complex, cyclosporine, quinidine, testosterone, and cortisol 6β-OH, mephenytoin (rat), benzphetamine, nifedipine, and other dihydropyridines Lauric acid and prostaglandin ω-OH

Substrates/ Reactions Inducers

peroxisome proliferators, DEHP, clofibrate pregnancy (4A4)

glucocorticoids DEX, PCN, macrolides, TAO clotrimazole, phenobarbital

Inhibitors cimetidine, naringenin


Figure 3.8 Cytochrome P450–catalyzed oxidations.




Figure 3.9 The cytochrome P450 oxidation cycle.

microsomal oxidizing enzyme, these enzymes are not subject to extensive induction (see discussion later). Monoamine oxidases, which are usually mitochondrial in location, oxidize by electron transfer to a flavin group. Monoamine oxidases are responsible for the normal metabolism of neurotransmitters, and exposure to agents, which are also metabolized by this enzyme, (e.g., tyramine) can result in toxicities or pharmacological effects arising from accumulation of the unmetabolized neurotransmitter. A neurotoxin of much recent interest, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP), which leads to Parkinson’s syndrome, is bioactivated by monoamine oxidase B (a form selectively inhibited by deprenyl and located in serotonergic neurons in the brain). Environmental compounds or drugs that are also tetrahydropyridines have been speculated to be causative agents in Parkinson’s disease in the elderly.

Phase I; Hydrolyses Hydrolysis reactions are catalyzed by esterases and amidases. While both can be microsomal, esterases are predominantly cytosolic in location. Hydrolysis of amides and esters produces two reactive centers,

TABLE 3.5 Compounds Metabolized by the Flavin-Containing Monooxygenases Heteroatom Nitrogen Sulfur

Class Tertiary amine Secondary amine Thiocarbamides Thioamides Thiols Sulfides

Examples N-Dimethylaniline, imipramine, amitryptyline N-Methylaniline, desipramine, nortryptyline Thiourea, propylthiouracil, methimazole Thioacetamide Dithiothreitol, β-mercaptoethanol Dimethylsulfide



both of which are suitable for conjugation, if the metabolites are not first excreted as the phase I products (Figure 3.10). Epoxide hydrolase activity is predominantly microsomal, but an enzyme is also present in the cytosol. Most hydrolyses occur to a significant extent in tissues other than liver. Their quantitative importance is variable, depending on the chemical challenge. One significant extrahepatic location of esterases is in the blood (plasma and erythrocytes), and of great concern is the enzyme normally responsible for the hydrolysis of acetylcholine. Blockade of this enzyme is the mode of action of many insecticides and “ nerve gases.”

Figure 3.10 Hydrolytic and reductive phase I reactions.



Phase I; Reductions Reductive metabolism in the liver endoplasmic reticulum can occur through the mediation of both hemoprotein (cytochrome P450) and flavoproteins. Reductions of azo and nitro groups are the most commonly encountered (Figure 3.10), but reduction of disulfides, sulfoxides, epoxides, and N-oxides can also occur. In many instances, the products of reductive metabolism can be reoxidized under aerobic conditions.

Phase II; Glucuronidation Glucuronidations are catalyzed by a group of closely related 55,000-dalton isozymes, termed UDPglucuronosyltransferases, located within the endoplasmic reticulum. They catalyze the transfer of glucuronic acid from a uridinediphosphoglucuronic acid (UDPGA) cofactor to a carboxyl or hydroxyl (phenol), or less often an amine group on the xenobiotic (or phase I metabolite) (Figure 3.3). The UDPGA is generated from the abundant carbohydrate supply in the liver as glucose-1-phosphate, and following the reaction with UTP, the resultant UDP-glucose is oxidized. The formation of the glucuronide does not involve the acid group of glucuronic acid, so the conjugate retains acid and ionized character at physiological pH, providing dramatic enhancement of water solubility and excretability to the xenobiotic. Glucuronides are actively secreted into bile and in the proximal tubule of the kidney. Xenobiotics conjugated as glucuronides can be released as either a phase I metabolite or the original molecule by the action of glucuronidases of both mammalian and microbial origin. UDP-glucuronosyltransferases occur in multiple forms. The most common classification utilized for the enzymes responsible for the metabolism of xenobiotics are those (GT1) that conjugate planar phenols (e.g., 1-naphthol, 4-nitrophenol) and are induced by polycyclic hydrocarbon-like molecules (see Table 3.6) and those (GT2) that conjugate nonplanar phenols (e.g. morphine, chloramphenicol) and are induced by phenobarbital and similar compounds. There are other forms which appear to be more selective for endogenous substrates, notably those for the 17 hydroxysteroids (testosterone), the 3 hydroxysteroids (androsterone) and bilirubin. More recent studies using the powerful techniques of molecular biology have provided a more rational classification system, but to aid the reader in understanding the bulk of existing literature, the old system has been used in this chapter. Like cytochrome P450s, UDP-glucuronosyltransferases are often substrate selective rather than substrate specific, being able to metabolize a wide range of compounds poorly (e.g., 4-nitrophenol is conjugated by almost all isozymes) while metabolizing substrates with particular characteristics very efficiently. Also like cytochrome P450s, more than one form may be induced by a xenobiotic inducing agent (both bilirubin and testosterone as well as morphine conjugations are induced by phenobarbital).

Phase II; Sulfation Sulfate conjugation is an important alternative to glucuronidation for phenolic compounds and occasionally arylamines. Sulfate availability within the cell may be limited, so this conjugation pathway decreases in importance with higher xenobiotic or phenolic metabolite concentrations. The 3′-phosphoadenosine-5′-phosphosulfate (PAPS) cofactor from which the sulfate group is transferred is generated from ATP and inorganic sulfate. The sulfate can be derived from the sulfur containing amino acids, cysteine and methionine. The enzymes catalyzing the sulfate conjugations are a family of cytosolic 64,000-dalton enzymes, termed sulfotransferases, and are one of the exceptions to the major groups of drug metabolizing enzymes in that they appear to not be induced by xenobiotic compounds (see Table 3.6). The sulfates are completely ionized at physiological pH and easily eliminated. Much like glucuronides, enzymes exist (termed sulfatases) that can break the conjugate and return the xenobiotic, if it is phenolic, or the phase I metabolite of a xenobiotic, if it was oxidized or hydrolyzed to that functional group.



TABLE 3.6 Changes in Rat Hepatic Drug-Metabolizing Enzymes Following Xenobiotic Administrationa

Agent (use)

Dose (mg/kg × days)

CYT P450


80 × 4 75 × 3 500 × 4 75 × 3 80 × 4 100 × 3 20 × 4 25 × 6

400b 320c 300b 290b 265 245 215 205

— 305 — 120 130 70 185 265

65 240 100 120 127 80 300 270

150 240 200 225 455 155 155 280

— 225 — 265 155 125 145 140

— 70 — 85 50 60 90 110






130 175 100 80 195 250 165 140 — — 215 110 160 120 80 100 85 240 145

225 170 150 85 260 130 110 95 — — 330 175 205 110 120 105 95 320 145

130 130 140 95 215 185 — 155 — — 170 100 180 — 95 85 95 140 155

90 115 120 — 100 85 — 100 — — 70 70 100 — 90 105 120 80 90

205 90 130 250

235 125 100 275

145 105 90 130

80 70 100 65


ST (pNP)

(% control) SKF 525-A 1-Benzylmidazole Troleandomycin (antibiotic) Clotrimazole (antifungal) Phenobarbital (anticonvulsant) Dexamethasone (glucocorticoid) 3-Methylcholanthrened PCBs (Aroclor 1254) (transformer fluid) 2,3,6,7-Tetrachlorodibenzodioxin (TCDD)d 4,4″-Dipyridyl Fluconazole (antifungal) Pregnenolone 16α carbonitriled Clofibrate (antihypertriglyceridemic) trans-Stibene oxided 5,6-Naphthoflavone [BNF] Benzo(a)pyrened Miconazole (antifungal) Phenytoin (anticonvulsant) Carbamazepine (anticonvulsant) Tioconazole (antifungal) Ketoconazole (antifungal) Isosafroled 7,8-Benzoflavoned Isoniazid (antitubercular) 1,10-Phenanthroline Cimetidine (antiulcer) 3,4 Benzoquinoline Butylated hydroxyanisole (antioxidant)d 2,2′-Dipyridyl Chloramphenicol (antibiotic) Cyclosporine (immunosuppressant) 4,7-Phenanthroline

0.01 × 1 185 75 × 3 75 × 3 75 × 4 400 × 3 400 × 4 80 × 3 20 × 4 150 × 3 100 × 7 100 × 7 150 × 3 150 × 4 150 × 4 100 × 4 100 × 4 75 × 3 350 × 3 75 × 3 500 × 10

190 180c 180 180 175 165 — 150 150 145 130c 130c 120 105e 105 100 90 85

245 180 85 65 170 495 165 165 110 125 300 150 150 125 60 105 95 230 150

75 × 3 300 × 3 25 × 10 75 × 3

85 80 80 80

310 95 — 390


Abbreviations: Enzymes: UGT = UDP-glucuronosyltransferase, GST = glutathione S-transferase, ST = sulfotransferase. Substrates: pNP = p-nitrophenol, N = 1-naphthol, M = morphine, CDNB = 1-chloro-2,4-dinitrobenzene. Full detection requires prior destruction of metabolic intermediate complex. c Full detection requires time-dependent displacement of azole ligand by CO. d From Watkins JB, Gregus Z, Thompson TN, Klaassen CD, Toxicol. Appl. Pharmacol. 64: 439 (1982); Thompson TN, Watkins JB, Gregus Z, Klaassen CD, Toxicol. Appl. Pharmacol. 66: 400 (1982). e Induction of P4502E isozyme (see Table 3.3) obscured by decreases in other forms. b

Phase II; Glutathione Conjugation Conjugation with glutathione (γ-glutamylcysteinylglycine) is an important reaction for sequestering reactive (toxic) metabolites, which may be generated by cytochrome P450 oxidations. Glutathione



S-transferases are located predominantly in the cytosol, and hepatic concentrations of the necessary nucleophilic glutathione cosubstrate are high (> 5 mM). The major transferases consist of homo- or heterodimers of a limited number of forms of approximately 25,000-dalton subunits. The different subunit combinations confer different but overlapping substrate selectivity and isoelectric points and are expressed differently in different organs within an animal species. The subunits also respond differently to xenobiotic-inducing agents. In addition to cytosolic enzymes, a glutathione transferase unrelated to the cytosolic proteins is present in the endoplasmic reticulum. Further metabolic products of glutathione conjugations include mercapturic acids (acetylated cysteine derivatives), which are the common excretory product. They are formed by sequential removal of glutamate and then glycine from the glutathione portion followed by acetylation of the amino group of the residual cysteine. Other metabolic products are methylated thiols and sulfones. Episulfonium ions and thioketenes can be formed from glutathione adducts and are reactive enough to form adducts with cell macromolecules and cause toxicity. Phase II; Acetylation, Amino Acid Conjugation, and Methylation The conjugations, involving acetylation of xenobiotics containing sulfonamide or amine groups, peptide conjugation of xenobiotics containing carboxylic acid groups, and methylation of xenobiotics containing amine or catechol groups (Figure 3.3), do not contribute much to enhanced excretability through an increase in water solubility, but serve to mask reactive centers. A problem with some early sulfonamides was that the acetylated metabolites were sufficiently less water-soluble that, they precipitated in the urine, resulting in renal damage. Both acetylations and amino acid conjugations utilize coenzyme A as a cofactor and require the formation of a thioester with the carboxylic acid group, either of acetate or of the xenobiotic. The thioester then reacts with an amine, either on the xenobiotic (acetylation) or amino acid (amino acid conjugation). In mammals, glycine and glutamate are the amino acids most commonly employed in xenobiotic conjugation, but taurine and aspartic acid conjugates are occasionally used, and in birds, ornithine is often used. Methylation reactions require the formation of S-adenosylmethionine (SAM) from ATP and the amino acid, methionine. All the abovementioned conjugates can be deconjugated; deacetylases can remove acetyl groups, cytochrome P450 can remove methyl groups, and peptidases can split amino acid conjugates. Most conjugations occur to varying degrees in tissues other than the liver. Quantitatively they are often minor, but can be very important for protection from reactive metabolites generated in extrahepatic tissues. Factors Affecting Drug Metabolizing Capabilities With all that has been documented in this chapter so far, it is easy to overlook the fact that as in most biological systems, xenobiotic metabolism is a dynamic situation undergoing constant change. Numerous factors affect the ability to catalyze xenobiotic metabolism. Many are an inherent property of the animal species or strain. In addition, these genetic differences may be further altered by such physiological factors as gender or age. Xenobiotic metabolism in different animal species differs quantitatively and qualitatively from that in human. Extrapolation from animals to human and the selection of the most appropriate animal model is difficult unless the role of species and physiological factors in modulating metabolism is clearly delineated. The contribution of these various factors is also an important consideration within experimental research when there is a need to compare or reproduce findings generated in different laboratories. Another factor of major concern is modification of xenobiotic metabolism by temporary stimuli, particularly chemical exposure. Typical human situations of chemical exposure can involve toxic accidental exposures but originate most often from ingestion of prescribed medications or ingestion of chemicals in the food, either as contaminants or as naturally occurring dietary constituents. The changes in xenobiotic metabolizing capability can be in either positive or negative directions, and each can occur by more than one mechanism. The response can be generalized over many enzymes catalyzing many different reactions or can be specific for a single isozyme and a single reaction.



Stimulation of Xenobiotic-Metabolizing Enzyme Activities by Xenobiotics. The activity of many microsomal and some cytoplasmic drug metabolizing enzymes can be increased by exposure to a wide range of drugs and other chemicals (Table 3.6). Generally, inducing agents possess two features in common: lipid solubility and a relatively long biological half-life (i.e., they gain access to the liver and remain there for a considerable period of time). The stimulation of enzyme activity, called induction, is most often the result of the increased amount of enzyme present. If it is the result of an increased efficiency of existing enzyme it is termed activation, a phenomenon seen under some conditions with UDP-glucuronosyltransferases. Although not currently well documented for xenobiotic metabolizing enzymes, the activity of many enzymes can be altered by structural modification from processes such as phosphorylation by kinases and dephosphorylation by phosphatases. Induction occurs by the inducing substance stimulating the synthesis of new enzyme. Because new protein (enzyme) synthesis requires time, the increase in activity is not an immediate event, and occurs over a period of many hours or days. Returning to a normal state following induction also takes a similar time course. The pattern of enzymes induced (both phase I and phase II) and the time course of induction varies with the agent. Induction is not open-ended, but rather, there appear to be limits to changes in each individual enzyme. Increases in liver microsomal enzyme activities determined in in vitro assays are often magnified for the metabolism of the xenobiotic in the whole animal because accompanying the increased enzyme activity per milligram of membrane protein are increased amounts of membrane per cell and increased overall size (most often an increased number of cells) of the liver. The mechanism of induction is best understood for one group of compounds, the polycyclic aromatic hydrocarbon type of inducers, although this receptor-mediated induction (Figure 3.11) may not be the only mechanism by which these agents induce. The cytosol contains a protein that has a high affinity for polycyclic aromatic hydrocarbon-like molecules. One of the chemicals most extensively utilized for these investigations has been 2,3,7,8,tetrachlorodibenzo-p-dioxin (TCDD). When the agent binds to this “ Ah receptor,” it displaces a heat-shock protein (hsp90), which enables the receptor to enter the nucleus. Through an interaction

Figure 3.11 Induction of xenobiotic-metabolizing enzymes by polycyclic aromatic hydrocarbons and related compounds.



with a transporter protein (ARNT) in the nucleus, it initiates the transcription of mRNA to a limited number of proteins, including certain isozymes of cytochrome P450 (e.g., CYP1A isozymes) and UDPglucuronosyltransferase (GT1), by binding to a regulatory region of these genes. The region of DNA to which it binds has been termed a xenobiotic response element (XRE). These mRNA molecules move out of the nucleus and are translated into new proteins on the ribosomes attached to the endoplasmic reticulum. The burst of mRNA production is usually seen within hours of exposure to the inducing agent. For increased amounts of active cytochrome P450, a coordinate induction of additional heme in the mitochondrion is also needed. Much of the information on this induction mechanism arose from work with the “ nonresponsive” strains of mice (e.g., D2, CF-1; see Table 3.6) in which the Ah receptor appears defective with respect to its affinity for the polycyclic aromatic hydrocarbon. No such well-defined deficiency has yet been found in rat strains or humans. The list of compounds that induce drug-metabolizing enzymes in a manner different from that of polycyclic hydrocarbons is much more extensive and includes chemicals of diverse chemical structure and biological effect. For some of these groups of chemicals (e.g., phenobarbital), no receptor has so far been identified. Different isozymes of the chemical/drug-metabolizing enzymes are induced (see Tables 3.4 and 3.6), and in contrast to the polycyclic hydrocarbons, many cause a marked proliferation of the endoplasmic reticulum and increase in liver size. Some of the induction seen with many of these agents has been attributed to a stabilization of existing enzyme in addition to the formation of new enzyme either via enhanced mRNA production (transcription) or changes in the translation rate of basal amounts of mRNA. Nonmicrosomal enzymes, including sulfotransferases, are not induced as extensively as are microsomal enzymes. Exceptions are the cytosolic GSH transferases, which are induced by a wide range of agents (see Table 3.6). Extrahepatic microsomal enzymes are induced by a more restricted number of compounds compared to those that are able to induce liver enzymes, and polycyclic aromatic hydrocarbon-type induction predominates. A similar degree of induction of both phase I and phase II enzymes does not always occur and can result in an imbalance in the ability of phase II reactions to conjugate all the reactive centers generated by the enhanced phase I activity (e.g., dexamethasone and pregnenolone 16α carbonitrile; Table 3.6). Sometimes, Phase II enzyme activities are increased with little (e.g., tioconazole, isosafrole; Table 3.6) or no (e.g., 2,2′-dipyridyl, 3,4-benzoquinoline; Table 3.6) effect on phase I enzymes. Changes in UDP-glucuronosyltransferases may be preferential for one or the other major isozyme (e.g., GT1 > GT2 for 5,6-naphthoflavone, 3-methylcholanthrene, and 2,3,6,7- tetrachlorodibenzodioxin; GT2 > GT1 for troleandomycin, phenobarbital, clotrimazole, and isosafrole). Changes in microsomal UDPglucuronosyltransferase enzymes may (e.g., clotrimazole, isosafrole, and β-naphthoflavone) or may not (e.g., fluconazole) be accompanied by major induction of the cytosolic glutathione S-transferase activity. The consequences of induction can be diverse. An inducing substance may increase the metabolism of one or more other xenobiotics and can even increase its own metabolism. Induction of microsomal enzymes can also enhance the metabolism of endogenous substrates such as steroids and bilirubin. Thus, induction may be important to consider in multiple drug therapy, chronic toxicity tests, crossover drug testing, and environmental toxicology. Some drug tolerance is explained by increased inactivation of the drug by induced enzymes. When major increases in phase I enzymes producing reactive intermediates are not matched by similar increases in the phase II enzymes responsible for their sequestration, increased toxicity may result. Induction is qualitatively, if not quantitatively, similar in most common laboratory animal species, although the rat is perhaps the most responsive (see Table 3.7). Induction is known to occur in humans, often necessitating a change in the therapeutic dosage regimen of drugs. However, for some agents (e.g., peroxisome proliferators), the inductive response seen in experimental animals is absent in humans at therapeutic doses. Although small differences are evident, the effects of inducers are also similar between strains of a species and between species. Thus, information derived from studies in one laboratory animal species can generally be assumed to occur in another. From the examples given in Table 3.7, the phenobarbi-


3.2 BIOTRANSFORMATION REACTIONS TABLE 3.7 Induction of Xenobiotic-Metabolizing Enzymes in Males of Various Animal Speciesa Phase II

Phase I UGT P450 Inducing Agent Ethanol Phenobarbital






(% of naive animal)

Species and Strain Rat: Fischer Rabbit: NZW Rat: Sprague–Dawley Hamster: Syrian Mouse: CF-1 Rabbit: NZW Rat: Sprague–Dawley Mouse: CF-1 Mouse: D2 Mouse B6 Hamster: Syrian Rabbit: NZW Rat: Sprague–Dawley Mouse: CF-1 Mouse D2 Hamster: Syrian


125 — 235 160 185 — 160 85 105 270 145 — 245 240 280 75

140 147 125 120 — 110 155 — 105 145 — 115 80 — 100 —

115 172 420 220 — 155 130 — 115 140 — 170 155 — 60 —

175 — 210 120 135 — 150 75 110 125 65 — 140 100 80 155

90 — 105 80 — — 85 — — — — — 30 — 160 —


Abbreviations: UGT = UDP-glucuronosyltransferase (two isozymes, GT1 and GT2); GST = glutathione S-transferase; ST = sulfotransferase; NZW = New Zealand White.

tal-induced increases in cytochrome P450, glutathione S-transferase, and preferential increase in GT2 UDP-glucuronosyltransferase activity over GT1 UDP-glucuronosyltransferase activity are similar in hamster and rat. Similarly, phenobarbital does not increase sulfotransferase activity in either species. β-Naphthoflavone, a polycyclic hydrocarbon-type inducer, has a similar effect in rat, mouse, and hamster, although the effect in the mouse depends on the strain employed. Two strains (CF-1 and D2) are considered nonresponsive with respect to induction by polycyclic hydrocarbon induction, and for these, in comparison with a B6 strain, there is no increase in cytochrome P450 nor induction of the GT1 UDP-glucuronosyltransferase. Dexamethasone produces large increases in cytochrome P450 with only minor increases in GT1 and GT2 UDP-glucuronosyltransferases and glutathione-S-transferases in either rat or mouse.

Inhibition of Xenobiotic-Metabolizing Enzymes Since the body contains numerous but relatively nonspecific enzymes to metabolize xenobiotics, many chemicals compete for the same enzymes and mutually inhibit the metabolism of each. This may or may not be of great consequence, depending on whether the activity of xenobiotic-metabolizing enzymes is rate limiting. In considering inhibitors, and their beneficial or adverse effects, it is important to consider the perspective from which it is viewed. Piperonyl butoxide is used to inhibit insect cytochrome P450 so that the insect does not metabolize and rid itself of the pesticide, thus increasing (synergizing) the effectiveness of the pesticide. N-substituted imidazoles (e.g., clotrimazole) inhibit cytochrome P450-dependent ergosterol biosynthesis in fungi and prevent growth. These beneficial agents, if they inhibit human hepatic cytochrome P450 and slow the metabolism of other xenobiotics (usually labeled as drug–drug interactions), are considered as acting in an adverse manner. Inhibition



of acetylcholine esterase by organophosphates is beneficial if it is being used as a pesticide, but not if it is directed against humans. Most studies of inhibition of xenobiotic metabolism have centered on cytochrome P450. Early studies identified a compound, SKF 525A, as one of the first cytochrome P450 inhibitors, and although used extensively in laboratory investigations, it has no therapeutic use. Like most cytochrome P450 inhibitors, much of its effect can be attributed to it being an alternative substrate, namely, a competitive inhibitor. Some compounds exhibit noncompetitive characteristics. Many of these are heme ligands, which do not bind to the apoprotein “ substrate site,” and this class includes many nitrogenous heterocyclic compounds such as substituted pyridines, N-substituted imidazoles, and triazoles. Carbon monoxide, although an inhibitor of cytochrome P450 via heme binding, does not do so in vivo because it is sequestered in the blood before reaching the liver. In addition to the two classes described above, a third group of inhibitors that exhibit both of the abovementioned characteristics has been described. Their dual nature arises from their being substrates for metabolism initially, but the products of that metabolism either disrupt the protein structure (e.g., chloramphenicol, cyclophosphamide) or heme function. Heme function can be compromised by alkylation of the heme (e.g., dihydropyridines, unsaturated compounds such as olefins), which produces green pigments, by covalent linking of the heme to the protein (carbon tetrachloride), or by binding as ligands to the heme iron. This latter subgroup includes methylenedioxybenzene derivatives such as isosafrole and piperonyl butoxide and many amines including SKF 525A, troleandomycin, and related compounds. The complexes they form are classified as cytochrome P450 metabolic-intermediate complexes, and these can be detected by their characteristic ferrous state absorbance spectrum around the same wavelength as seen with the carbon monoxide, about 450 nm. Because both competitive and suicide (mechanism-based) inhibitors require active-site recognition, inhibitors can be extremely selective for the enzyme or isozyme they inhibit. Some such selective cytochrome P450 isozyme inhibitors are given in Table 3.4. For some compounds, the exact nature of their inhibition of cytochrome P450 remains obscure; ethanol is one such example. Despite all the information available on drug interactions and toxic episodes resulting from inhibition, it is likely that the mechanism(s) of many of them have yet to be fully elucidated. The biological consequences of inhibition of metabolism are two fold. In the acute phase, interactions can manifest themselves as either the potentiation of the biological effect of each, if metabolism results in inactivation, or protection from toxicity if toxicity arises from the bioactivation of the parent molecule. With chronic exposure, many agents generally considered as inhibitors (e.g., SKF 525A and clotrimazole) are also inducing agents (see Table 3.6). It appears that the compensation for long-term cytochrome P450 inhibition can be induction, perhaps as a response designed to circumvent the block. It should be noted, however, that more xenobiotic metabolizing enzymes than cytochrome P450 are induced by cytochrome P450 inhibitors. The induction seen with chronic exposure to inhibitors can thus result in drug interactions that are opposite to those listed as acute effects. In addition to substrate-binding (active) site inhibition, drug metabolizing capability can be reduced by cosubstrate or cofactor depletion (e.g., glutathione, SO4, NAD+), by their diversion to other biochemical pathways, or by an inhibition of enzymes responsible for their formation. In laboratory investigations, glutathione conjugation can be inhibited by either buthionine sulfoximine, which inhibits the synthesis of glutathione; or diethylmaleate, which sequesters available glutathione. Galactosamine, prior to its hepatotoxic effect can deplete UDPGA by sequestering UTP. For multicomponent reactions, the xenobiotic metabolism reaction can be inhibited at a distance (e.g., cytochrome P450 oxidations can be inhibited by the interruption of electron flow by heavy-metal ions, such as mercury, because the flavoprotein contains a more susceptible sulfhydryl group). Since xenobiotic metabolism is catalyzed by enzymes, many of the reactions can be inhibited nonselectively by protein denaturants such as heavy-metal ions and detergents, the degree of inhibition depending on the concentration. For enzymes that require a suitable membrane environment for activity, xenobiotics with lipid solvent properties can inhibit activity by destroying that necessary environment. Changes in lipid often lead to conformational changes that alter activity.



Not all inhibitors relate to enzymes located in membranes. There are inhibitors of nonmicrosomal xenobiotic-metabolizing enzyme activities that have toxicological importance and clinical usefulness. Disulfiram, an inhibitor of aldehyde dehydrogenase, is used as an adjunct to behavioral modification in the treatment of alcoholism since the unpleasant symptoms elicited by the accumulating acetaldehyde are sufficient to dissuade further ethanol ingestion. Monoamine oxidase inhibitors are available as drugs for the treatment of depression. If chemicals (e.g., tyramine) normally adequately metabolized by these enzymes are ingested simultaneously, they may accumulate to a sufficient concentration to cause severe toxicity (hypertensive crisis). Esterases where the active center contains a serine residue are readily inhibited by organophosphates and carbamates. Such inhibition results in the accumulation of other chemicals undergoing hydrolysis, particularly the endogenous substrate, acetylcholine, to the point of toxicity. Other Factors Responsible for Variations in Xenobiotic Metabolizing Enzymes

Animal Species and Strain Much has been made of species differences in xenobiotic metabolism, both for the purposes of extrapolation to humans and for exploiting differences in the understanding of species selective toxicities. Rodents have higher cytochrome P450 concentrations than other mammalian species, birds, and fish. Among mammals, cats are particularly deficient in UDP-glucuronosyltransferase activities and fish are deficient in all conjugations. This latter point has been attributed to the lesser need of aquatic animals to render foreign compounds to their most water-soluble form, since the volume of water that xenobiotics can diffuse into via the gills compensates for the lower partition coefficient. In comparison to most laboratory animal species, the rat is well endowed with sulfotransferase activity, a little lower in cytochrome P450 concentration, and relatively deficient in glutathione transferase activity (Table 3.8).

TABLE 3.8 Species and Strain Variations in Xenobiotic-Metabolizing Enzymesa Phase I

Phase II UGT

Species and Strain


pNA deM





575 470 55 — 225 325 265 200 415 75 85 75 60

140 25 — — 320 235 — 50 30 50 30 35 10

(vs. Male Sprague–Dawley Rat) Rabbitb Hamster Rat: Fischer Rat: Gunn Mouse: D2 Mouse: B6 Mouse: CF-1 Mouse: SWb Guinea pigb Catb Dogb Quailb Troutb

140 160 90 125 85 — 120 105 105 60 70 45 68

— 300 125 — 245 85 490 — — — — — —

250 155 115 30 60 325 — 65 180 5 335 220 5

275 235 115 120 170 90 — 220 95 50 355 25 20

a Abbreviations: pNA deM—p-nitroanisole demethylase; UGT = UDP-glucuronosyltransferase (two isozymes: GT1 and GT2); GST = glutathione S-transferase; ST = sulfotransferase; SW = Swiss Webster. b Gregus Z, Watkins JB, Thompson TN, Harvey MJ, Rozman K, Klaassen CD, Toxicol. Appl. Pharmacol. 67: 430 (1983).



In some species, while the rate of metabolism of a compound may be similar, the products may be different. Amphetamines are oxidatively metabolized by cytochrome P450 in both rabbits and rats but deamination products predominate in rabbits and phenyl hydroxylation products predominate in rats. Within an animal species there are also strain differences in xenobiotic-metabolizing capabilities. While a Fischer rat is very similar to a Sprague–Dawley rat, the Gunn strain of rat exhibits a deficiency in one of the two major classes of UDP-glucuronosyltransferase activity (GT1), yet its cytochrome P450 concentration is very similar to that in other strains. Two mouse strains (D2 and B6) that are very different in their response to induction by polycyclic hydrocarbons are very similar in their drug metabolizing enzymes activities before such exposure. Clear examples of human genetic polymorphisms are known. The best documented is the variation in acetylator phenotype and its consequences for the use of the antituberculosis drug, isoniazid (see Figure 3.12). Fast acetylators, occurring as a higher percent in Asian and Eskimo populations, tend to metabolize the drug and show liver toxicity from the reactive metabolite generated. Slow acetylators show the toxicity associated with the accumulation of unchanged drug, a peripheral neuropathy resulting from

Figure 3.12 The pathways of isoniazid metabolism and related toxicities.



pyridoxal phosphate depletion by the unmetabolized isoniazid. Slow acetylators have an increased risk of developing bladder cancer when exposed to arylamine compounds but are less represented than the overall population among colorectal cancer patients. Recently much attention has been directed at the number of isozymes, of human cytochrome P450 since variations in the amounts of the various isozymes that have some degree of substrate selectivity, could explain the variations in responses observed to standard doses of drugs. It could also partly explain the susceptibility of certain individuals to toxicity by chemicals that are bioactivated via oxidative metabolism. The cytochrome P450 2D6 polymorphism divides the Caucasian population into poor (5–10 percent) and extensive metabolizers of over 40 drugs and has been implicated in the development of some forms of cancer. Interestingly, extensive metabolizers are overrepresented in tobacco-smoke-associated lung cancer patients and underrepresented in leukaemia and melanoma patients. Polymorphism in the 2C19 form shows interracial differences, with an incidence of poor metabolizers of > 5 percent in Caucasian populations and 20 percent in Asian populations, although to date such differences have not been implicated in toxicities other than those arising from the slow metabolism of drugs. Gender Although there is no evidence of major gender differences in hepatic xenobiotic metabolism in humans, a major difference has been well documented for rats, particularly with respect to cytochrome P450. (Limited studies in humans suggest that females have slightly greater oxidative metabolism rates than do males.) It is also well documented that there is gender-dependent expression of certain cytochrome P450 isozymes in the rat (see Table 3.4). Sex differences would appear to be independent of the strain of rat and also apparently occur in at least one other rodent species, the hamster (Table 3.9). The mouse generally displays a higher cytochrome P450 concentration and activity in females. Phase II conjugations can also show sex differences, and these, like cytochrome P450, may be isozyme specific. Most of the gender-related differences in cytochrome P450 expression in rodents have been related to gender differences in growth hormone secretion. Although small differences are evident, the effects of inducers are similar between sexes of a species. From the examples given in Table 3.9, the phenobarbital-induced increases in cytochrome P450, glutathione S-transferase, and preferential increase in GT2 UDP-glucuronosyltransferase activity over GT1 UDP-glucuronosyltransferase activity are similar in males and females of both hamster and rat. Similarly, phenobarbital does not increase sulfotransferase activity in either sex of either species. TABLE 3.9 Gender Differences in Xenobiotic-Metabolizing Enzymesa Phase I


Species and Strain/Gender

Phase II

pNA P450






1.00 0.90 0.80

1.05 1.15 1.00

0.30 — 0.85

4.20 3.30 2.20 1.95

2.10 1.55 1.20 1.15

1.05 0.90 0.80 0.60

(Female vs. Male) Naive

Rat: SD Rat: Fischer Hamster

0.85 0.70 0.90


Rat: SD male Rat: SD female Hamster: male Hamster: female

2.35 1.65 1.60 1.40


0.75 0.70 0.85 0.35 0.80 0.80 (Induced vs. Naive) 6.60 3.75 3.26 3.35

1.25 1.50 1.20 1.30

Abbreviations: pNA deM = p-nitroanisole demethylase; UGT = UDPglucuronosyltransferase (two isozymes: GT1 and GT2); GST = glutathione S-transferase; ST = sulfotransferase; SD = Sprague–Dawley.



Maturation The age of a rat can also cause changes in its complement of drug metabolizing enzymes. Old age decreases the cytochrome P450 concentration and activity, particularly in the male. Of the phase II enzymes, a decline in sulfotransferase activity is apparent. Glutathione S-transferase appears marginally increased in old rats and lower in immature rats as compared to the mature adult. Neonatal animals generally exhibit lower drug-metabolizing activities than adults (Table 3.10). In humans, the activity of microsomal and perhaps nonmicrosomal xenobiotic-metabolizing enzymes is low in premature and neonatal infants. The effective glucuronidation of the bulky-type molecules, (e.g., morphine and chloramphenicol) appears to develop much later than does that for the planar phenol-type compounds. The activity of microsomal enzymes in neonates can be induced. Although there is evidence that the elderly have a decreased rate of hepatic microsomal metabolism of some drugs, the clinical importance of this is not clear because drug clearance remains unchanged as a consequence of changes in the volume of distribution of many drugs. The metabolism of many drugs and xenobiotic chemicals is fastest in adolescents.

Environment Human chemical drug metabolism can be influenced by the environment and diet. All diets contain naturally occurring nutrients and may also contain pesticide residues and food additives that are capable of altering the activity of chemical/drug-metabolizing enzymes. Among the more recent nutrient interactions is an inhibition of cytochrome P450 3A by grapefruit juice flavonoids, naringin and quercetin. Other flavonoids (catechin, myricetin, rutin, etc.) are able to induce phase II enzymes and protect against bioactivated intermediates (see discussion below). Even the quality of the diet can have an effect. Protein deficiency or diets deficient in essential fatty acids or certain vitamins (e.g., A, C, E) can decrease xenobiotic metabolism. Supplementation of diets with these nutrients (e.g., high protein) can increase chemical metabolism above normal. Administration of drugs and exposure to toxic compounds in burned carbonaceous material such as cigarette smoke and charcoal-broiled foods are among the best known modifiers of xenobiotic metabolizing capabilities. It is evident from the foregoing illustrations concerning the complex mixture of factors responsible for variations in xenobiotic metabolism that while a basic understanding of chemical metabolism can provide guidelines, prediction of actual situations is infinitely more difficult. While genetic differences between species may be obvious, subtle differences in physiology and diet all tend to confound extrapolations between experimental animals and, more critically, between experimental animal models and humans. These differences may even confound extrapolation between subgroups within the human population.

TABLE 3.10 Influence of Maturity on Rat Xenobiotic-Metabolizing Enzymesa Phase I

Phase II

pNA Age

Gender and Strain



1.00 1.40 0.65 0.90

1.05 1.00 0.70 0.80





— — 1.10 1.05

— — 0.60 0.70

(Video) Principles Of Toxicology

(vs. Young Adult) Immature Male Sprague–Dawleyb Female Sprague–Dawleyb Old Male Fischer Female Fischer

0.50 0.55 0.95 1.00

0.90 1.05 1.00 1.10

a Abbreviations: pNA deM = p-nitroanisole demethylase; UGT = UDP-glucuronosyltransferase (two isozymes; GT1 and GT2); GST = glutathione S-transferase; ST = sulfotransferase. b Chengelis CP, Xenobiotica 18: 1225 (1988).



Biotransformation: A Balance between Bioactivation and Detoxification The balance between bioactivation and inactivation (detoxification) can often determine whether a chemical is “ toxic” to cellular systems; toxic in this situation is usually defined as any cell damage leading to modified cell function, not necessarily cell and tissue death. Reactive intermediates may cause enzyme and protein modification and inactivation, membrane lipid peroxidation, and changes in DNA. Reactive intermediates are therefore implicated in carcinogenesis and tissue allergic responses in addition to tissue necrosis. Although enzymatic bioactivation of many environmental chemicals is a necessary step in the process of detoxification and elimination of the xenobiotic (Figure 3.4), the chemistry of the enzymatic product often precludes the elimination of the chemical without damage to critical cellular targets. When such reactive metabolites are produced in sufficient quantities, it is then that they cause cell and tissue damage. Thus, while the metabolism of certain toxic chemicals should represent a protective biological process, it frequently becomes a bioactivation process that is highly effective in the production of toxins within the organism. The reactive intermediates produced by the enzymes are often electrophiles, free radicals, or chemicals that can rearrange nonenzymatically to such intermediates. The chemical nature of the enzymatically generated reactive intermediates can be useful in providing a classification of toxic compounds (Table 3.11). The best documented example of enzymatic processes that produce reactive intermediates is the oxidation of chemicals by members of the cytochrome P450 superfamily; however, one should not overlook the fact that many xenobiotics are oxidized to nontoxic products such as phenols, N-oxides, alcohols, amines, aldehydes, and carboxylic acids. Thus, the same enzymes that detoxify one chemical can be responsible for the bioactivation of another. The reactive electrophilic intermediates produced by cytochrome P450 range from epoxides to iminium ions and include the formation of free radicals. In addition to cytochrome P450, reactive intermediate generation by a variety of other enzymes also occurs (Figure 3.13). The conjugation of xenobiotics with glutathione, glucuronic acid, sulfate, or acetate (phase II reactions) was originally thought to embody solely a detoxication process for drugs and environmental chemicals. In the vast majority of the examples that have been studied, these products of conjugation reactions are, in fact, detoxication metabolites. However, a significant number of studies with a variety of toxicants have shown that many of the conjugates are not innocuous. Glucuronidation has recently

TABLE 3.11 Classification of Toxicants by Reactive Intermediates Acyl glucuronides of Bilirubin, clofibric acid, diflunisal, indomethacin, tolmetin, valproic acid, zomipirac Carbonium ions 2-Acetylaminofluorene, dimethylnitrosamine, nitrosonornicotine, procarbazine, pyrolizidine alkaloids Epoxides Aflatoxins B1 and B2, benzo[a]pyrene, benzo[e]pyrene, benzy[b]fluoranthene, chrysene, 7,12dimethylbenz[a]anthracene, 3-methylcholanthrene Glutathione adducts of Chlorotrifluoroethylene, 1,2-dibromo-3-chloropropane, dibromoethane, N-(3,5-dichlorophenyl)succinimide, hexachlorobutadiene, tetrachloroethylene, tetrafluoroethylene, trichlorethylene, tris(2,3dibromopropyl)phosphate Imines Acetaminophen, amodiaquine, 3-methylindole, 2,6-dimethylaniline, ellipticine acetate, nicotine, phencyclidine Nitrenium 2-Acetylaminofluorene, 4-aminobiphenyl, 2-aminonaphthalene, 2-aminophenanthrene, benzidine Quinones Adriamycin, o-benzoquinone, p-benzoquinone, bleomycin, menadione, mitomycin c, 1,2-naphthoquinone, streptonigrin



Figure 3.13 The enzymatic generation of reactive intermediates.



been shown to be responsible for the production of glucuronides, which, via protein alkylation, can result in the formation of immunogens. The immune response mounted to these aberrant molecules can be highly toxic to organisms. Examples of toxic conjugates (acyl-linked glucuronides and glutathione adducts) are shown in Figure 3.13 and Table 3.11. Glutathione conjugates can also serve as transport forms of reactive intermediates. Methyl isocyanate, a highly reactive electrophile, is such an example. The glutathione conjugate is transported to sites distant from the initial absorption site to cause toxicity to other organs. The balance between detoxication and bioactivation of xenobiotics by metabolism enzymes can be dramatically changed by the induction or inhibition of the enzymes. Enzymes that are normally present at low levels, and therefore do not bioactivate toxicants to reactive intermediates, can become active participants in the toxicity of chemicals when the levels and activities of the enzymes are increased. Many examples of this situation exist. For example, induction of CYP2E1 by ethanol results in the greater bioactivation of hepatotoxins like CCl4 and acetaminophen or carcinogens such as dimethylnitrosamine. Although the toxicants can produce damage normally, their potency is greatly increased after induction of CYP2E1; specifically, toxicity is elicited at much lower doses because more of the chemical is oxidized to a reactive intermediate. Conversely, the toxicity of many chemicals can be ameliorated by induction of enzymes that are responsible for the detoxication of the compound. Bilirubin can cause significant central nervous system damage in neonates where the UDP-glucuronosyltransferase(s) that detoxify this naturally occurring heme breakdown product are present in low amounts. Inducing the levels of the necessary UDP-glucuronosyltransferase by drugs such as phenobarbital increases the glucuronidation of bilirubin and decrease its toxicity. In the same way that induction of bioactivation enzymes can increase toxicity and induction of detoxification enzymes can decrease toxicity, the inhibition of bioactivation enzymes or the inhibition of detoxification enzymes should decrease or increase toxicity, respectively. The carcinogenicity of complex mixtures of polycyclic aromatic hydrocarbons is sometimes found to be less than one would expect if the relative carcinogenicity of each component were summed. A probable reason for this decrease in toxicity lies in the inhibition, by components of the mixture, of the cytochrome P450 enzymes that bioactivate the carcinogens to their DNA-reactive intermediates. The mechanisms by which xenobiotics cause toxicity can be highly diverse, and elucidating the precise biochemical and chemical mechanisms that induce toxicity can be a difficult process. There are many tools available that can be used to evaluate toxic mechanisms. They include the use of animal species, gender, or cellular differences that vary widely in their response to the toxin. For example, naphthalene is highly toxic to mice when administered by the intraperitoneal route or by inhalation but is much less toxic to rats. Investigators have used this species differences to provide vital information about a cytochrome P450 (CYP2F2) that is highly expressed only in murine lung and is responsible for the bioactivation of this toxicant. Limonene causes severe renal toxicity to male rats but not female rats. The primary cause for the toxicity was eventually linked to the expression of a globulin that is not expressed to a significant degree in female rats. An example of the use of specific cellular targets is with the nephrotoxic glutathione conjugates of halogenated hydrocarbons, such as hexachlorobutadiene, which are selective for the proximal tubule cells of the nephron. Analysis showed that these cells contain high amounts of the enzyme C-S lyase, and it is this enzyme which is responsible for the production of the electrophilic intermediates from these toxicants. Similarly, it is the high content of monoamine oxidase B in dopamine-containing neurons linked with the cellular selectivity of the toxicity of MPTP that has enabled the mechanism of bioactivation of this toxicant to be elucidated.

3.3 SUMMARY By altering a portion of a chemical or by adding another molecule to it, drug-metabolizing enzymes can alter the toxicity of the chemical, its tissue-binding properties, and its distribution and duration within the body.



While the main benefit of biotransformation is to protect the body from attaining high chemical levels within the various tissues, the lack of specificity and predictability in the process sometimes leads to bioactivation or an increase in the toxicity of a chemical. Some such products are mutagenic and/or carcinogenic. Thus, metabolism of a toxic chemical has three possible outcomes, and it is the balance between these possible occurrences that determines the eventual outcome of the chemical exposure in question. 1. It may form a nontoxic metabolite. 2. It may generate a toxic metabolite that is subsequently detoxified. 3. It may generate a toxic metabolite that is not rendered harmless by detoxification before cellular and tissue injury have ensued. The primary organ functioning to metabolize chemicals is the liver. It is the first organ to be exposed to chemicals absorbed from the gut, the major portal of entry of xenobiotics. The characteristics of a person exposed to a toxic chemical that may alter the metabolism can include diet or nutritional status, age, gender, and/or hormonal status, and the genetic makeup of the individual. These characteristics may account for the interindividual variations observed in the human response to chemical exposure and the extent or type of toxicity observed in various model animals. A major determinant of metabolizing capabilities is prior exposure to chemicals that induce or inhibit the enzymes involved.

SUGGESTED READING Gibson, G., and P. Skett, Introduction to Drug Metabolism, 2nd ed., Chapman & Hall, London, 1993. Mulder, G. J., ed., Conjugation Reactions in Drug Metabolism: An Integrated Approach, Taylor & Francis, Bristol, PA, 1990. Ortiz de Montellano, P. R., ed., Cytochrome P450. Structure, Mechanism and Biochemistry, 2nd ed., Plenum Press, New York, 1995. Pacifici, G. M., and G. N. Fracchia, eds., Advances in Drug Metabolism in Man, European Commision, Luxembourg, 1995. Timbrell, J. A., Principles of Biochemical Toxicology, 2nd ed., Taylor & Francis, Bristol, PA, 1991. Witmer, C. M., R. R. Snyder, D. J. Jollow, G. F. Kalf, J. J. Kocsis, and I. G. Sipes, eds., Biological Reactive Intermediates IV; Molecular and Cellular Effects and Their Impact on Human Health, Plenum Press, New York, 1991.

4 Hematotoxicity: Chemically Induced Toxicity of the Blood HEMATOTOXICITY: CHEMICALLY INDUCED TOXICITY OF THE BLOOD


This chapter describes toxicities affecting blood. The following subjects are covered:

• The origin, formation, and differentiation of blood cells • Clinical tests used to evaluate hematotoxicity • Oxygen transport by erythrocytes (red blood cells or RBCs) and interference with oxygen • • • •

transport by drugs and chemicals Chemicals that affect the formation of red blood cells, platelets, and white blood cells (bone marrow suppression) Leukemias and lymphomas (cancers of the white blood cells) Neurological and cardiovascular toxicities caused by interference with oxygen utilization (e.g., cyanide, hydrogen disulfide) Medical treatment of hematotoxicity

This chapter describes common occupational and environmental chemicals and drugs that affect blood formation and function. Only the recognized chemically induced blood toxicities are included in this chapter, although there are many examples of anecdotal reports linking a large number of chemicals and drugs with hematotoxicity.

4.1 HEMATOTOXICITY: BASIC CONCEPTS AND BACKGROUND Hematotoxicity essentially involves two basic homeostatic functions: (1) RBC-mediated oxygen transport and (2) the production of red and white blood cells and platelets. Perhaps the earliest experiences with hematotoxicity involved the consumption of fava beans and the development of favism among people in the Mediterranean region. Favism is the development of acute hemolytic anemia in individuals with a deficiency in the red blood cell enzyme glucose-6-phosphate dehydrogenase following the ingestion of fava beans. It is also likely that methemoglobinemia, a blood condition characterized by cyanosis, was observed when individuals consumed well water containing large amounts of nitrates and nitrates. Chemically induced hematotoxicity has been reported in the medical literature for over a century. For example, the 1919 publication by Dr. Alice Hamilton, entitled Industrial Poisoning by Compounds of the Aromatic Series, described a number of blood toxicities that were commonly encountered in occupational settings such as benzene-induced bone marrow suppression, aniline and nitrobenzeneinduced methemoglobinemia, and hydrogen sulfide–induced effects. The major hematotoxicities encountered in the workplace involve benzene (bone marrow suppression and acute myelogenous leukemia), carbon monoxide (impairment of oxygen transport), aniline/nitrobenzene analogs Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




(hemolytic anemia, which reduces the oxygen transport capacity of blood), and hydrogen sulfide. Exposure to hydrogen sulfide can be a significant industrial hygiene concern in the refining of petroleum products and the biological degradation of silage (fermented corn, grain, etc., used to feed livestock) and sewage. As for the number of workers affected, benzene and hydrogen sulfide probably constitute the most significant risk factors for toxicity. Hematotoxicity is also an important concern in the administration of pharmaceuticals. For example, dapsone (used to treat leprosy) and primaquine (used to treat malaria) can produce a fatal hemolytic anemia in certain genetically predisposed individuals (those with a deficiency in glucose-6-phosphate dehydrogenase). Unfortunately, individuals most likely to require primaquine or dapsone therapy live in tropical areas of Africa, Asia, and the Mediterranean and are most likely to inherit a deficiency in glucose-6-phosphate dehydrogenase. Of widespread concern are the risks of bone marrow injury and suppression caused by cancer chemotherapeutics, complications that can often limit the administration of cancer-curing drugs. Another longstanding problem involves carbon monoxide poisoning, which results from exposure to improperly ventilated combustion products. Outside the workplace, the most common occurrences of hematotoxicity involve carbon monoxide poisoning, due to faulty gas heating, and adverse hematologic effects due to prescription medications. Fortunately, hematotoxicity is rarely encountered due to the resiliency of bone marrow, the redundancy of various hematologic controls and functions, and the implementation of more conservative occupational hygiene standards. However, when it occurs it is often life threatening. Likewise, examples of hematotoxicity resulting from exposure to environmental chemicals are relatively rare and generally involve foods or medications. Although hematotoxicity is not prevalent, it is useful for industrial hygienists, toxicologists, and occupational physicians to be aware of the chemicals that cause hematotoxicity, relevant signs and symptoms, and any antidotes and treatments that are available.

4.2 BASIC HEMATOPOIESIS: THE FORMATION OF BLOOD CELLS AND THEIR DIFFERENTIATION All blood cells originate from undifferentiated mesenchymal cells, which are located in the bone marrow. The various stages of blood cell formation are depicted in Figure 4.1. From stem cells, clones of immature blood cells differentiate along one of two pathways: the myelogenous series or the lymphocytic series. The myeloid series gives rise to erythrocytes, macrophages, platelets, neutrophils, eosinophils, and basophils. The lymphoid series gives rise to T (thymus) and B (Bursa) lymphocytes. Bone marrow production of blood cells is highly dependent on, and controlled by, a number of growth factors. Erythropoietin, a glycoprotein growth factor produced in the peritubular cells of the kidney, is essential for the differentiation and maturation of red blood cells. Under conditions of hypoxia (low oxygen), such as that occurring at higher altitudes or during anemia (a reduction in red blood cells or hemoglobin content) affliction, the release of erythropoietin by the kidney is enhanced. Conversely, the release of erythropoietin is inhibited by polycythemia (the increased number of circulating red blood cells) or hyperoxia. Other important glycoproteins that act alone or in conjunction with erythropoietin to control red blood cell formation include interleukins such as IL3, IL1, and IL2; granulocyte-macrophage colony stimulating factor (GM-CSF); insulin-like growth factor; and granulocyte colony stimulating factor. White blood cell formation also depends upon stimulation and control by various growth factors. IL3 stimulates all of the myeloid series cells. GM-CSF stimulates the formation of granulocytes and macrophages. Additionally, specific G-CSF and M-CSF proteins stimulate the granulocyte series or the macrophage series, respectively. These growth factors, unlike erythropoietin, are produced by various cells including T lymphocytes, macrophages, fibroblasts, and endothelial cells. All the growth factors work in concert to regulate different stages of myeloid and lymphoid differentiation and replication.


Figure 4.1 Formation of blood. Mature differentiated blood cells arise from a pluripotential stem cell via sequential complex interactions of regulatory molecules and cellular controls within the bone marrow environment. Disruption of the differentiation and maturation process gives rise to the various types of chemically induced blood deficits and/or leukemias/lymphomas.



When bone marrow is injured or suppressed, the number of specific types of blood cells (or all blood cells) may decline or even disappear. The decline in the number of cells from a specific blood cell lineage has its own diagnostic term and is based on the expected normal range that exists in healthy individuals. For example, a decrease in the normal number of circulating red blood cells leads to the clinical condition of anemia, a decrease in circulating platelets is known as thrombocytopenia, and a decrease in white blood cells is called leukopenia. Table 4.1 provides definitions for various clinical terms used to describe the abnormal number of circulating red blood cells, neutrophils, lymphocytes, and platelets. Some terms describe the same condition and may create confusion when used interchangeably. The suffix penia means an abnormal reduction, and the suffix cytosis refers to an abnormal excess. Changes in the number of circulating white blood cells provide an important diagnostic parameter for many diseases. Granulocytopenia, when the granulocyte count (primarily the neutrophils) falls to less than 1000 cells/mm3, may arise from chemical-induced bone marrow damage following administration of cancer chemotherapeutic drugs or the antibiotic chloramphenicol, antiinflammatory agents such as butazolidin, or exposure to benzene. When this occurs, an insufficient number of granulocytic cells are available to maintain the first line of defense against infectious agents, and recurrent infection is likely. Granulocytosis, or an increased number of circulating granulocytes (exceeding 10,000 cells/mm3), often occurs in patients with leukemia or can be triggered by an underlying infection. In cancers of the myeloid series, such as myelogenous leukemia, neutrophil numbers may exceed 30,000 cells/mm3. The leukemias generally consist of cells that lack the normal morphology and function of mature white blood cells (e.g., they resemble precursors of immature white blood cells found in the bone marrow). The cancer biology of the various types of leukemias (myeloid and lymphoid) and lymphomas, as well as multiple myeloma is quite complex. The types of leukemias and lymphomas are described in Table 4.2 according to the International Classification of Disease codes (frequently abbreviated the ICD). These cancers can differ in morbidity as well as clinical presentation, symptoms, management, and long-term survival.

TABLE 4.1 Definitions of Hematological Clinical Terms (Normal Adult) Clinical Term Anemia Aplastic anemia Agranulocytosis Granulocytopenia, neutropenia Granulocytosis Leukopenia Leukocytosis, neutrophilia Lymphopenia Lymphocytosis Eosinophilia Thrombocytopenia

Definition/Description A reduction in either the number or the volume of red blood cells, i.e., less than 3,500,000 RBC/mm3 or 14 g of hemoglobin per 100 mL of blood A cessation of the normal regenerative production of red blood cells in the bone marrow A reduction in the number of polymorphonuclear leukocytes (PMNs) less than 500/mm3 A reduction in the normal number of granulocytic leukocytes in the blood; normal granulocytes number around 3000–4000/mm3 An increase above normal of the normal number of circulating granulocytic leukocytes A reduction of the number of circulating leukocytes (white blood cells) below 5000 cells/mm3 An increase in the number of leukocytes, typically PMNs, above 10,000 cells/mm3 A reduction in the number of circulating lymphocytes less than the normal 2500/mm3 An increase in the number of circulating lymphocytes from their normal number of around 2500/mm3 Increase number of eosinophils above 200/mm3 Abnormal number of circulating platelets less than 250,000–500,000/mm3



TABLE 4.2 Leukemias and Lymphomas ICD No. ICD 200: lymphosarcoma and reticulosarcoma

ICD 201: Hodgkin’s disease ICD 202: other malignant neoplasms of lymphoid and histiocytic tissue ICD 203: multiple myeloma ICD 204: lymphoid leukemia ICD 205: myeloid leukemia ICD 206: monocytic leukemia ICD 207: erythroleukemia ICD 208: leukemia not otherwise specified

General Description Lymphocytic lymphomas involving resident lymphocytes in various tissues such as the liver, spleen, lung, skin, bone marrow, and gastrointestinal tract Lymph node cancers exhibiting Reed–Sternberg cells Involves resident macrophage cell types located in the peripheral tissues such as lymph nodes, spleen, skin, liver, and connective tissue Cancer of a specific plasma cell (antibody producing B lymphocyte) Cancer of circulating B and T lymphocytes Cancer of the myeloid series, usually granulocytes (neutrophils or PMNs) Cancers of the monocytic series involving the bone marrow–derived monoblast Cancer of RBC precursors

Source: International Classification of Disease (ICD), 9th ed.


Red Blood Cells (RBC): Erythrocytes During erythrocyte maturation, the blast forming unit-erythrocyte or BFU-E matures into the colony forming unit-erythrocyte (CFU-E), which finally evolves into a mature erythrocyte after extrusion of the nucleus and acquisition of hemoglobin and a mature cytoskeletal support system. Accelerated erythrocyte formation due to anemia, may result in the release of RBCs prior to the loss of the nucleus. These immature erythrocytes known as reticulocytes can be found circulating in the blood in increased numbers (i.e., reticulocytosis). Physicians may suspect anemia or a bleeding disorder if the number of reticulocytes exceeds their normal blood level (normal circulating reticulocyte levels are 5 percent of the total number of RBCs) or if other clinical parameters are abnormal, specifically, the number of RBCs, the percent volume of packed RBCs or the hemoglobin content. Once the red blood cell is released into circulation, it has a normal lifespan of 120 days. Damaged or senescent RBCs are sequestered in the spleen and destroyed by splenic macrophages. Mature RBCs are discoid-shaped and devoid of a nucleus and mitochondria. They are rich in the heme-containing protein, hemoglobin, which transports oxygen to peripheral tissues and carries carbon dioxide from the periphery to the lungs, where it is exhaled. Erythrocytes also provide a pH buffering function by converting carbon dioxide to carbonic acid via the enzymatic activity of carbonic anhydrase. Erythrocytes possess an outer membrane supported by a complex cytoskeletal system of various proteins that are essential for carrying out the normal functions of the RBC. Erythrocytes normally constitute around 40 percent by volume of whole blood and number around 3,500,000– 5,000,000 per cubic millimeter. The normal hemoglobin content of 100 mL of blood is 14 g. When the 120-day lifespan of erythrocytes is shortened to the extent that stimulated erythrocyte replacement cannot keep pace, varying degrees of anemia may result. There are numerous causes of anemia, including abnormal hemoglobin inherited in the form of sickle cell anemia or thalassemia; congenital nonspherocytic anemias involving defects in biochemical pathways in the RBC (e.g., glucose-6-phosphate dehydrogenase deficiency); congenital spherocytosis,



in which inherited defects in cytoskeletal membrane proteins predispose RBCs to hemolysis and destruction by the spleen; proxysmal nocturnal hemoglobinuria involving nighttime episodes of hemolysis; autoimmune-induced RBC destruction; chemical-induced RBC damage or destruction; infectious diseases such as malaria; and hypersplenism. Many of the underlying pathological conditions that give rise to the various types of anemia can be quite complex and require careful clinical assessment. Anemias may present with different RBC morphology (shape or appearance). The types of anemia and information regarding specific anemias are listed in Table 4.3. An essential characteristic of RBCs is their ability to undergo reversible shape changes or deformations as they travel through the narrow capillary beds. The reversible deformability involves a highly organized submembrane cytoskeletal scaffolding system. In this system, spectrin chains are attached to the membrane via ankyrin and further controlled by actin and numerous other cytoskeletal membrane proteins. Not surprisingly, if these cytoskeletal membranes are damaged by active oxygen (the high oxygen environment of the erythrocyte or redox cycling caused by chemicals, as discussed below) or by some other chemically induced perturbation or hereditary disease, the deformability of the erythrocyte is lost, and bizarre shape changes may ensue. Alterations in the normal cytoskeletal protein structure may also impart fragility to the RBC, inducing it to burst (hemolyze) under stressful conditions. In hereditary hemolytic anemias, familial mutations in various cytoskeletal membrane proteins induce premature destruction of the red blood cells. Chemical-induced hemolytic anemia is a similar condition in which exposure to certain chemicals results in damage to the cytoskeletal membrane proteins and the ultimate destruction of RBCs.

Platelets (or Thrombocytes) Platelets are formed in a similar manner to erythrocytes, originating from the same myeloid stem cells. Platelets are actually cell fragments that pinch off the megakaryocyte. Mature platelets are the smallest cells found in circulation, and once a platelet is formed, its circulating life span is only 9 days. Circulating platelets provide the first line of defense against blood loss by monitoring the integrity of the endothelial lining of arteries and veins. If a blood vessel ruptures, the connective tissue underneath the normally smooth, continuous, and nonreactive endothelial lining is exposed. The exposed collagen has a strong negative charge and is one of the most powerful inducers of platelet aggregation. Platelets bind to the collagen and to other platelets, and then the aggregated platelets begin

TABLE 4.3 Anemias Types Microcytic hypochromic anemia; small RBCs

Macrocytic; large RBCs

Normocytic; normal-sized RBCs

Description Results when red blood cells are formed having a low hemoglobin content (this occurs when red cells are produced rapidly in response to rapid blood loss); frequently an iron deficiency coincides with an inability to meet the demand of increased erythrocyte production, thereby causing a reduction in hemoglobin, hence hypochromic or reduced red pigmentation; iron-deficient diets may also produce hypochromic anemia; genetic diseases such as thalassemia result in abnormal hemoglobin production and microcytic anemia This anemia involves two different types: (1) megaloblastic and (2) nonmegaloblastic—it is an anemia resulting from a defect in DNA synthesis possibly secondary to folic acid or vitamin B12 deficiency; other etiologies may include liver disease and hypothyroidism that result in destruction of recticulocytes (reticulocytosis), or the myelodysplastic syndrome Early stages of microcytic and macrocytic anemia may present as normocytic; causes of normocytic anemia may be primary bone marrow failure such as aplastic anemia, hemorrhage, hemolytic anemia (immune or drug-induced), and mild hypothyroidism



to contract. The contraction stimulates the release of ADP (adenine diphosphate), serotonin, and other locally active compounds that initiate the clotting cascade mechanism, which culminates with the formation of a fibrin clot. Thus, platelets serve two purposes: (1) they create a physical barrier via formation of a plug to seal a vascular break, and (2) they initiate the intrinsic and extrinsic clotting mechanism involving proteins (clotting factors) synthesized by the liver. Platelet activation is normally a life-saving process. However, in the presence of atherosclerotic plaques (fatty, fibrous, calcified lesions in the arterial endothelium), repeated rupture of the blood vessel may lead to thickening and closure of the arterial lumen (strokes and hypertension). This can result in sudden death if a platelet clot occludes the coronary artery (myocardial infarction, starving downstream heart tissue of oxygen). Cigarette smoke–induced arteriosclerosis is a chemical toxicity that indirectly involves the platelets. In atherosclerosis accelerated by cigarette smoking, plaques, composed of a complex mixture of lipids (e.g., cholesterol), form underneath the normal smooth endothelial lining of the artery/arterioles. If the plaque ruptures, which it can do unpredictably, the connective tissue underneath the endothelial lining becomes exposed, and this event triggers platelet aggregation. If the platelet aggregation and fibrin deposition progresses to the point of occluding the blood vessel, an infarct occurs, and all tissue distal to the occlusion (platelet clot) dies from anoxia. The muscle that dies downstream of the clot (infarct or inclusion) is replaced by scar tissue, and hence the efficiency of the cardiac muscle is compromised. The scar may also impact normal electrical conduction pathways in the heart. Hence, platelets are involved in the final stages of cigarette-induced atherosclerosis, a condition that can lead to vascular disease, strokes, angina, and heart attacks. Overall, few toxicologically effects result from direct stimulation of platelet aggregation. On the other hand, aspirin and inhibitors of prostaglandin synthetase, such as ibuprofen, inhibit platelet aggregation and can prolong bleeding times. In individuals with bleeding disorders, inhibition of platelet aggregation can lead to excessive blood loss. However, the impairment of clotting is actually beneficial to individuals who are at risk of strokes, angina (ischemia of the cardiac muscle, causing pain and potential arrhythmias), and heart attacks (myocardial infarction). The most common platelets toxicity involves suppression of normal platelet number (thrombocytopenia). Alkylating agents used in the treatment of cancer are notorious for causing thrombocytopenia. For many of these cancer chemotherapeutics, their dosages are limited by potentially life-threatening thrombocytopenia, which can lead to hemorrhaging and death. The oncologist will frequently regulate the dose of alkylating agents based on the patient’s platelet count. As one would anticipate, hemorrhaging from the mucous membranes of the mouth, nose, and kidneys often reveals the onset of a deficiency in platelet-controlled clotting. Sometimes it is necessary to administer platelets or whole blood to treat a cancer patient who develops serious thrombocytopenia. Granulocytes or Neutrophils, Monocytes or Macrophages, Eosinophils, and Basophils Differentiation pathways similar to RBCs exist for neutrophils, basophils, eosinophils, and monocytes (which mature into macrophages after leaving the bloodstream). Neutrophils provide the first line of defense against bacterial invasion. They are also commonly referred to as polymononuclear neutrophils (or PMNs) because they possess a multilobed nucleus. They generally amplify in number in response to infectious agents, which gives the physician a diagnostic endpoint to suspect an infection; specifically, PMNs exceeding 10,000 mm3 in blood may indicate an infectious process. Neutrophils spend less than a day in circulation before attaching to vascular epithelial cells and migrating to extravascular locations in response to foreign invasion. Monocytes circulate in the blood for 3–4 days before migrating to reticuloendothelial tissues, such as the liver, spleen, and bone marrow, where they set up residence as fixed macrophages. Macrophages are recruited into an area of infection or injury and remove cellular debris and phagocytize pathogens. Macrophages may also act as antigen presenting cells (APCs) by presenting a digested antigen to T lymphocytes in order to activate cellular immunity (the immune response mediated by T lymphocytes). Overall, granulocytes (including neutrophils, eosinophils, and basophils) and monocytes or macrophages can act as phagocytic cells by physically attaching to foreign particles via receptor recognition. Following attachment, the macro-



phage engulfs (phagocytizes) the particle or foreign cell, and enzymatic processes within these cells facilitate the digestion of the engulfed particle. Eosinophils provide protection against infectious organisms by releasing proteolytic enzymes and active oxygen and conducting phagocytotic activities. An increased number of eosinophils in the blood and tissue is normally observed in allergic (atopic) individuals who suffer from chronic hay fever or asthma. However, in certain toxicities, such as the L-tryptophan eosinophilia myalgia syndrome (LTEMS), eosinophil excess resulted from contaminants that were present an over-the-counter amino acid sleep aid. In this case, the increase in eosinophils constituted a harmful autoimmune response. Basophils, when stimulated, release histamine, proteolytic enzymes, and inflammatory mediators. Toxicities involving basophils are almost non-existent.

4.4 THE LYMPHOID SERIES: LYMPHOCYTES (B AND T CELLS) The lymphoid series gives rise to cells involved in both humoral (B cells) and cellular (T cells) immunity. B cells function to produce antibodies, while T cells kill virus-infected cells and mediate the actions of other white blood cells. In the last 15–20 years (at the time of writing), considerable progress has been made toward understanding (1) the various types of T cells and how they differ in

Figure 4.2 Thymic Maturation of T-Lymphocytes. Immature T-lymphocytes pass through the various layers and cavities of the thymus gland while acquiring specific functional capabilities. These capabilities result, in part, from the acquisition of receptors expressed on the cell surface of the T-lymphocytes. T-lymphocytes are identified by the phenotype expression of specific receptors such as T-suppressor cells which express the CD8 receptor while not expressing the CD4 receptor. Conversely, T-helper cells are defined by expression of the CD4 receptor protein while lacking CD8 expression.



function and response to stimuli, (2) the unique membrane-bound T cell receptors responsible for antigen recognition, and (3) many of the complex events that regulate T-cell maturation. Although the process of T-cell production begins in the bone marrow, the immature pre–T cell must migrate to the thymus gland, via the bloodstream, for further development and differentiation. The thymus-dependent differentiation of T cells into specific subpopulations is governed by the expression of unique cell surface proteins or receptors known as cluster determinants. Specific types of T cells are defined by their cluster determinant repertoire, namely, CD4 for T-helper cells, CD8 for T-suppressor cells, and CD3 as a marker for all T cells. The cluster determinant expression (phenotype of the mature T cell) ultimately determines the precise function of the mature T cell that leaves the thymus (T helper, suppressor, memory, and killer cells for example). Thymic maturation of T cells involving the acquisition and deletion of specific cluster determinants is depicted in Figure 4.2. The post–bone marrow maturation of B cells in humans is not well understood. Like T cells, B lymphocytes may also be defined by their own distinct repertoire of cluster determinants (membrane proteins and protein receptors). Chemicals that affect T and B lymphocyte function are more appropriately discussed under the topic of immunotoxicity.

4.5 DIRECT TOXICOLOGICAL EFFECTS ON THE RBC: IMPAIRMENT OF OXYGEN TRANSPORT AND DESTRUCTION OF THE RED BLOOD CELL Two types of toxicities essentially affect red blood cells: (1) competitive inhibition of oxygen binding to hemoglobin and (2) chemically induced anemia in which the number of circulating erythrocytes is reduced in response to red blood cell damage. Inhibition of oxygen transport is the more commonly observed toxicity directly affecting the RBC. Carbon monoxide, cyanide, and hydrogen sulfide bind to hemoglobin and can potentially interfere with its ability to transport oxygen. Carbon monoxide directly inhibits oxygen binding to hemoglobin, which can result in a spectrum of adverse effects ranging from mild subjective complaints to life-threatening hypoxia. The mechanism underlying carbon monoxide toxicity is one of the simpler toxicological phenomena, in terms of its binding to the iron molecule in hemoglobin. However, some of the consequences of carbon monoxide poisoning, such as cardiovascular and neurological effects, are much more complex and occasionally are associated with somewhat controversial outcomes (i.e., delayed neurological injury, such as memory loss, purportedly expressed as a reduction in neuropsychological test performance). While cyanide and hydrogen sulfide can also bind to the heme iron in hemoglobin, their significant toxic effects relate to inhibition of mitochondrial energy production. Chemically induced methemoglobin and methemoglobinemia associated with hemolytic anemia occur by two different mechanisms. The first mechanism involves oxidation of hemoglobin (methemoglobin formation). The second mechanism involves oxidation of hemoglobin coupled to modification of RBC membrane proteins causing the RBC to be recognized as foreign by the immune system. The ultimate outcome of either type of toxicity is hypoxia.

Oxygen Transport: Hemoglobin An understanding of hemoglobin’s protein structure is necessary to fully appreciate how carbon monoxide, cyanide, and hydrogen sulfide bind to the heme iron of hemoglobin and prevent oxygen from binding or being released. Hemoglobin (Hb) consists of four separate peptide chains (two alpha and two beta peptides). Each peptide chain is irregularly folded and surrounds a porphyrin molecule (protoporphyrin) located in a hydrophobic pocket. An iron molecule is located in the center of the protoporphyrin ring and forms a coordinate–ligand bond with oxygen. The oxidation state of the iron atom is an important factor in oxygen binding. Oxygen can only bind to iron when it is in its ferrous state (+2 oxidation state). Oxidation of the iron atom to its ferric state (+3 oxidation state) produces methemoglobin, a derivative of hemoglobin that does not form a coordinated ligand bond with oxygen.



The binding of one molecule of the iron molecule induces a conformational change in the tertiary structure of hemoglobin. The resulting shape change increases hemoglobin’s affinity for subsequent oxygen binding. Thus, the binding of each oxygen molecule facilitates the binding of the next in a process known as positive cooperativity. Positive cooperativity produces a characteristic sigmoidal shaped oxygen binding curve (Figure 4.3), demonstrating that a disproportionately greater increase in oxygen binding to hemoglobin occurs as the oxygen concentration (PO ) of blood increases by only a small amount. 2 The release of oxygen from hemoglobin is caused by the tissue PO gradient from the arteriole to 2 the venous side. The release of the first oxygen molecule facilitates the release of the second oxygen molecule, and so on. The first oxygen is released in an area of relatively higher tissue oxygen content whereas the remaining oxygen is released in areas further down the capillary bed where the tissue oxygen content is lower. The transit of oxygenated blood from the arteriole to the venous side results in a loss of approximately 5 mL of oxygen from each 100 mL of blood. Hypoxia Hypoxia is defined as a decreased concentration of oxygen in inspired air, oxygen content in arterial blood, or oxygen content in tissue. Anoxia, on the other hand, is the complete absence of oxygen.

Figure 4.3 Characteristic sigmoid shape of the hemoglobin-O2 dissociation curve. To liberate the first 4–5 ml of oxygen, the partial pressure of oxygen within the blood must drop about 60 mm Hg. The second 5 ml of oxygen per 100 ml of blood is liberated with a drop in pressure of only 15 to 20 mm Hg.



Hypoxia can result from a variety of conditions including anemia; a reduction in the iron carried by the RBC; ischemia (physical barrier to blood flow) caused by occlusion or vasoconstriction of an artery; or by an increased oxygen affinity (shift to the left of the oxygen-hemoglobin binding curve), which reduces the release of oxygen. In situations involving oxygen-deficient atmospheres, the blood oxygen concentration can drop to a level in which the central nervous system and cardiovascular system risk impairment. Hypoxia typically occurs when workers enter confined spaces where the atmospheric oxygen (normally at 21 percent) is too low to sustain the oxygen saturation of hemoglobin above 80 percent. Under circumstances of reduced oxygen delivery to the lungs, serious cardiovascular and central nervous system impairment can develop. The symptoms range in severity from euphoria to loss of consciousness, seizures, and cardiac arrhythmias. Hemoglobin oxygen saturation less than 80 percent results in a sense of euphoria, impaired judgment, and memory loss. As the oxygen desaturation of hemoglobin worsens, the extent of central nervous system effects increase. If oxygen pressure drops to 30 mm Hg, a level corresponding to approximately 55–60 percent oxygen saturation, consciousness may be lost. Individuals with ischemic heart disease, such as atherosclerotic coronary vascular disease, may be more sensitive to hypoxic conditions than in healthy individuals. Individuals with atherosclerosis may be more prone to hypoxia-induced ischemia, which may lead to arrhythmias (irregular electrical conduction in the heart) or ischemia-like pains (i.e., chest pain encountered during angina or a myocardial infarction). Subjects with serious atherosclerosis of the cerebral vasculature are more likely to develop CNS impairment related to hypoxia than are healthy subjects. Hence, hypoxia resulting from either low oxygen concentrations or interference with oxygen transport must be assessed according the subject’s cardiovascular status. Physiological adaptations can affect oxygen’s affinity for hemoglobin, especially when chronic low levels of hypoxia are present. 2,3-Diphosphoglycerate (or 2,3-bisphosphoglycerate) concentrations increase within RBCs under conditions of chronic hypoxia (e.g., high altitudes, various anemias). By complexing with deoxygenated hemoglobin, 2,3-diphosphoglycerate decreases hemoglobin’s affinity for oxygen and facilitates oxygen release in peripheral tissues. This is illustrated by a shift to the right in the oxygen–hemoglobin binding curve. An increase in hydrogen ions (acidity of blood) also causes the hemoglobin–oxygen binding curve to shift to the right. Hydrogen ions are generated when carbon dioxide (formed during respiration or oxygen consumption) is converted to bicarbonate. When the hydrogen ions are then taken up by hemoglobin, oxygen is released. Consequently, ischemic tissue, where the oxygen tension is low and carbon dioxide is high, is benefited by the increased oxygen release that occurs in the presence of hydrogen ions. Conversely, if the oxygen–hemoglobin binding curve is shifted to the left, oxygen binds more avidly to hemoglobin. When this occurs, an even lower tissue oxygen concentration is required before oxygen can be released.


Carbon Monoxide Carbon monoxide binds to hemoglobin, decreasing the available sites for oxygen while increasing the binding affinity of the oxygen that is already bound. The hemoglobin binding affinity of carbon monoxide is explained by the Haldane equation, named after the scientist who studied the effects of carbon monoxide in the late 1800s. The carbon monoxide binding affinity is denoted by M in the Haldane equation [HbCO] M[PCO] = [HbO2] [PO2] where HbCO represents the percentage of carboxyhemoglobin (the carbon monoxide-hemoglobin complex), and HbO2 represents the percentage of hemoglobin bound by oxygen. PCO and PO represent 2



the carbon monoxide and oxygen tensions (percentages), respectively, in air. In humans, M is reported to be anywhere from 210 to 245, demonstrating that carbon monoxide binds to hemoglobin approximately 200 times more avidly than oxygen. To illustrate this further, consider the concentration of carbon monoxide that is required to decrease hemoglobin oxygenation by 50 percent. First, the concentrations of carboxyhemoglobin and oxyhemoglobin are equal so that the left side of the equation becomes one, that is, 50 percent of the blood exists as HbCO and 50 percent exists as HbO2. The equation then simplifies to [PCO] =

[PO2] M

Since the normal oxygen concentration in air is 21 percent, solving the Haldane equation yields a carbon monoxide concentration in air of 0.1 percent or approximately 1000 ppm. When equilibrium is achieved, an individual inhaling 1000 ppm of CO will develop 50 percent carboxyhemoglobin and a serious hypoxic situation. Compounding this hypoxia is the increased binding affinity of oxygen caused by carbon monoxide inhibiting the release of oxygen to tissue. The ability of carbon monoxide to decrease oxygen’s binding to hemoglobin and to increase oxygen’s affinity for hemoglobin is called the Haldane effect. Low level background carboxyhemoglobin concentrations of 1.0% or less normally exist in the blood as a result of porphyrin metabolism. Cigarette smoking increases carboxyhemoglobin concentrations to as much as 5–10 percent in heavy smokers—two packs per day, for example. If exposure to carbon monoxide from exogenous sources increases carboxyhemoglobin concentrations to around 20 percent, subjective complaints may be reported. As shown in Table 4.4, the adverse effects of carbon monoxide are concentration dependant. Significant hypoxia caused by carboxyhemoglobin has been reported to produce brain injury resulting in a Parkinson’s disease-like condition, cognitive impairment, and serious neurobehavioral changes. Some of these neurological sequelae may not be apparent for a number of days or even weeks following exposure. The more severe neurological effects generally occur in only a few individuals under circumstances of life-threatening hypoxia. Fortunately, most individuals with mild to moderate carbon monoxide poisoning experience complete recovery. Recovery is aided by the use of 100% oxygen or hyperbaric oxygen treatment along with supportive measures. Assessment of carbon monoxide poisoning is typically performed in the emergency room. However, significant time may lapse between the exposure, emergency room arrival and the determination of carboxyhemoglobin. The time between loss of consciousness or serious clinical effects and drawing

TABLE 4.4 Carboxyhemoglobin and Effects Carboxyhemoglobin (% Hemoglobin Saturation with Carbon Monoxide) 0.3–0.7 1–5

2–9 16–20 20–30 30–40 50+ 67–70

Effect Background concentrations due to endogenous production of carbon monoxide Increase in blood flow via compensating mechanisms such as increased heart rate or increased contractility (these concentrations are typically observed in cigarette smokers) A reduction in exercise tolerance and an increase in the visual threshold for light awareness Headache; abnormal visual responses A throbbing headache accompanied by nausea, vomiting, and a decrease in finemotor movement Severe headaches, nausea, vomiting, and weakness Coma and convulsions Lethal if not aggressively treated



a blood sample may lead to a significant decline in carboxyhemoglobin levels, especially if the patient is treated with oxygen. If breathing room air, the half-life of carboxyhemoglobin is approximately 4–5 hours; if 100 percent oxygen is administered, the half-life can be reduced by 4-fold. If hyperbaric oxygen treatment is implemented, the normal half-life can be shortened 10-fold. Hence, carboxyhemoglobin determinations at the time of medical intervention may not accurately gauge the extent of carboxyhemoglobin that occurred during exposure. Carbon monoxide is generated by incomplete combustion; automobile fumes and cigarettes are among the most familiar sources. A common example of carbon monoxide poisoning occurs from heating with natural gas, especially natural gas of lesser quality, namely, individuals overcome by carbon monoxide heating homes with wet natural gas and without proper ventilation. Another potential occupational, as well as environmental, source of carbon monoxide results from methylene chloride exposure. Methylene chloride is metabolized to carbon monoxide by cytochrome P450 enzymes resulting in elevations in carboxyhemoglobin levels. Case reports have documented elevated carboxyhemoglobin levels in individuals stripping furniture with methylene chloride–based paint strippers. Physical activity, which increases the respiratory rate, will increase the amount of inhaled methylene chloride and the resulting carboxyhemoglobin levels. The current OSHA standard and ACGIH TWA for carbon monoxide is 50 and 25 ppm, respectively. By the time equilibrium is achieved, 50 ppm carbon monoxide will produce carboxyhemoglobin concentrations of approximately 5–6 percent after about 8 h of exposure. It should be noted that the binding equilibrium of carbon monoxide is not achieved instantaneously but requires time.

4.7 INORGANIC NITRATES/NITRITES AND CHLORATE SALTS In blood, an equilibrium exists between ferrous and ferric hemoglobin. The oxygen-rich environment surrounding the RBC continually oxidizes hemoglobin to methemoglobin. Since methemoglobin does not bind and transport oxygen, the accumulation of methemoglobin is detrimental. Therefore, the accumulation of methemoglobin is prevented by the enzymatic reduction of ferric iron to ferrous iron via the enzyme methemoglobin reductase (also known as diaphorase). The normal concentration of methemoglobin is generally 0.5 percent or less, which produces no adverse health effects. Methemoglobin formation results in a noticeable change in the color of blood from its normal red color to a brownish hue. In humans and animals, significant methemoglobinemia creates a bluish discoloration of the skin and mucous membranes. Mild to moderate concentrations of methemoglobin can be tolerated, and low levels of less than 10 percent may be asymptomatic, except for a slightly bluish color imparted to the mucous membranes. If blood methemoglobin concentrations achieve 15–20 percent of the total hemoglobin, clinical symptoms of hypoxia can develop, and above 20 percent, cardiovascular and neurological complications related to hypoxia may ensue. Methemoglobin concentrations exceeding 40 percent are often accompanied by headache, dizziness, nausea, and vomiting, and levels surpassing 60 percent may be lethal. Other than supportive care to maximize oxygen transport, such as oxygen administration, little can be done to treat methemoglobinemia. One available antidote is the intravenous administration of methylene blue, which provides reducing equivalents to methemoglobin reductase and thus facilitates the reduction of methemoglobin back to ferrous hemoglobin. Inorganic nitrites such as sodium nitrite (NaNO2) and chlorates (ClO−3 ) oxidize ferrous hemoglobin (Fe2+) to ferric-hemoglobin (Fe3+ or methemoglobin). Nitrite and chlorate directly oxidize hemoglobin; nitrate, however, must first be reduced to nitrite by nitrifying bacteria in the gut. Exposures to nitrates, nitrites, and chlorates occur mostly in industrial settings or from contaminated drinking water. The typical concentrations of nitrate and nitrite found in foods and drinking water, however, do not present a risk in terms of methemoglobin production. If the rate of hemoglobin oxidation caused by nitrite/chlorate exceeds the capacity of methemoglobin reductases activity, a buildup in methemoglobin results. The oxidative conversion of hemoglobin to methemoglobin by nitrites and chlorates, combined with the reduction of methemoglobin back to ferrous-hemoglobin, is referred to as a redox cycle.



Nitrates, in addition to their conversion to methemoglobin-causing nitrite, can produce a complex array of vascular changes, such as venous pooling (reduced blood return to the right side of the heart). Episodes of as venous pooling aggravate the clinical complications of methemoglobinemia; cardiac output is reduced and tissue hypoxia is exacerbated. Thus, nitrate toxicity presents a complicated clinical picture that integrates the production of methemoglobin with a reduction in blood perfusion to tissues most in need of oxygen. The hematologic hazards regarding nitrite and chlorate, on the other hand, appear to be limited to the direct oxidation of hemoglobin to methemoglobin.

4.8 METHEMOGLOBIN LEADING TO HEMOLYTIC ANEMIA: AROMATIC AMINES AND AROMATIC NITRO COMPOUNDS Aromatic amines and nitro compounds such as aniline and nitrobenzene cause methemoglobinemia by initiating a redox cycle in the RBC. The aromatic amines and nitro compounds are important building blocks in the dye, pharmaceutical, and agricultural chemical industries. Aromatic amines are also important structural components of numerous prescription medications. By in large, amineinduced methemoglobinemia and hemolytic anemia develop most often following treatment with antibiotics such as dapsone and primaquine, pharmaceuticals used to treat infectious diseases such as leprosy and malaria, respectively. However, unlike those for nitrites and chlorates, the potential hazards of aromatic amines are not limited to methemoglobinemia. RBC changes occurring during or after methemoglobin formation may result in damage to the RBC membrane. The damaged RBCs are recognized by splenic macrophages, which remove and destroy them. Hemolytic anemia can result if the number of red blood cells destroyed exceeds the bone marrow’s capacity to replenish them; for example, by amplification of RBC production in response to increased release of erythropoietin. Reactive metabolite(s) of the parent aromatic amine compound, formed via cytochrome P450 metabolism, are also capable of causing methemoglobinemia and hemolytic anemia. Aromatic nitro compounds, like inorganic nitrate, must first be reduced to their respective aromatic amine by gut bacteria before being metabolized to an arylhydroxylamine. It is the N-hydroxyl metabolite that is directly responsible for initiating hemoglobin oxidation via a redox cycle. The redox cycle results in the formation of reactive oxygen species in the RBC (i.e., hydrogen peroxide). The reactive oxygen species oxidize proteins in the RBC cytoskeleton and damage the RBC membrane by crosslinking adjacent proteins. The crosslinked proteins can be visualized in the form of Heinz bodies, which consist of hemoglobin covalently linked to cytoskeletal proteins on the inner side of the red blood cell membrane. RBC membrane damage may alter the normal RBC discoid morphology, depicted in Figure 4.4 for dapsone N-hydroxylamine-induced RBC morphology alteration. These spike-shaped RBCs produced by dapsone N-hydroxylamine are known as echinocytes. Other abnormally shaped RBCs that may result from exposure to various aromatic amines include anisocytes (asymetrically shaped RBCs); spherocytes (round RBCs); elliptocytes (ellipse or egg-shaped RBCs); sickle cell–shaped RBCs (known as drepanocytes); acanthocytes, which are round RBCs with irregular spiny projections; and stomatocytes, which are RBCs with a slit-like concavity. A senescent (aging) signal may appear on the membrane of the damaged red blood cell and serve as a recognition sign for the spleen. In effect, active oxygen species produced during redox cycles appear to cause premature aging and altered morphology of RBCs, leading to their early removal from circulation. Another name for redox cycle formation of reactive oxygen species and damage to the RBC is “ oxidative stress.” Instances of aromatic amine-induced methemoglobinemia and hemolytic anemia are rather rare. This is due to their low volatility, which reduces inhalation exposure, and the fact that many of the amines are used in the form of salts, which reduces their potential for dermal absorption. The free amines, however, are dermally absorbed and can pose a potential hazard if directly contacted by the skin. Another serious concern with exposure to aromatic amines is their potential to induce hemorrhagic cystitis (bleeding from bladder damage) and bladder cancer.



Figure 4.4 Dapsone N-hydroxylamine-induced Red Blood Cell Changes. Chemically induced damage to red blood cells is typically expressed as changes in red blood cell shape. The altered shape (morphology) results from damage to the cytoskeleton proteins or lipid membrane of the red blood cell.

Exposure to aromatic amines can be potentially life-threatening to individuals with a deficiency in the enzyme glucose-6-phosphate dehydrogenase (G6PDH). Individuals with deficiencies in G6PDH are limited in their ability to maintain sufficient levels of reduced glutathione (GSH) in their RBC. GSH acts as a scavenger of active oxygen species such as hydrogen peroxide that are formed during the redox cycle. In the event of oxidative stress caused by an activated redox cycle, these individuals cannot withstand the oxidations of GSH to GS-SG (glutathione disulfide) or GS-S-protein, and they will suffer oxidative damage to the RBC membrane proteins at lower blood concentrations of N-hydroxy metabolites than normal people. G6PDH deficiency exists primarily among individuals of Mediterranean, African, and Asian decent. It can be tested for prior to initiation of drug therapy that may cause hemolytic anemia. Treatment modalities for chemically induced hemolytic anemia are limited. Methylene blue may be administered to maximize the ability of methemoglobin reductase, which reduces methemoglobin back to ferrous hemoglobin. Transfusions may be necessary to replace red blood cells prematurely sequestered and destroyed by the spleen. There is no information on the use of glutathione-related antidotes such as N-acetyl cysteine. Mild conditions of chemically induced hemolytic anemia are not fatal and can be treated supportively. The extent of hemolysis induced by aromatic amines is proportionate to the amount of methemoglobin produced. Therefore, low levels of methemoglobin, in the general range of 20–30 percent or less, do not typically lead to extensive removal of red blood cells and anemia.

4.9 AUTOIMMUNE HEMOLYTIC ANEMIA Hemolysis mediated by the immune system occurs via a different mechanism than direct oxidative stress. In this instance, the drug or drug metabolites cause immunoglobulins (either IgG or IgM) to nonspecifically or specifically bind to the RBC. The IgG or IgM bound to the RBC attracts complement. Complement then binds to the surface of the RBC and initiates destruction of the RBC membrane. The damage to the RBC imparts fragility to the membrane, the RBC ruptures in the vasculature, and hemoglobin is released. The intravascular hemolysis can provoke disseminated intravascular coagulation (DIC), a serious consequence of autoimmune hemolytic anemia. Free circulating hemoglobin can also induce renal failure when it is excreted by the kidney. Hence, autoimmune hemolytic anemia, primarily caused by prescription drug use, can result in a battery of serious health effects. Fortunately, only a few drugs are known to provoke this adverse drug reaction, and most cases are considered idiosyncratic.




Bone Marrow Suppression A variety of industrial chemicals and pharmaceuticals can cause partial or complete bone marrow suppression. Pancytopenia occurs when all cellular elements of the blood are reduced. Bone marrow suppression may be reversible or permanent depending on the chemical agent and the extent of exposure. Clinical signs of bone marrow suppression include bleeding, caused by a reduction in platelet counts; anemia, which leads to fatigue and altered cardiovascular/respiratory parameters; and a heightened susceptibility to various infectious processes. The cells with the shorter lifespans are the first to disappear, such as the platelets, which have a circulating lifespan of only 9 or 10 days. Therefore, if the bone marrow injury involves the myeloid series, thrombocytopenia (i.e., reduction in the number of blood platelets) bleeding is one of the first complications to be observed. Patients with this condition are at a high risk for life-threatening internal hemorrhaging. Examples of occupational chemicals and drugs reported to cause blood dyscrasias (e.g., thrombocytopenia, neutropenia, pancytopenia) are listed in Table 4.5. Some of the examples listed in Table 4.5 are based solely on case reports and do not represent confirmed examples of chemically induced bone marrow suppression. For example, the evidence regarding the effects of pentachlorophenol-induced aplastic anemia is based on isolated case reports. However, larger clinical studies performed on wood-treatment workers and animal testing show no evidence of bone marrow suppression. Hence, concrete evidence that pentachlorophenol causes bone marrow suppression is lacking. In contrast to the numerous single case reports weakly implicating specific chemicals with bone marrow suppression, there are a number of chemicals with undisputed bone marrow toxicity; benzene is the best-known example among industrial chemicals. A known marrow suppressant, benzene was experimentally used decades ago to inhibit the uncontrollable production of leukemia cells. Today, the cancer chemotherapeutics are the most frequently encountered causes of bone marrow suppression. The alkylating agents used in cancer chemotherapy are notorious for damaging the bone marrow and are often administered until the patient develops bone marrow suppression. In this event, the administration of further chemotherapy is discontinued, or more commonly, a reduction in the dose of the anticancer drug is attempted. Oncologists constantly monitor the patient’s platelet and white blood cell count in order to evaluate the bone marrow suppressive effects of the cancer chemotherapy. Chloram-

TABLE 4.5 Chemicals Reported to Cause Bone Marrow Suppression Benzene (an important industrial solvent and component of many refined petroleum products, e.g., gasoline) Procainamide (an antiarrhythmic used to control cardiac arrhythmias) Methyldopa (an antihypertensive used to treat high blood pressure) Sulfasalazine (a drug used to treat inflammatory bowel disease)

Isoniazid (a mainstay antibiotic in treating tuberculosis) Diphenylhydantoin (an important drug used in the treatment of epilepsy)

Chloramphenicol (an important antibiotic used to treat resistant bacterial infections)

Phenylbutazone (antinflammatory used to treat arthritic conditions)

Allopurinol (a drug used to treat gout)

Tolbutamide (used to treat maturity onset or type II diabetes)

Sulindac (antiinflammatory agent)

Aminopyrine (analgesic and antipyretic) Sodium valproate (used to treat Alkylating and antimetabolite certain epileptic conditions) (cancer chemotherapy agents, e.g., nitrogen mustard, 5fluorouracil, cytoxan) Cephalothin (a cephalosporin Gold (used as an antiflammatory antibiotic) agent in arthritic conditions) Pentachlorophenol (a chemical used Carbamazepine (used to treat to treat wood) certain forms of epilepsy)



phenicol is an important antibiotic used to combat strains of bacteria that are resistant to first-line antibiotics; however, it bears a well-recognized risk of bone marrow suppression. The drug phenylbutazone, once commonly used as an antiinflammatory agent for treating arthritic conditions, is now conservatively prescribed for only a few weeks at a time in order to reduce the chance of developing bone marrow suppression. The marrow suppressive effects of benzene were described long before benzene was established as a cause of acute myelogenous leukemia (AML). Benzene’s suppressant effects range from mild and reversible to lethal, namely, life-threatening aplastic anemia or pancytopenia. Preleukemia or myelodysplasia, often viewed as a precursor to leukemia, is characterized by abnormal morphology of blood cells and may be associated with chronic bone marrow suppression. Evidence of benzeneinduced bone marrow suppression in humans is based on many studies. One of the most highly publicized cases involved the Ohio Pliofilm workers of the 1940s and 1950s. The Pliofilm worker studies provided evidence that benzene exposures exceeding 50–75 ppm were associated with reductions in white blood cell counts. More recent evidence, using more sophisticated cell counting methods, suggest that lymphocytes may be the most sensitive target of benzene . Metabolite(s) of benzene is (are) the actual cause(s) of marrow suppression. Benzene is metabolized by hepatic cytochrome P450 mixed function oxidases. Benzene is a substrate of cytochrome P450 IIE, which is one of the many isozymes among the family of cytochrome P450 mixed-function oxidases. Benzene oxide, the first intermediate in CYP 2EI-mediated metabolism, is converted into a number of metabolites including phenol, hydroquinone, and muconic acid/muconaldehyde (see Figure 4.5). Two benzene metabolites not shown in Figure 4.5 include catechol and trihydroxy benzene. In the bone marrow, myeloperoxidase further oxidizes phenolic metabolites of benzene to form free radicals capable of damaging the bone marrow.

Figure 4.5 Benzene’s Metabolism. Benzene is both bioactivated and detoxified via a number of different enzymatic-mediated steps. The bioactivated metabolites of benzene, such as hydroquinone and muconaldehyde, disrupt the various stages of blood formation in the bone marrow gives rise to any number of blood dyscrasias, myelodyplastic syndrome, and acute myelogenous leukemia.



Not all of benzene’s metabolites cause bone marrow suppression. Phenol, hydroquinone, catechol, trihydroxy benzene, and muconaldehyde act in concert to cause bone marrow changes; by themselves these metabolites have less marrow toxicity. The precise mechanism by which these metabolites act alone or in concert to cause marrow suppression is uncertain, although these issues are among the topics of ongoing research.

4.11 CHEMICAL LEUKEMOGENESIS Bone marrow injury may promote the development of myelodysplastic syndromes and acute myelogenous leukemia. Therefore, by damaging the bone marrow, benzene, chloramphenicol, and cancer chemotherapeutic agents increase an individual’s risk of contracting bone marrow cancer. However, critical issues regarding exposure and dose, as well as the weight of evidence from epidemiologic and animal studies all influence the relative risk. The cancer biology of chemically induced leukemia is complex, and one or more of the following mechanisms may be involved in the progression toward myelodysplastic syndrome and possibly leukemia: bioactivation of the parent molecule to reactive intermediates, disruption of marrow physiology (e.g., interference with the mitotic spindle), inhibition of topoisomerases, formation of DNA adducts, chromosomal alterations, oncogene activation, and suppressor gene inactivation. As with any chemically induced cancer, benzene-induced AML follows a continuum or progression of events that includes repeated bone marrow injury and suppression, chromosomal changes, the development of dysplastic and metaplastic features, and the ultimate expression of AML. Awareness of benzene’s role in acute myelogenous leukemia came later. The mounting evidence of benzene-induced leukemias finally surfaced in the 1970s and 1980s with publication of NIOSHconducted studies of Pliofilm workers from two plants in Ohio, the Turkish studies of shoemakers who used glues with high benzene content, and Italian rotogravure printers who used benzene-containing solvents, for example. The collective findings of these studies clearly implicated benzene in the development of AML. Recent Chinese studies suggest that other hematological tumors may occur at a higher incidence among benzene-exposed workers. However, the evidence for benzene-induced hematological cancers, other than AML, is still rather limited, and further investigations are needed. Industries with less benzene exposure (average benzene exposures of 1 part per million or less among refinery workers, rubber workers, and gasoline workers) and chemical workers exposed to benzene have not shown an increased incidence of AML. effects. The Pliofilm studies have contributed information involving exposure estimates and dose–response relationships. For instance, Rinsky et al. (1988) first proposed a risk–exposure relationship: OR = e(0.0126 × ppm⋅year) where OR stands for the odds ratio for leukemia relative to the unexposed workers in a worker who has acquired a specific cumulative ppm⋅year of benzene exposure. Based on this risk model a background exposure of 0.1 ppm⋅year generates a risk estimate no greater than background, that is, an odds ratio of 1.0. More recent studies of the Pliofilm workers have concluded that a threshold level of benzene as high as 50 ppm (or even higher) must be exceeded before a significant risk of developing AML exists. In summary, epidemiologic evidence has established that high-level benzene exposure in the workplace is associated with an increased risk of acute myelogenous leukemia. Clear evidence that a causal relationship exists between benzene exposure and AML comes primarily from the studies on Pliofilm workers. When these studies are further evaluated for a dose–response relationship, the level of occupational exposure that bears a significant risk may be 50 ppm or greater. There is no sound evidence that benzene causes other types of cancer such as other types of leukemia, non-Hodgkin’s lymphoma, or solid tumors such as lung cancer. Currently, the OSHA standard of 1.0 ppm should provide adequate protection against both benzene-induced bone marrow depression and a risk of AML.



4.12 TOXICITIES THAT INDIRECTLY INVOLVE THE RED BLOOD CELL Two important chemicals interact with blood, and yet their toxicological effects directly involve the nervous and cardiovascular system. Both cyanide and hydrogen sulfide bind to the heme portion of hemoglobin. At toxic dosages, however, they first inhibit energy production by mitochondrial heme oxidase. Heme oxidase contains a porphyrin ring such as hemoglobin, which is essential for transporting electrons during oxidative phosphorylation. Cyanide and hydrogen sulfide are respiratory poisons that shut down energy production in cells carrying out aerobic metabolism. The selectivity of hydrogen sulfide and cyanide’s apparent toxicity (on the nervous and cardiovascular system) is related to the high oxygen and energy demands of these two tissues. It has been suggested that carbon monoxide toxicity also affects the electron transport chain in the mitochondria.

4.13 CYANIDE (CN) POISONING Cyanide inhibits cytochrome oxidase, thus halting electron transport, oxidative phosphorylation, and aerobic glucose metabolism. Inhibition of glucose metabolism results in the buildup of lactate (lactic acidemia) and the increase in the concentration of oxygenated hemoglobin in venous blood returning to the heart. Increased oxyhemoglobin in the venous circulation reflects the fact that oxygen is not being utilized in the peripheral tissues. The most serious consequences of oxidative phosphorylation inhibition are related to neurological and cardiovascular problems, including adverse neurological sequelae, respiratory arrest, arrhythmia, and cardiac failure. Cyanide exposure can occur via inhalation of hydrogen cyanide gas or through ingestion of sodium or potassium cyanide. Approximately 100 mg of sodium or potassium cyanide is lethal. Sublethal doses of cyanide are quickly metabolized to thiocyanate via the enzyme rhodenase (a sulfurtransferase): Na2S2O3 + CN– → SCN– + Na2SO3 The detoxification of cyanide to thiocyanate is facilitated by adding the substrate sodium thiosulfate, which reacts with cyanide through the action of rhodenase. Thiocyanate (SCN–) is a relatively nontoxic substance eliminated in the urine.

4.14 HYDROGEN SULFIDE (H2S) POISONING Hydrogen sulfide also inhibits mitochondrial respiration by inhibiting cytochrome oxidase thus halting the production of adenosine triphosphate, or ATP. Central nervous system effects ranging from reversible CNS depression to loss of consciousness and death may occur. Cardiac effects may include alterations in the rhythm and contractility of the heart. Less serious consequences of hydrogen sulfide include irritation, inflammatory changes, and edema of the mucous membranes of the eyes, nose, throat, and respiratory tract. The ppb odor threshold for hydrogen sulfide (i.e., the rotten-egg odor) in normal individuals far precedes concentrations causing adverse health effects, and for a short period of time can serve as a warning signal. Hydrogen sulfide exposure can occur around sewers and petroleum refinery wastestreams and in situations involving natural gas production or fermentation, such as with manure or silage (fodder for livestock stored in silos). Fortunately, most individuals are relatively sensitive to the odor of hydrogen sulfide and can detect it at ppb air concentrations, which provides an early warning. However, odor fatigue occurs with time and may result in a serious exposure if the individual remains in an area containing high or increasing concentrations of hydrogen sulfide. There are reports of individuals who are rapidly rendered unconscious and die from exposures to high levels of hydrogen sulfide, such as those exceeding 1000 ppm. For example, there are documented episodes



of workers who collapse and died within minutes of entering silos storing silage. Table 4.6 lists increasing air concentrations of hydrogen sulfide and the effects that may result from exposure at each level. The current OSHA acceptable ceiling concentration for hydrogen sulfide is 20 ppm, with maximum 10-min peak concentrations of 50 ppm allowed over an 8 h workshift (29 CFR, Part 1910). The American Conference of Governmental and Industrial Hygienists recommend a time-weighted average (TWA) exposure of 10 ppm. Once absorbed from the lungs, hydrogen sulfide is rapidly metabolized in the blood and liver. A series of enzymatic and non-enzymatic pathways convert hydrogen sulfide (the sulfide anion) to thiosulfate and then sulfate, which is eliminated from the body. If a blood sample is drawn shortly after exposure, elevated blood concentrations of sulfide can be detected. However, in general, blood sulfide determinations are usually forgotten during an emergency since the immediate concern is to treat the patient. The delay between exposure and blood sampling is usually too long to determine the blood sulfide concentration that was responsible for the observed acute effects. Furthermore, blood sulfide determination is usually considered a specialty analysis that must be conducted by laboratories outside the hospital. The term sulfhemoglobin has been used to describe hemoglobin with unique spectral characteristics distinguishable from simple methemoglobin. Sulfhemoglobin spectral changes were originally observed when hydrogen sulfide was bubbled through whole blood. The observation between high concentrations of hydrogen sulfide in the test tube and sulfhemoglobin formation has led to the misconception that hydrogen sulfide poisoning also produces measurable sulfhemoglobin (often used as a biomarker of hydrogen sulfide poisoning). However, this is an erroneous concept since sulfhemoglobin formation requires concentrations of hydrogen sulfide that far exceed those required to completely shut down oxidative phosphorylation. Thus, sulfhemoglobin determinations are not useful in verifying toxicity or lethality caused by hydrogen sulfide exposure.

TABLE 4.6 Dose-Response Relationship for Hydrogen Sulfide Air Concentrations of Hydrogen Sulfide (parts per million) 0.022 0.025–0.13 0.3 0.77 3–6 20 20–30 150 200 250 500

700–1000 5,000

Effect No odor. Noticeable to minimally detectable odor. Distinct odor. Generally perceptible. Quite noticeable, offensive, moderately intense. OSHA acceptable ceiling level. Strong intense odor but not intolerable. Olfactory nerve paralysis and mucous membrane irritation. Less intense odor due to eventual sensory fatigue. Prolonged exposure may cause pulmonary edema. Increasing mucous membrane irritation. Dizziness over a few minutes to severe central nervous system impairment and unconsciousness if inhaled for more than a few minutes. Increasing mucous membrane irritation. Unconsciousness may develop rapidly followed by respiratory paralysis and death within minutes. Increasing mucous membrane irritation. Imminent death.



4.15 ANTIDOTES FOR HYDROGEN SULFIDE AND CYANIDE POISONING Unfortunately, there are no failproof antidotes for hydrogen sulfide poisoning, although methods that induce methemoglobinemia have been suggested. In instances of cyanide poisoning, and occasionally hydrogen sulfide exposure, the administration of nitrite in the form of amyl nitrite or intravenous sodium nitrite is recommended to purposely convert the patient’s blood to a safe-level of methemoglobin. Methemoglobin has a very strong binding affinity for cyanide and hydrogen sulfide. The relative large amount of methemoglobin binds up and acts as a sink to remove cyanide or hydrogen sulfide from cellular spaces and the mitochondria. Once bound to methemoglobin, cyanide and hydrogen sulfide are no longer available to bind to (and thus inhibit) cytochrome oxidase, an mitochondrial enzyme essential to the aerobic metabolism of glucose. The chemicals are eventually released into the blood where they can be metabolized to thiocyanate (in the case of cyanide) and sulfite/sulfate (in the case of hydrogen sulfide). The ability of methemoglobin to trap cyanide and hydrogen sulfide is illustrated in Figure 4.6.

Figure 4.6 Schematic depiction of the electron transport chain through which the oxidation of NADH derived from sugar metabolism generates ATP. Both the cyanide (CN–) and hydrogen sulfide (HS–) anions bind to and inhibit cytochrome oxidase. However, both anions also bind the Fe+++ ion methemoglobin (MetHb) formed by the oxidation of hemoglobin with nitrate (NO−2 ). CN-MetHb denotes cyanmethemoglobin; HS-MetHb denotes sulfmethemoglobin.



4.16 MISCELLANEOUS TOXICITIES EXPRESSED IN THE BLOOD Lead poisoning may affect normal red blood cell parameters. For one, lead interferes with heme synthesis in the liver, which can lead to anemias. This interference results in the accumulation of protoporphyrin, a heme precursor that is measurable in the form of zinc protoporphyrin in the blood. The term basophilic stippling is often associated with RBCs that are prematurely destroyed in response to lead-induced anemia. Basophilic stippling is characterized by various-sized purple granules that are microscopically observed within the RBC. The purple granules are comprised of pyrimidine compounds that accumulate because lead inhibits erythrocyte pyrimidine-5-nucleotidase, the enzyme responsible for the normal degradation of these pyrimidine nucleotides. The apparent blood lead threshold affecting porphyrin biochemistry is around 25–30 µg/dL and the threshold for affecting hemoglobin is around 50 µg/dL. Treatment of lead poisoning generally involves chelation therapy with drugs such as penicillamine, EDTA, Dimercarpol, or BAL (British anti-lewisite). A number of chemicals affect the formation and action of clotting factors. Many of these chemicals inhibit clot formation and are extremely useful as anticoagulants in individuals with atherosclerotic cardiovascular and cerebrovascular disease. Thus, the anticoagulants aid in the prevention of heart attacks and strokes. For example, the drug warfarin effectively reduces circulating clotting factors within a few hours to days following treatment. Warfarin’s mechanism of action involves the antagonism of vitamin K, which is involved in the carboxylation of clotting factor proteins. Anticoagulants such as warfarin are also used as pesticides. Several additional rodenticides include difenacoum, chlorphacinone, and brodifacoum. Unless ingested, these anticoagulants are relatively safe since they are nonvolatile and cannot be absorbed through the skin. Poisoning by the anticoagulants usually occurs in infants and suicide cases. In the clinical setting, physicians monitor the patient’s clotting times to control for the desired therapeutic effect and to avoid excessive anticoagulation, which could result in a fatal hemorrhage. Vitamin K is the recommended antidote for treating individuals poisoned by anticoagulants.

4.17 SUMMARY Hematotoxicity involves a wide range of effects ultimately affecting oxygen delivery, maintenance of a viable immune or clotting system, and cancer. It is fortunate that hematotoxicity is a relatively uncommon occurrence. Overall, the real concern regarding bone marrow injury is related to benzene exposure in occupational settings. Benzene exposure in the workplace has dramatically declined since the days of the Pliofilm workers and before, and currently, there is little evidence to suggest that existing occupational settings pose a risk of AML. The threshold for benzene-induced AML has been reported to range of 0.1–50 ppm, although the actual concentration which poses a serious threat is still heavily debated. On the other hand, benzene exposure resulting from ingestion of ppb concentrations in ambient air or drinking water does not pose a risk of AML or any other hematopoietic tumors. In general, hematotoxicity is an occupational concern since the exposures and doses of chemicals required to cause a toxic response cannot be achieved from the low levels found in the environment (i.e., ppb air concentrations). The exceptions, of course, are carbon monoxide poisonings, which frequently occur in home settings, or toxicities from medications, such as chemotherapeutic agents used to treat cancer.

REFERENCES AND SUGGESTED READING Ellenhorn’s Medical Toxicology Diagnosis and Treatment of Human Poisoning. Matthew J. Ellenhorn editor, 2nd edition. Williams & Wilkins, Baltimore, (1997). Fishbeck, W. A., J. C. Townsend, and M. G. Swank, “ Effects of chronic occupational exposure to measured concentrations of benzene.” J. Occup Med. 20(8): 539–542 (1978).



Hamilton, A., “ Industrial poisoning by compounds of the aromatic series.” J. Ind. Hygi. 200–212 (1919). Hancock, G., A. E. Moffitt, Jr., and E. B. Hay, “ Hematological findings among workers exposed to benzene at a coke oven by-product recovery facility,” Arch. Environ. Health 39(6): 414–418 (1984). Kipen, H. M., R. P. Cody, K. S. Crump, B. C. Allen, and B. D. Goldstein, “ Hematological effects of benzene: A thirty-five year longitudinal study of rubber workers,” Toxicol. Ind. Health 4: 411–430 (1988). Peterson, J. E., and R. D. Stewart, “ Absorption and elimination of carbon monoxide by inactive young men.” Arch. Environ. Health 21: 165–171 (1970). Rinsky, R. A., A. B. Smith, R. Hornung, T. G. Filloon, R. J. Young, A. H. Okun, and P. J. Landrigan, “ Benzene and Leukemia. An epidemiologic risk assessment,” N. Engl. J. Med. 316: 1044–1050 (1987). Stewart, R. D., “ The effects of low concentrations of carbon monoxide in man,” Scand. J. Respir. Dis. Suppl. 91: 56–62 (1974). Yin, S.-N., Q. Li, Y. Liu, F. Tian, C. Du, and C. Jin. “ Occupational exposure to benzene in China,” Br. J. Ind. Med. 44: 192–195 (1987).

5 Hepatotoxicity: Toxic Effects on the Liver HEPATOTOXICITY: TOXIC EFFECTS ON THE LIVER


This chapter will familiarize the reader with

• • • • • •

The basis of liver injury Normal liver functions The role the liver plays in certain chemical-induced toxicities Types of liver injury Evaluation of liver injury Specific chemicals that are hepatotoxic

5.1 THE PHYSIOLOGIC AND MORPHOLOGIC BASES OF LIVER INJURY Physiologic Considerations The liver is the largest organ in the body, accounting for about 5 percent of total body mass. It is often the target organ of chemical-induced tissue injury, a fact recognized for over 100 years. While the chemicals toxic to the liver and the mechanisms of their toxicity are numerous and varied, several basic factors underlie the liver’s susceptibility to chemical attack. First, the liver maintains a unique position within the circulatory system. As Figure 5.1 shows, the liver effectively “ filters” the blood coming from the gastrointestinal tract and abdominal space before this blood is pumped through the lungs and into the general circulation. This unique position in the circulatory system aids the liver in its normal functions, which include (1) carbohydrate storage and metabolism; (2) metabolism of hormones, endogenous wastes, and foreign chemicals; (3) synthesis of blood proteins; (4) urea formation; (5) metabolism of fats; and (6) bile formation. When drugs or chemicals are absorbed from the gastrointestinal tract, virtually all of the absorbed dose must pass through the liver before being distributed through the bloodstream to the rest of the body. Once a chemical reaches the general circulation, regardless of the route of absorption, it is still subject to extraction and metabolism by the liver. The liver receives nearly 30 percent of cardiac output and, at any given time, 10–15 percent of total blood volume is present in the liver. Consequently, it is difficult for any drug or chemical to escape contact with the liver, an important factor in the role of the liver in removing foreign chemicals. The liver’s prominence causes it to have increased vulnerability to toxic attack. The liver can particularly affect, or be affected by, chemicals ingested orally or administered intraperitoneally (i.e., into the abdominal cavity) because it is the first organ perfused by blood containing the chemical. As discussed in Chapter 2, rapid and extensive removal of the chemical by the liver can drastically reduce the amount of drug reaching the general circulation—termed the first-pass effect. Being the first organ Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




Figure 5.1 The liver maintains a unique position within the circulatory system.

encountered by a drug or chemical after absorption from the gastrointestinal tract or peritoneal space also means that the liver often sees potential toxicants at their highest concentrations. The same drug or chemical at the same dose absorbed from the lungs or through the skin, for example, may be less toxic to the liver because the concentrations in blood reaching the liver are lower, from both dilution and distribution to other organs and tissues.



A second reason for the susceptibility of the liver to chemical attack is that it is the primary organ for the biotransformation of chemicals within the body. As discussed in Chapter 3, the desired net outcome of the biotransformation process is generally to alter the chemical in such a way that it is (1) no longer biologically active within the body and (2) more polar and water-soluble and, consequently, more easily excreted from the body. Thus, in most instances, the liver acts as a detoxification organ. It lowers the biological activity and blood concentrations of a chemical that might otherwise accumulate to toxic levels within the body. For example, it has been estimated that the time required to excrete one-half of a single dose of benzene would be about 100 years if the liver did not metabolize it. The primary disadvantage of the liver’s role as the main organ metabolizing chemicals, however, is that toxic reactive chemicals or short-lived intermediates can be formed during the biotransformation process. Of course, the liver, as the site of formation of these bioactivated forms of the chemical, usually receives the brunt of their effects.

Morphologic Considerations The liver can be described as a large mass of cells packed around vascular trees of arteries and veins (see Figure 5.2). Blood supply to the liver comes from the hepatic artery and the portal vein, the former normally supplying about 20 percent of blood reaching the liver and the latter about 80 percent. Terminal branches of the hepatic artery and portal vein are found together with the bile duct (Figure 5.2). In cross section, these three vessels are called the portal triad. Blood is collected in the terminal hepatic venules, which drain into the hepatic vein. The functional microanatomy can be viewed in different ways. In one view, the basic unit of the liver is termed the lobule. Blood enters the lobule Bile canaliculi


Hepatic artery Bile ductule

Central vein

Opening of sinusoid

Hepatic lamina

Fenestration in lamina

Portal vein Figure 5.2 Hepatic architecture, showing arrangement of blood vessels and cords of liver cells. Reproduced with permission from Textbook of Human Anatomy, Second Edition, C.V. Mosby Co., St. Louis, MO, 1976.



from the hepatic artery and portal veins, traverses the lobule through hepatic sinusoids, and exits through a hepatic venule. In the typical lobule view, cells near the portal vein are termed periportal, while those near the hepatic venule are termed perivenular. The hepatic venule is visualized as occupying the center of the lobule, and cells surrounding the venule are sometimes termed centrilobular, while those farther away, near the portal triad, are called peripheral lobular. Rappaport proposed a different view of hepatic anatomy in which the basic anatomical unit is called the simple liver acinus. In this view (Figure 5.3, left), cells within the acinus are divided into zones. The area adjacent to small vessels radiating from the portal triad is zone 1. Cells in zone 1 are first to receive blood through the sinusoids. Blood then travels past cells in zones 2 and 3 before reaching the hepatic venule. As can be seen in Figure 5.3, zone 3 is roughly analogous to the centrilobular region of the classic lobule, since it is closest to the central vein. Zone 3 cells from adjacent acini form a star-shaped pattern around this vessel. Zone 1 cells surround the terminal afferent branches of the portal vein and hepatic artery, and are often stated as occupying the periportal region, while cells between zones 1 and 3 (i.e., in zone 2) are said to occupy the midzonal region. A modification of the typical lobule and acinar models has been provided by Lamers and colleagues (1989) (Figure 5.3, right). Based on histopathologic and immunohistochemical studies, they propose that zone 3 should be viewed as a circular, rather than star-shaped, region surrounding the central vein. Zone 1 cells surround the portal tracts, and zone 1 cells from adjacent acini merge to form a reticular pattern. As with the Rappaport (1979) model, cells in zone 3 may be described as centrilobular (matching closely the classic lobular terminology), cells in zone 1 as periportal, and the cells in zone 2 in between are called midzonal. Each of these viewpoints has in common a recognition that the cells closest to the arterial blood supply receive the highest concentrations of oxygen and nutrients. As blood traverses the lobule, concentrations of oxygen and nutrients diminish. Differences in oxygen tension and nutrient levels are reflected in differing morphology and enzymatic content between cells in zones 1 and 3. Consistent with their greater access to oxygen, hepatocytes in zone 1 are better adapted to aerobic metabolism. They have greater respiratory activity, greater amino acid utilization, and higher levels of fatty acid oxidation. Glucose formation from gluconeogenesis and from breakdown of glycogen predominate in zone 1 cells, and most secretion of bile acids occurs here. On the other hand, most forms of the biotransformation enzyme cytochrome P450 are found in highest concentrations in zone 3 cells. As the site of biotransformation for most drugs and chemicals, zone 3 cells have greatest responsibility for their detoxification. This also means that zone 3 cells are often the primary targets for chemicals that are bioactivated by these enzymes to toxic metabolites in the liver.

Figure 5.3 Alternative views of the liver acinus. Reproduced with permission from Lamers et al., 1989.



Figure 5.4 Liver section from mouse given an hepatotoxic dose of acetaminophen. With acetaminophen, liver cell swelling and death characteristically occurs in regions around the central vein (Zone 3, arrow); cells near the portal triad (Zone 1, arrow head) are spared.

There are several types of liver cells. Hepatocytes, or parenchymal cells, constitute approximately 75 percent of the total cells in the human liver. They are relatively large cells and make up the bulk of the hepatic lobule. By virtue of their numbers and their extensive xenobiotic metabolizing activity, these cells are the principal targets for hepatotoxic chemicals. The sinusoids are lined with endothelial cells. These cells are small but numerous, making up most of the remaining cells in the liver. The hepatic microvasculature also contains resident macrophages, called Kupffer cells. Although comparatively few in number, these cells play an important role in phagocytizing microorganisms and foreign particulates in the blood. While these cells are a part of the liver, they are also part of the immune



system. They are capable of releasing reactive oxygen species and cytokines, and play an important role in inflammatory responses in the liver. The liver also contains Ito cells (also termed fat-storing cells, parasinusoidal cells, or stellate cells) which lie between parenchymal and endothelial cells. These cells appear to be important in producing collagen and in vitamin A storage and metabolism.

5.2 TYPES OF LIVER INJURY All chemicals do not produce the same type of liver injury. Rather, the type of lesion or effect observed is dependent on the chemical involved, the dose, and the duration of exposure. Some types of injury are the result of acute toxicity to the liver, while others appear only after chronic exposure or treatment. Basic types of liver injury include the anomalies described in the following paragraphs.

Hepatocellular Degeneration and Death Many hepatotoxicants are capable of injuring liver cells directly, leading to cellular degeneration and death. A variety of organelles and structures within the liver cell can be affected by chemicals. Principal targets include the following: 1. Mitochondria. These organelles are important for energy metabolism and synthesis of ATP. They also accumulate and release calcium, and play an important role in calcium homeostasis within the cell. When mitochondria become damaged, they often lose the ability to regulate solute and water balance, and undergo swelling that can be observed microscopically. Mitochondrial membranes can become distorted or rupture, and the density of the mitochondrial matrix is altered. Examples of chemicals that show damage to hepatic mitochondria include carbon tetrachloride, cocaine, dichloroethylene, ethionine, hydrazine, and phosphorus. 2. Plasma Membrane. The plasma membrane surrounds the hepatocyte and is critically important in maintaining the ion balance between the cytoplasm and the external environment. This ion balance can be disrupted by damage to plasma membrane ion pumps, or by loss of membrane integrity causing ions to leak in or out of the cell following their concentration gradients. Loss of ionic control can cause a net movement of water into the cell, resulting in cell swelling. Blisters or “ blebs” in the plasma membrane may also occur in response to chemical toxicants. Examples of chemicals that show damage to plasma membrane include acetaminophen, ethanol, mercurials, and phalloidin. 3. Endoplasmic Reticulum. The endoplasmic reticulum is responsible for synthesis of proteins and phospholipids in the hepatocyte. It is the principal site of biotransformation of foreign chemicals and, along with the mitochondria, sequesters and releases calcium ions to promote calcium homeostasis. As discussed in Chapter 3, hepatic biotransformation enzyme activity is substantially increased in response to treatment or exposure to a variety of chemicals. Many of these enzymes, including cytochrome P450, are located in the endoplasmic reticulum, which undergoes proliferation as part of the enzyme induction process. Because the endoplasmic reticulum is the site within the cell of most oxidative metabolism of foreign (xenobiotic) chemicals, it is also the site where reactive metabolites from these chemicals are formed. This makes it a logical target for toxicity for chemicals that produce injury through this mechanism. Morphologically, damage to the endoplasmic reticulum often appears in the form of dilation. Examples of chemicals that show damage to endoplasmic reticulum include acetaminophen, bromobenzene, carbon tetrachloride, and cocaine. 4. Nucleus. There are several ways in which the nuclei can be damaged by chemical toxicants. Some chemicals or their metabolites can bind to DNA, producing mutations (see Chapter 12). These mutations can alter critical functions of the cell leading to cell death, or can contribute to malignant transformation of the cell to produce cancer. Some chemicals appear to cause activation of endonucleases, enzymes located in the nucleus that digest chromatin material. This leads to uncontrolled digestion of the cell’s DNA—obviously not conducive to normal cell functioning. Some chemicals



cause disarrangement of chromatin material within the nucleus. Morphologically, damage to the nucleus appears as alterations in the nuclear envelope, in chromatin structure, and in arrangement of nucleoli. Examples of chemicals that produce nuclear alterations include aflatoxin B, beryllium, ethionine, galactosamine, and nitrosamines. 5. Lysosomes. These subcellular structures contain digestive enzymes (e.g., proteases) and are important in degrading damaged or aging cellular constituents. In hepatocytes injured by chemical toxicants, their numbers and size are often increased. Typically, this is not because they are a direct target for chemical attack, but rather reflects the response of the cell to the need to remove increased levels of damaged cellular materials caused by the chemical. Not all hepatocellular toxicity leads to cell death. Cells may display a variety of morphologic abnormalities in response to chemical insult and still recover. These include cell swelling, dilated endoplasmic reticulum, condensed mitochondria and chromatin material in the nucleus, and blebs on the plasma membrane. More severe morphological changes are indicative that the cell will not recover, and will proceed to cell death, that is, undergo necrosis. Examples of morphological signs of necrosis are massive swelling of the cell, marked clumping of nuclear chromatin, extreme swelling of mitochondria, breaks in the plasma membrane, and the formation of cell fragments. Necrosis from hepatotoxic chemicals can occur within distinct zones in the liver, be distributed diffusely, or occur massively. Many chemicals produce a zonal necrosis; that is, necrosis is confined to a specific zone of the hepatic acinus. Table 5.1 provides examples of drugs and chemicals that produce hepatic necrosis and the characteristic zone in which the lesion occurs. Figure 5.4 shows an example of zone 3 hepatic necrosis from acetaminophen. Confinement of the lesion to a specific zone is thought to be a consequence of the mechanism of toxicity of these agents and the balance of activating and inactivating enzymes or cofactors. Interestingly, there are a few chemicals for which the zone of necrosis can be altered by treatment with other chemicals. These include cocaine, which normally produces hepatic necrosis in zone 2 or 3 in mice, but in phenobarbital-pretreated animals causes necrosis in zone 1. Limited observations of liver sections from humans experiencing cocaine hepatotoxicity are consistent with this shift produced by barbiturates. The reason for the change in site of necrosis with these chemicals is unknown. Necrotic cells produced by some chemicals are distributed diffusely throughout the liver, rather than being localized in acinar zones. Galactosamine and the drug methylphenidate are examples of chemicals that produce a diffuse necrosis. Diffuse necrosis is also seen in viral hepatitis and some forms of idiosyncratic liver injury. The extent of necrosis can vary considerably. When most of the cells of the liver are involved, this is termed massive necrosis. As the name implies, this involves destruction of most or all of the hepatic acinus. Not all the acini in the liver are necessarily affected to the same extent, but at least some acini will have necrosis that extends across the lobule from the portal triad to the hepatic vein, called bridging necrosis. Massive necrosis is not so much a characteristic of specific hepatotoxic chemicals as of their dose. Because of the remarkable ability of the liver to regenerate itself, it is able to withstand moderate zonal or diffuse necrosis. Over a period of several days, necrotic cells are removed and replaced with new cells, restoring normal hepatic architecture and function. If the number of damaged cells is too great, however, the liver’s capacity to restore itself becomes overwhelmed, leading to hepatic failure and death. Another form of cell death is apoptosis, or programmed cell death. Apoptosis is a normal physiological process used by the body to remove cells when they are no longer needed or have become functionally abnormal. In apoptosis, the cell “ commits suicide” through activation of its endonucleases, destroying its DNA. Apoptotic cells are morphologically distinct from cells undergoing necrosis as described above. Unlike cells undergoing necrosis, which swell and release their cellular contents, apoptotic cells generally retain plasma membrane integrity and shrink, resulting in condensed cytoplasm and dense chromatin in the nucleus. There are normally few apoptotic cells in liver, but the number may be increased in response to some hepatotoxic chemicals, notably thioacetamine and ethanol. Also, some chemicals produce hypertrophy, or growth of the liver beyond its normal size.



TABLE 5.1 Drugs and Chemicals that Produce Zonal Hepatic Necrosis Site of Necrosis Chemical Acetaminophen Aflatoxin Allyl alcohol Alloxan α-Amanitin Arsenic, inorganic Beryllium Botulinum toxin Bromobenzene Bromotrichoromethane Carbon tetrachloride Chlorobenzenes Chloroform Chloroprene Cocainea Dichlorpropane Dioxane DDT Dimethylnitrosamine Dinitrobenzene Dinitrotoluene Divinyl ether Ethylene dibromide Ethylene dichloride Ferrous sulfate Fluoroacetate Iodobenzene Iodoform Manganese compounds Methylchloroform Naphthalene Ngaione Paraquat Phalloidin Pyridine Pyrrolidizine alkaloids Rubratoxin Tannic acid Thioacetamide Urethane Xylidine


Zone 2

Zone 3 X X






Source: Adapted from Cullen and Reubner, 1991. a Necrosis is shifted to zone 1 in phenobarbital-pretreated animals.




Examples include lead nitrate and phenobarbital. When exposure or treatment with these agents has ended, the liver will return to its normal size. During this phase, the number of apoptotic cells is increased, reflecting an effort by the liver to reduce its size, in part by eliminating some of its cells. Drugs and chemicals can produce hepatocellular degeneration and death by many possible mechanisms. For some hepatotoxicants, the mechanism of toxicity is reasonably well established. For example, galactosamine is thought to cause cell death by depleting uridine triphosphate, which is essential for synthesis of membrane glycoproteins. For most hepatotoxicants, however, key biochemical effects responsible for hepatocellular necrosis remain uncertain. The search for a broadly applicable mechanism of hepatotoxicity has yielded several candidates: Lipid Peroxidation Many hepatotoxicants generate free radicals in the liver. In some cases, such as carbon tetrachloride, the free radicals are breakdown products of the chemical generated by its cytochrome P450-mediated metabolism in the liver. In other cases, the chemical causes a disruption in oxidative metabolism within the cell, leading to the generation of reactive oxygen species. An important potential consequence of free-radical formation is the occurrence of lipid peroxidation in membranes within the cell. Lipid peroxidation occurs when free radicals attack the unsaturated bonds of fatty acids, particularly those in phospholipids. The free radical reacts with the fatty acid carbon chain, abstracting a hydrogen. This causes a fatty acid carbon to become a radical, with rearrangement of double bonds in the fatty acid carbon chain. This carbon radical in the fatty acid reacts with oxygen in a series of steps to produce a lipid hydroperoxide and a lipid radical that can then react with another fatty acid carbon. The peroxidation of the lipid becomes a chain reaction, resulting in fragmentation and destruction of the lipid. Because of the importance of phospholipids in membrane structure, the principal consequence of lipid peroxidation for the cell is loss of membrane function. The reactive products generated by lipid peroxidation can interact with other components of the cell as well, and this also could contribute to toxicity. The list of chemicals that produce lipid peroxidation as part of their hepatotoxic effects is extensive, and includes halogenated hydrocarbons (e.g., carbon tetrachloride, chloroform, bromobenzene, tetrachloroethene), alcohols (e.g., ethanol, isopropanol), hydroperoxides (e.g., tert-butylhydroperoxide), herbicides (e.g., paraquat), and a variety of other compounds (e.g., acrylonitrile, cadmium, cocaine, iodoacetamide, chloroacetamide, sodium vanadate). Consequently, it is an attractive common mechanism of hepatotoxicity. There is some question, however, as to whether it is the most important mechanism of toxicity for these chemicals. For some of these hepatotoxic compounds, experiments have been conducted in which lipid peroxidation was blocked by concomitant-treatment with an antioxidant. In many cases, hepatotoxicity still occurred. This argues that for at least some agents, lipid peroxidation may contribute to their hepatotoxicity, but is not sufficient to explain all of their toxic effects on the liver. Irreversible Binding to Macromolecules Most of the conventional hepatotoxicants must be metabolized in order to produce liver toxicity, producing one or more chemically reactive metabolites. These reactive metabolites bind irreversibly to cellular macromolecules—primarily proteins, but in some cases also lipids and DNA. This binding precedes most manifestations of toxicity, and the extent of binding often correlates well with toxicity. In fact, histopathology studies with some of these chemicals have found that only cells with detectable reactive metabolite binding undergo necrosis. Examples of hepatotoxic chemicals that produce reactive metabolites include acetaminophen, bromobenzene, carbon tetrachloride, chloroform, cocaine, and trichloroethylene. It is certainly plausible that irreversible binding of a toxicant to a critical protein or other macromolecule in the cell could lead to loss of its function, and the fact that binding precedes most, if not all, toxic responses in the cell make it a logical initiating event. However, demonstrating precisely how irreversible binding causes cell death has been extremely challenging. Several studies have been conducted attempting to identify the macromolecular targets for binding and to determine whether this binding results in an effect that could lead to cell death. Acetaminophen, in particular, has been studied in this regard. While several proteins bound by the acetaminophen reactive metabolite, N-acetyl-p-



benzoquinone imine, have been identified, none as yet has been clearly shown to be instrumental in acetaminophen-induced hepatic necrosis. Without identification of the critical target(s) for irreversible binding for hepatotoxicants, this remains an attractive but unproven mechanism. Loss of Calcium Homeostasis Intracellular calcium is important in regulating a variety of critical intracellular processes, and the concentration of calcium within the cell is normally tightly regulated. The plasma membrane actively extrudes calcium ion from the cell to maintain cytosolic concentrations at a low level compared with the external environment (the ratio of intracellular to extracellular concentration is about 1:10,000). Both the mitochondria and endoplasmic reticulum are capable of sequestering and releasing calcium ion as needed to modulate calcium concentrations for normal cell functioning. Loss of control of intracellular calcium can lead to a sustained rise in intracellular calcium levels, which, in turn, disrupts mitochondrial metabolism and ATP synthesis, damages microfilaments used to support cell structure, and activates degradative enzymes within the cell. These events could easily account for cell death from hepatotoxic chemicals. Early studies of toxic effects of chemicals on liver cells in culture suggested that an influx of calcium from outside the cell (e.g., from plasma membrane failure) was responsible for their toxic effects. Later experiments showed that this was probably not the case, but nonetheless supported disregulation of intracellular calcium as a key event in toxicity. Intracellular calcium levels were observed to rise substantially in response to a number of hepatotoxicants, apparently due to chemical effects on mitochondria and/or the endoplasmic reticulum leading to loss of control of intracellular calcium stores. Impaired extrusion of calcium out of the cell by the plasma membrane might also be important, at least for some chemicals. In general, increases in intracellular calcium preceded losses of viability, suggesting a cause–effect relationship. It is sometimes difficult, however, to discern to what extent elevated calcium levels are the cause of, or merely the result of, cytotoxicity. Immune Reactions This mechanism of hepatotoxicity is not common, but nonetheless important. Characteristically, an initial exposure is required that does not produce significant hepatotoxicity—a sensitizing event. Subsequent exposure to the drug or chemical can lead to profound liver toxicity that may be accompanied by hepatic inflammation. Consistent with a hypersensitivity reaction, there is little evidence of a dose–response relationship, and even small doses can trigger a reaction. This response is usually rare and difficult to predict; hence it is often considered an idiosyncratic reaction. Typically, this kind of hepatotoxicity for a drug or chemical is very difficult to demonstrate in laboratory animals, and unfortunately becomes known only after widespread use or exposure in humans. Perhaps the most familiar example of a drug or chemical producing this type of hepatoxicity is the general anesthetic halothane. Studies suggest that halothane is metabolized to a reactive metabolite that binds with proteins. These proteins become expressed on the cell surface where they are recognized by the immune system as being foreign. The immune system then mounts a cell-mediated response, resulting in destruction of the hepatocytes. This response, called halothane hepatitis, seldom occurs (only about 1 in 10,000 anesthetic administrations in adults) but has a 50 percent mortality rate. A similar phenomenon has been observed with other drugs, including diclofenac. Fatty Liver Many chemicals produce an accumulation of lipids in the liver, called fatty liver or steatosis. Examples of chemicals that produce fatty liver are provided in Table 5.2. Just as hepatocellular necrosis preferentially occurs in specific acinar zones in response to certain chemicals, so does fatty liver. For example, zone 1 is the primary site of lipid accumulation from white phosphorus, while zone 3 is where most of the lipid accumulation is observed with tetracycline and ethanol. The lipid accumulates in vacuoles within the cytoplasm, and these vacuoles are usually present as either one large, clear vacuole (called macrovesicular steatosis) or numerous small vacuoles (microvesicular steatosis). The type of steatosis (macro- or microvesicular) is characteristic of specific hepatotoxicants and, in some cases, of certain diseases or conditions. For example, microvesicular steatosis has been associated with



TABLE 5.2 Drugs and Chemicals that Produce Fatty Liver Antimony Barium salts Borates Carbon disulfide Chromates Dichloroethylene Dimethylhydrazine Ethanol Ethionine Ethyl bromide

Ethyl chloride Hydrazine Methyl bromide Orotic acid Puromycin Safrole Tetracycline Thallium compounds Uranium compounds White phosphorus

tetracycline, valproic acid, salicylates, aflatoxin, dimethylformamide, and some of the antiviral nucleoside analogs used to treat HIV. It is also associated with Reye’s syndrome and fatty liver of pregnancy. Macrovesicular steatosis has been associated with antimony, barium salts, carbon disulfide, dichloroethylene, ethanol, hydrazine, methyl and ethyl bromide, thallium, and uranium compounds. There are several potential chemical effects that can give rise to accumulation of lipids in the cell. These include: 1. Inhibition of Lipoprotein Synthesis. A number of chemicals are capable of inhibiting synthesis of the protein moiety needed for synthesis of lipoproteins in the liver. These include carbon tetrachloride, ethionine, and puromycin. 2. Decreased Conjugation of Triglycerides with Lipoproteins. Another critical step in lipoprotein synthesis is conjugation of the protein moiety with triglyceride. Carbon tetrachloride, for example, can interfere with this step. 3. Interference with Very-Low-Density Lipoprotein (VLDL) Transfer. Inhibition of transfer of VLDL out of the cell results in its accumulation. Tetracycline is an example of an agent that interferes with this transfer. 4. Impaired Oxidation of Lipids by Mitochondria. Oxidation of nonesterified fatty acids is an important aspect of their hepatocellular metabolism, and decreased oxidation can contribute to their accumulation within the cell. Carbon tetrachloride, ethionine, and white phosphorus have been shown to inhibit this oxidation. 5. Increased Synthesis of Fatty Acids. The liver is capable of synthesizing fatty acids from acetyl-CoA (coenzyme A), and increased fatty acid synthesis can increase the lipid burden of the cells. Ethanol is an example of a chemical that produces this effect. Other possible mechanisms might contribute to fatty liver, such as increased uptake of lipids from the blood by the liver, but the role of these processes in drug- or chemical-induced steatosis is less clear. The mechanisms listed above are not mutually exclusive. Indeed, it is likely that many of the chemicals that produce steatosis do so by producing more than one of these effects. Fatty liver may occur by itself, or in conjunction with hepatocellular necrosis. Many chemicals produce a lesion that consists of both effects. Examples include: aflatoxins, amanitin, arsenic compounds, bromobenzene, carbon tetrachloride, chloroform, dimethylnitrosamine, dinitrotoluene, DDT, dichloropropane, naphthalene, pyrrolizidine alkaloids, and tetrachloroethane. Drug- or chemical-induced steatosis is reversible when exposure to the agent is stopped. Phospholipidosis is a special form of steatosis. It results from accumulation of phospholipids in the hepatocyte, and can be caused by some drugs as well as by inborn errors in phospholipid metabolism. Liver sections from patients with phospholipidosis reveal enlarged hepatocytes with



“ foamy” cytoplasm. Often this condition progresses to cirrhosis. Examples of drugs associated with phospholipidosis include amiodarone, chlorphentermine, and 4,4′-diethylaminoethoxyhexoestrol.

Cholestasis The term cholestasis refers to decreased or arrested bile flow. Many drugs and chemicals are able to produce cholestatic injury, and examples are listed in Table 5.3. There are several potential causes of impaired bile flow, many of which can become the basis for drug- or chemical-induced cholestasis. Some of these are related to loss of integrity of the canalicular system that collects bile and carries it to the gall bladder, while others are related to the formation and secretion of bile. For example, α-naphthylisothiocyanate disrupts the tight junctions between hepatocytes that help form the canaliculi, the smallest vessels of the bile collection system. This causes a leakage of bile contents out of the canaliculi into the sinusoids. Other toxicants, such as methylene dianiline and paraquat, impede bile flow by damaging the bile ducts. The primary driving force for bile formation is the secretion of bile acids into the canalicular lumen. This requires uptake of bile acids from the blood into hepatocytes, and then transport into the canaliculus. Anabolic steroids are an example of a class of compounds that produce cholestatic injury by inhibiting these transport processes. Some cholestatic injury can be expected whenever there is severe hepatic injury of any type. This is because normal bile flow requires functioning hepatocytes as well as a reasonably intact cellular architecture in the liver. Whenever this is disrupted, some impairment of bile flow can be expected as a secondary consequence. Many agents produce primarily hepatic necrosis with perhaps limited cholestasis (see Table 5.1), others produce primarily cholestasis with some necrosis (chlorpromazine and erythromycin are examples), and still others are capable of producing cholestasis with little or no damage to the hepatocytes. The contraceptive and anabolic steroids are examples of this last category of agents.

Vascular Injury Cells lining the vasculature within the liver are also potential targets for hepatotoxicants. Injury of vascular cells leads to occlusion (impaired blood flow), which in turn leads to hypoxia. Cells in zone 3 are most vulnerable, since the oxygenation of blood reaching these cells is low even under normal conditions. Typically, hypoxia results in necrosis, and continuing injury over time leads to fibrosis. Severe cases can result in fatal congestive cirrhosis. There are several examples of chemicals known

TABLE 5.3 Drugs and Chemicals that Produce Cholestasis Amitryptyline Ampicillin Arsenicals, organic Barbiturates Carbamazepine Chlorpromazine Cimetidine Cyproheptadine 4,4-Diaminodiphenylmethane 4,4-Diaminodiphenylamine 1,1-Dichloroethylene Dinitrophenol Erythromycin estolate Estrogens

Ethanol Haloperidol Imipramine Methylene dianiline Methyltestosterone α-Naphthylisothiocyanate Norandrostenolone Paraquat Phalloidin Phenytoin Prochlorperazine Tolbutamide Troleandomycin



to produce hepatic venoocclusive disease, including many of natural origin such as pyrrolizidine alkaloids in herbal teas. Oral contraceptives and some anticancer drugs have also been associated with this effect. Peliosis hepatis is another vascular lesion characterized by the presence of large, blood-filled cavities. It is unclear why these cavities form, but there is reason to suspect that it may be due to a weakening of sinusoidal supporting membranes. Use of anabolic steroids has been associated with this effect. Although patients with peliosis hepatis are usually without symptoms, the cavities occasionally rupture causing bleeding into the abdominal cavity. Cirrhosis Chronic liver injury often results in the accumulation of collagen fibers within the liver, leading to fibrosis. Fibrotic tissue accumulates with repeated hepatic insult, making it difficult for the liver to replace damaged cells and still maintain normal hepatic architecture. Fibrous tissue begins to form walls separating cells. Distortions in hepatic microcirculation cause cells to become hypoxic and die, leading to more fibrotic scar tissue. Ultimately, the organization of the liver is reduced to nodules of regenerating hepatocytes surrounded by walls of fibrous tissue. This condition is called cirrhosis. Hepatic cirrhosis is irreversible and carries with it substantial medical risks. Blood flow through the liver becomes obstructed, leading to portal hypertension. To relieve this pressure, blood is diverted past the liver through various shunts not well suited for this purpose. It is common for vessels associated with these shunts to rupture, leading to internal hemorrhage. Even without hemorrhagic episodes, the liver may continue to decline until hepatic failure occurs. The ability of chronic ethanol ingestion to produce cirrhosis is widely appreciated. Occupational exposures to carbon tetrachloride, trinitrotoluene, tetrachloroethane, and dimethylnitrosamine have also been implicated as causing cirrhosis, as well as the medical use of arsenicals and methotrexate. Some drugs (e.g., methyldopa, nitrofurantoin, isoniazid, diclofenac) produce an idiosyncratic reaction resembling viral hepatitis. This condition, termed chronic active hepatitis, can also lead to cirrhosis if the drug is not withdrawn. Tumors Many chemicals are capable of producing tumors in the liver, particularly in laboratory rodents. In fact, in cancer rodent bioassays for carcinogenicity, the liver is the most common site of tumorigenicity. Hepatic tumors may be benign or malignant. Conceptually, the distinction between them is that benign tumors are well circumscribed and do not metastasize (i.e., do not invade other tissues). Malignant tumors, on the other hand, are poorly circumscribed and are highly invasive (see Chapter 13 for additional discussion on benign and malignant tumors). Benign tumors, despite their name, are capable of producing morbidity and mortality. However, they are easier to manage and have a much better prognosis than malignant tumors. Tumors are also classified by the tissue of origin, that is, whether they arise from epithelial or mesenchymal tissue, and by the specific cell type from which they originate. The nomenclature for naming tumors is complex, and the reader is referred elsewhere for a complete discussion of the topic. Basically, malignant tumors arising from epithelial tissue are termed carcinomas, while malignant tumors of mesenchymal origin are sarcomas. Thus, malignant tumors derived from hepatocytes, which are of epithelial origin, are termed hepatocellular carcinomas. Malignant tumors from bile duct cells, also of epithelial origin, are termed cholangiocarcinomas (the prefix cholangio- refers to the bile ducts). Cells of the vascular lining are of mesenchymal origin. Consequently, a malignant tumor in the liver arising from these cells may be called hemangiosarcoma. Benign tumors are also named on the basis of tissue of origin and their appearance. For example, benign tumors of epithelial origin with gland, or glandlike structures are called adenomas, and in the liver these can occur among hepatocytes or bile duct cells. Benign tumors of fibrotic cell origin are termed fibromas, and those in the bile ducts are called cholangiofibromas.



To make things more complicated, cells go through a series of morphological changes as they progress to become a benign or malignant tumor. Thus, groups of cells that represent proliferation of liver tissue, but are not (or not yet) tumors, may be described as nodular hyperplasia, focal hepatocellular hyperplasia, or foci of hepatocellular alteration, depending on their morphological characteristics. The foci of hepatocellular alteration represent the earliest stages that can be detected microscopically. These foci are small groups of cells that are abnormal, but have no distinct boundary separating them from adjacent cells. Their growth rate is such that they are producing little or no compression of surrounding cells. The abnormalities are subtle at this stage, and special stains and markers are sometimes used to help visualize them. Nodular hyperplasia is more readily observed; the group of cells is more circumscribed and compression of adjacent cells is apparent. These cells are thought to represent an intermediate step in tumor development. The significance of these lesions is not that they are associated with any clinical signs or symptoms of disease, but rather that they may represent an area from which a tumor may develop. Consequently, their appearance is important in the assessment of the ability of a drug or chemical to cause cancer. For most chemicals, only a very small percentage—or perhaps none—of the neoplastic areas will go on to produce a malignant tumor. Consequently, the issue of how to use data regarding the appearance of these lesions in the assessment of carcinogencity of a chemical is one of considerable discussion and debate among toxicologists. Liver tumors from chemical exposure can arise through numerous mechanisms. Some hepatocarcinogens form DNA adducts leading to mutations. Nitrosoureas and nitrosamines are examples of hepatocarcinogens thought to produce tumors through this mechanism (see also Chapters 12 and 13 for further discussion of genotoxicity and carcinogenicity). Many chemicals that produce liver tumors are not genotoxic, however, and appear to work through epigenetic mechanisms. Nongenotoxic hepatocarcinogens are many and diverse, and include tetrachlorodibenzo-p-dioxin, sex steroids, synthetic antioxidants, some hepatic enzyme inducing agents (e.g., phenobarbital), and peroxisome proliferators (e.g., clofibrate). A discussion of the mechanisms underlying epigenetic carcinogenesis (e.g., inhibition of cell-to-cell communication, recurrent cellular injury, receptor interactions) is beyond the scope of this chapter, and the reader is referred to Chapter 12 for more information on this subject. Despite the many chemicals found to produce benign and malignant liver tumors in mice and rats, relatively few have been clearly associated with liver tumors in humans. Adenomas have been associated with the use of contraceptive steroids, and clinical and epidemiologic studies implicate anabolic steroids, arsenic, and thorium dioxide as causing hepatocellular carcinoma in humans. Hemangiosarcoma is a rare tumor that has been strongly linked to occupational exposure to vinyl chloride, and has also been associated with arsenic and thorium dioxide exposure.

5.3 EVALUATION OF LIVER INJURY Symptoms of Liver Toxicity As discussed above, liver injury may be either acute or chronic, and may involve liver cell death, hepatic vascular injury, disruption of bile formation and/or flow, or the development of benign or malignant tumors. Obviously, the signs and symptoms that accompany this array of types of liver injury can vary significantly. There are some generalizations that can be made, however. Common symptoms of liver injury include anorexia (loss of appetite), nausea, vomiting, fatigue, and abdominal tenderness. Physical examination may reveal hepatomegaly (swelling of the liver) and ascites (the accumulation of fluid in the abdominal space). Patients whose liver toxicity involves impaired biliary function may develop jaundice, which results from the accumulation of bilirubin in the blood and tissues. Jaundice will appear as a yellowish tint to the skin, mucous membranes, and eyes. Pruritis, or an itching sensation in the skin, will often accompany the jaundice. If the injury is particularly severe, it may lead to fulminant hepatic failure. When the liver fails, death can occur in as little as 10 days. There are several complications associated with fulminant hepatic



failure. Because the liver is no longer able to produce clotting factor proteins, albumin, or glucose, hemorrhage and hypoglycemia are common. Also, failure of the liver leads to renal failure and deterioration of the central nervous system (hepatic encephalopathy). Inability to sustain blood pressure and accumulation of fluid in the lungs may also result. Prognosis is poor for patients with fulminant hepatic failure, with a mortality rate of about 90 percent.

Morphologic Evaluation For laboratory animal studies of hepatotoxicity, histopathologic examination of liver tissue by light or electron microscopy can be extremely valuable. Histopathologic evaluation can provide information on the nature of the lesion and the regions of the liver affected. This, in turn, can provide insight as to the mechanism of toxicity. For example, the presence of fatty liver would suggest that the chemical may interfere with triglyceride metabolism and/or lipoprotein secretion by the liver. Hepatocellular necrosis confined to the centrilobular region might suggest bioactivation of the chemical by cytochrome P450, since most of the activity of this enzyme normally exists in centrilobular cells. Altered morphology of mitochondria as an early event in toxicity might suggest that mitochondrial toxicity is an important initiating event in the sequence of events leading up to cell death. Histopathologic observations alone cannot establish the mechanism of toxicity, and additional experimentation would be required to explore these hypotheses. Nevertheless, morphologic observation provides important clues, and is an integral part of any comprehensive study of potential hepatotoxicity of a chemical. In humans, morphologic evaluation of liver biopsies is sometimes used in the diagnosis and management of chronic liver toxicity, particularly liver cancer. Also, noninvasive techniques such as computerized tomography (CT) or magnetic resonance imaging (MRI) scans are used to detect liver cancer, obstructive biliary injury, cirrhosis, and venoocclusive injury to the liver.

Blood Tests A great deal of insight into the nature and extent of hepatic injury can often be gained through tests on blood samples. There are two fundamental types of blood tests that can be performed. One type is an assessment is based on measuring the functional capabilities of the liver. This can involve an evaluation of the liver’s ability to carry out one or more of its basic physiological functions (e.g., glucose metabolism, synthesis of certain proteins, excretion of bilirubin) or its capacity to extract and metabolize foreign compounds from the blood. The second type of assessment involves a determination of whether there are abnormally high levels in the blood of intracellular hepatic proteins. The presence of elevated levels of these proteins in blood is presumptive evidence of liver cell destruction. Examples of these two types of tests are described below: 1. Serum Albumin. Albumin is synthesized in the liver and secreted into blood. Liver damage can impair the ability of the liver to synthesize albumin, and serum albumin levels may consequently decrease. The turnover time for albumin is slow, and as a result it takes a long time for impaired albumin synthesis to become evident as changes in serum albumin. For this reason, serum albumin measurements are not helpful in assessing acute hepatotoxicity. They may assist in the diagnosis of chronic liver injury, but certain other diseases can alter serum albumin levels, and the test is therefore not very specific. 2. Prothrombin Time. The liver is responsible for synthesis of most of the clotting factors, and a decrease in their synthesis due to liver injury results in prolonged clotting time. In terms of clinical tests, this appears as an increase in prothrombin time. Several drugs and certain diseases also increase prothrombin time. As with serum albumin measurement, this is a relatively insensitive and nonspecific tool for detecting or diagnosing chemical-induced liver injury. 3. Serum Bilirubin. The liver conjugates bilirubin, a normal breakdown product of the heme from red blood cells, and secretes the glucuronide conjugate into the bile. Impairment of normal conjugation



and excretion of bilirubin results in its accumulation in the blood, leading to jaundice. Serum bilirubin concentrations may be elevated from acute hepatocellular injury, cholestatic injury, or biliary obstruction. This test is always included among the battery of tests to assess liver function clinically, although it is not a particularly sensitive test for acute injury. 4. Dye Clearance Tests. These tests involve administration of a dye that is cleared by the liver and measurement of its rate of disappearance from the blood. Delayed clearance is interpreted as evidence of liver injury. One such dye is sulfobromophthalein (Bromsulphalein; or BSP). Clearance of BSP from the blood is dependent on its active transport into liver cells, conjugation with glutathione, and then active transport into the bile. Conceivably, disruption of any of these processes could result in delayed clearance, although the biliary excretion step is regarded as most critical. The test consists of administering a dose of the dye intravenously and measuring its concentration in blood spectrophotometrically over time. Another dye used for this purpose is indocyanine green (ICG). Unlike BSP, ICG is excreted into the bile without conjugation. Following an intravenous dose, the disappearance of ICG from blood can be measured with repeated blood samples or noninvasively by ear densitometry. The dye tests, although well established, are seldom used clinically. 5. Drug Clearance Tests. This test relies on the principle that liver injury will result in impaired biotransformation. The biotransformation capacity of the liver is assessed by following the rate of elimination of a test drug whose clearance from blood is dependent on hepatic metabolism (i.e., a drug for which other elimination processes, such as renal excretion, are insignificant). A test drug such as antipyrine, aminopyrine, or caffeine is administered, and its rate of disappearance from blood is followed over time through serial blood sampling. This rate is compared with a value considered “ normal” to determine whether impaired biotransformation exists. This can also be used to test for hepatic enzyme induction, in which the rate of elimination from blood would be increased, rather than decreased as in liver injury. This test is primarily used for research purposes. 6. Measurement of Hepatic Enzymes in Serum. Cells undergoing acute degeneration and injury will often release intracellular proteins and other macromolecules into blood. The detection of these substances in blood above normal, baseline levels signals cytotoxicity. This is true for any cell type, and in order for the presence of intracellular proteins in blood to be diagnostic for any particular type of cell injury (e.g., liver toxicity versus renal toxicity versus cardiotoxicity), the proteins must be associated rather specifically with a target organ or tissue. Fortunately, several proteins are found primarily in hepatocytes, and their presence in blood in elevated levels is the basis for some of the most commonly used tests for hepatotoxicity. Table 5.4 shows many of the most common proteins measured in these tests. The reader will note that all of these proteins are enzymes. This is not a coincidence. While any intracellular protein specific to the liver would be useful theoretically, enzymes are proteins that can be measured specifically (by measuring the rate of their particular enzyme activity) using

TABLE 5.4 Serum Enzyme Indicators of Hepatotoxicity Enzyme


Alanine aminotransferase


Aspartate aminotransferase


Alkaline phosphatase γ-Glutamyl transferase; γ-glutamyltranspeptidase 5′-Nucleotidase Sorbitol dehydrogenase


Ornithine carbamoyltransferase


Comments Found mainly in the liver; increase reflects primarily hepatocellular damage Less specific to the liver than ALT; increase reflects primarily hepatocellular damage Increases reflect primarily cholestatic injury Increases reflect primarily cholestatic injury, although elevated in hepatocellular damage as well Increases reflect primarily cholestatic injury High specificity for liver; increase reflects primarily hepatocellular damage High specificity for liver; increase reflects primarily hepatocellular damage



assays that are rapid and inexpensive. In fact, the concentrations of each of these proteins are typically measured as an enzyme activity rate, rather than a true concentration per se. Aminotransferase activities [alanine aminotransferase (ALT) and aspartate aminotransferase (AST)], alkaline phosphatase activity, and gamma glutamyltransferase transpeptidase (GGTP) are included in nearly all standard clinical test suites to assess potential hepatotoxicity. The value of performing a battery of these tests is that each test responds slightly differently in the various forms of liver injury, and evaluating the pattern of responses can offer insight into the type of injury that has occurred. For example, severe hepatic injury from acetaminophen can result in dramatic increases in serum ALT and ALT activities (up to 500 times normal values), but only modest increases in alkaline phosphatase activity. Pronounced increases in alkaline phosphatase is characteristic of cholestatic injury, where increases in ALT and AST may be limited or nonexistent. In alcoholic liver disease, AST activity is usually greater than ALT activity, but for most other forms of hepatocellular injury ALT activities are higher. Serum GGTP is an extremely sensitive indicator of hepatobiliary effects, and may be elevated simply by drinking alcoholic beverages. It is not a particularly specific indicator (it is increased by both hepatocellular and cholestatic injury) and is best utilized in combination with other tests. Serum levels of enzymes such as lactate dehydrogenase have been used to evaluate liver toxicity, but this enzyme has such low specificity for the liver that interpretation of these results is impossible without other confirming tests. Other enzymes such as sorbitol dehydrogenase (SDH) and ornithine carbamoyltransferase (OCT) are quite specific to the liver.

5.4 SUMMARY Both the anatomic location and its role as a primary site for biotransformation make the liver uniquely susceptible to drug- and chemical-induced injury. Many chemicals encountered in the workplace and environment are capable of producing toxic effects in the liver:

• There are many types of liver injury, including hepatocellular degeneration and death (necrosis), fatty liver, cholestasis (decreased or arrested bile flow), vascular injury, cirrhosis, and tumor development.

• Hepatic injury from drugs and chemicals can arise from a variety of mechanisms. While the mechanism of toxicity for some chemicals is reasonably well established, many aspects of toxic mechanisms for most chemicals remain unclear.

• Hepatotoxic chemicals can attack a variety of subcellular targets. Principal organelles and structures affected include the plasma membrane, mitochondria, the endoplasmic reticulum, the nucleus, and lysosomes.

• Liver injury can be evaluated morphologically (microscopic examination of liver tissue) or through blood tests. Blood tests are designed to either measure the functional capacity of the liver or the appearance of intracellular hepatic contents in the blood.

REFERENCES AND SUGGESTED READING Cullen, J. M., and B. H. Ruebner, “ A histopathologic classification of chemical-induced injury of the liver,” in Hepatotoxicity, R. G. Meeks, S. D. Harrison, and R. J. Bull, eds., CRC Press, Boca Raton, FL, 1991, pp. 67–92. Delaney, K., “ Hepatic principles,” in Goldfrank’s Toxicologic Emergencies, L. R. Goldfrank, N. E. Flomenbaum, N. A. Lewin, R. S. Weisman, M. A. Howland, and R. S. Hoffman, eds., Appleton & Lange, Stamford, CT, 1998, pp. 213–228. Kedderis, G. L. “ Biochemical Basis of Hepatocellular Injury.” Toxicologic Pathology, 24 (1): 77–83 (1996).



Lamers, W. H., A. Hilberts, E. Furt, J. Smith, G. N. Jonges, C. J. F. von Noorden, J. W. G. Janzen, R. Charles, and A. F. M. Moorman, “ Hepatic enzymic zonation: A reevaluation of the concept of the liver acinus,” Hepatology 10: 72–76 (1989). Marzella, L., and B. F. Trump, “ Pathology of the liver: Functional and structural alterations of hepatocyte organelles induced by cell injury” in Hepatotoxicity, R. G. Meeks, S. D. Harrison, and R. J. Bull, eds., CRC Press, Boca Raton, FL, 1991, pp. 93–138. MacSween, R. N. M., and R. J. Scothorne, “ Developmental anatomy and normal structure,” in Pathology of the Liver, R. N. M. MacSween, P. P. Anthony, P. J. Scheuer, A. D. Burt, and B. C. Portmann, eds., Churchill Livingstone, Edinburgh, 1994, pp. 1–49. Miyai, K., “ Structural organization of the liver,” in Hepatotoxicity, R. G. Meeks, S. D. Harrison, and R. J. Bull, eds., CRC Press, Boca Raton, FL, 1991, pp. 1–65. Moslen, M. T., “ Toxic responses of the liver,” Casarett and Doull’s Toxicology. The Basic Science of Poisons, 5th ed., C. D. Klaasen, M. O. Amdur, and J. Doull, eds., McGraw-Hill, New York, 1996, pp. 403–416. Popper, H., “ Hepatocellular degeneration and death,” in The Liver: Biology and Pathobiology, I. M. Arias, W. B. Jakoby, H. Popper, D. Schachter, and D. A. Shafritz, eds., Raven Press, New York, 1988, pp. 1087–1103. Rappaport, A. M., “ Physioanatomical basis of toxic liver injury,” in Toxic Injury of the Liver, Part A, E. Farber and M. M. Fisher, eds., Marcel Dekker, New York, 1979, pp. 1–57. Zimmerman, H. J., and K. G. Ishak, “ Hepatic injury due to drugs and toxins,” in Pathology of the Liver, R. N. M. MacSween, P. P. Anthony, P. J. Scheuer, A. D. Burt, and B. C. Portmann, eds., Churchill Livingstone, Edinburgh, 1994, pp. 563–633.

6 Nephrotoxicity: Toxic Responses of the Kidney NEPHROTOXICITY: TOXIC RESPONSES OF THE KIDNEY


This chapter will give the environmental and occupational health professional information about

• • • •

The importance of kidney functions How toxic agents disrupt kidney functions Measurements performed to determine kidney dysfunctions Occupational and environmental agents that cause kidney toxicity

6.1 BASIC KIDNEY STRUCTURES AND FUNCTIONS The principal excretory organs in all vertebrates are the two kidneys. The primary function of the kidney in humans is removing wastes from the blood and excreting the wastes in the form of urine. However, the kidney plays a key role in regulating total body homeostasis. These homeostatic functions include the regulation of extracellular volume, the regulation of calcium metabolism, the control of electrolyte balance, and the control of acid–base balance. The adult kidneys of reptiles, birds, and mammals (including humans) are nonsegmental and drain wastes only from the blood (principally breakdown products of protein metabolism). The kidneys are paired organs that lie behind the peritoneum on each side of the spinal column in the posterior aspect of the abdomen. The adult human kidney is approximately 11 cm long, 6 cm broad, and 2.5 cm thick. In human adults individual kidneys weigh 125–170 g for males and 115–155 g for females. The renal artery and vein pass through the hilus, which is a slit in the medial or concave surface of each kidney (Figure 6.1b). From each kidney a common collecting duct, the ureter, carries the urine posteriorly to the bladder where it can be voided from the body. Each human kidney consists of an outer cortex and an inner medulla (see Figures 6.1b and 6.2). The cortex constitutes the major portion of the kidney and receives about 85 percent of the total renal blood flow. Consequently, if a toxicant is delivered to the kidney in the blood, the cortex will be exposed to a very high proportion.

Blood Flow to the Kidneys The kidneys represent approximately 0.5 percent of the total body weight, or approximately 300 g in a 70-kg human. Yet the kidneys receive just under 25 percent of the total cardiac output, which is about 1.2–1.3 L blood/min, or 400 mL/100 g tissue/min. The rate of blood flow through the kidneys is much greater than through other very well perfused tissues, including brain, heart, and liver. If the normal blood hematocrit (i.e., that proportion of blood that is red blood cells) is 0.45, then the normal renal plasma flow is approximately 660 to 715 mL/min. Yet only 125 mL/min of the total plasma flow is Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




Figure 6.1 The human renal excretory system: (a) the complete excretory system; (b) cross section of kidney; (c) representative section for the enlargement in Figure 6.2.

actually filtered by the kidney. Of this, the kidney reabsorbs approximately 99 percent, resulting in a urine formation rate of only about 1.2 mL/min. Thus, the kidneys, which are perfused at approximately 1 L/min, form urine at approximately 1 mL/min or 0.1 percent of the perfusion. Because of the high volume of blood flow to the kidneys, a chemical in the blood is delivered to this organ in relatively large quantities. The kidney requires large amounts of metabolic energy to remove wastes from the blood by tubular secretion and to return filtered nutrients back to the blood. Roughly 10 percent of the normal resting oxygen consumption is needed for the maintenance of proper kidney function. Therefore, the kidney is sensitive to agents, such as barbiturates, that induce ischemia, a lack of oxygen caused by a decrease in blood flow. Acute intoxication by barbiturates induces severe hypotension (i.e., low blood pressure) and shock. The severe decrease in blood pressure results in a decrease in filtration of the plasma, resulting in a decrease (oliguria) or cessation (anuria) of urine formation. At an early stage this is called pre–renal failure, and a reversal in the blood deficit to the kidney will restore normal renal function. However, a critical point is reached when renal sufficiency cannot be restored because of the cell death caused by ischemic anoxia, and the resultant renal failure is irreversible. In this situation, the accumulation in the blood of wastes normally excreted (uremia) results in death. It should be remembered, then, that any agent or physical trauma that causes severe hypotension and shock may produce acute renal failure and eventually death by a similar mechanism. Nephrons: The Functional Units of the Kidney The cortex of each kidney in humans contains approximately one million excretory units called nephrons. Agents toxic to the kidney generally injure these nephrons, and such agents are therefore referred to as nephrotoxicants. Degeneration, necrosis, or injury to the nephron elements is referred to as a nephrosis or nephropathy. An individual nephron may be divided into three anatomic portions: (1) the vascular or bloodcirculating portion, (2) the glomerulus, and (3) the tubular element (Figures 6.2 and 6.3). The glomerulus, which is about 200 µm in diameter, is formed by the invagination of a tuft of capillaries



Figure 6.2 Cortical and juxtamedullary nephrons. Enlargement of representative kidney section in Figure 6.1c. (Based on B. Brenner and F. Rector, The Kidney, Saunders, Philadelphia, 1976.)

into the dilated, blind end of the nephron (Bowman’s capsule). The capillaries are supplied by an afferent arteriole and drained by an efferent arteriole. These vascular elements deliver waste and other materials to the tubular element for excretion, return reabsorbed and synthesized materials from the tubular element to the blood circulation, and deliver oxygen and nourishment to the nephron. The Glomerulus and Glomerular Filtration The glomerulus behaves as if it were a filter with pores 100 Å in diameter, or about 100 times more permeable than the capillaries in skeletal muscle. Substances as great as 70,000 daltons can appear in the glomerular filtrate, but most proteins in the plasma are still too large to pass through the glomerulus. Therefore, a substance that is, for example, 75 percent bound to plasma proteins has an effective filterable concentration of 25 percent its total plasma concentration. Small amounts of protein, principally the albumins, which are important chemical-binding proteins, may appear in the glomerular filtrate, but these are then normally reabsorbed. The glomerular filter can be made more permeable in certain disease states and by actions of certain nephrotoxicants. Both circumstances may result in the appearance of protein in the urine (proteinuria). If damage to the glomerular element is severe, the result is a loss of a large amount of the plasma proteins. If this occurs at a rate greater than the rate at which the liver can synthesize the plasma proteins, the result will be hypoproteinemia (lower than normal levels of proteins in the blood) and a concomitant edema due to the reduction in osmotic pressure. This clinical picture is sometimes referred to as the nephrotic syndrome. However, transient but significant proteinuria occurs normally after prolonged standing or strenuous exercise, so a single measurement of high protein levels in the urine may not indicate kidney damage. Nephron Tubules and Tubular Reabsorption The tubular element of the nephron selectively reabsorbs 98–99 percent of the salts and water of the initial glomerular filtrate. The tubular element of the



Figure 6.3 Juxtamedullary nephron: (1) afferent arteriole; (2) efferent arteriole; (3) glomerulus; (4) proximal convoluted tubule; (5) proximal straight tubule (pars recta); (6) descending limb of the loop of Henle; (7) thin ascending limb of the loop of Henle; (8) thick ascending limb of the loop of Henle; (9) distal convoluted tubule; (10) collecting duct. (Based on J. Doull, et al., eds., Casarett and Doull’s Toxicology: The Basic Science of Poisons, 2nd ed., Macmillan, New York, 1980.)

nephron consists of the proximal tubule, the loop of Henle, the distal tubule, and the collecting duct (see Figure 6.3). The proximal tubule consists of a proximal convoluted section (pars convoluta) and a distal straight section (pars recta). Substances that are actively reabsorbed in the proximal tubule include glucose, sodium, potassium, phosphate, amino acids, sulfate, and uric acid. Essentially all amino acids and glucose are reabsorbed in the proximal tubule, and virtually none normally appear in the urine. Agents toxic to the proximal tubule cause amino acids and glucose to appear in the urine (aminoaciduria and glycosuria). Even though 250 g of glucose normally passes through the kidney daily, no more than 100 mg is usually excreted in 24 h. However, glucose does appear in excess quantities in the urine if high blood glucose levels produce a glucose load in the filtrate and this exceeds the resorptive capacity of the proximal tubule of the nephrons. This occurs in diabetes mellitus, in which excess glucose appears in urine because excessive amounts of glucose in the blood plasma filtrate have overwhelmed the glucose transport system in the nephron. Water is also reabsorbed in the proximal tubule because of an osmotic gradient between the filtrate in the tubule and the blood plasma. Thus, isotonicity is maintained in the proximal tubule even though there is a selective reabsorption of solutes. Approximately 75 percent of the glomerular filtrate fluid is reabsorbed in the proximal tubule.



If tubular reabsorption of substances is compromised, then less water is reabsorbed. The result is diuresis (increased urine flow) and polyuria (excess urine production). Toxic agents can cause polyuria by affecting active solute reabsorption. Tubular Secretion Active transport of certain organic compounds into the tubular fluid also occurs in the proximal tubule. There are two separate active secretory systems in the proximal tubule: one for anionic (negatively charged) organic chemical species, and a similar but separate system for cationic (positively charged) organic chemical species. The organic anion secretory system is the better studied. Organic cations such as tetramethyl ammonium are actively secreted, but this system is not as well studied as the organic anion secretory system. The two secretory systems also have unique competitors and inhibitors. Penicillin and probenecid are actively secreted by the organic anion secretory system. As a consequence, they inhibit the excretion of PAH (p-amminohippuric acid) and each other. In fact, probenicid has been used to prolong the half-life of penicillin in the blood since probenicid inhibits secretion of penicillin into the proximal tubules and its subsequent excretion in the urine. These organic anions do not inhibit secretion of organic cations or compete with them for secretion. The reverse is also true. The result is that substances reabsorbed from the tubule will have a clearance significantly less than the glomerular filtration rate (approximately 125 mL/min), while those secreted into the tubules will have a clearance greater than the glomerular filtration rate in the adult human. The Loop of Henle After the glomerular filtrate has passed the proximal tubule in the nephron, it moves into the loop of Henle. A nephron with a glomerulus in the outer portion of the renal cortex has a short loop of Henle, whereas a nephron with a glomerulus close to the border between the cortex and medulla (juxtamedullary nephrons) has a long loop of Henle extending into the medulla and papilla (Figures 6.2 and 6.3). Approximately 15 percent of the nephrons in humans are juxtamedullary. As the tubule descends into the medulla there is an increase in osmolality of the interstitial fluid. In the descending limb the tubular fluid becomes hypertonic (high in salt) as water leaves the tubule to maintain isoosmolality with the hypertonic interstitial fluid. However, in the thick segment of the ascending portion of the loop of Henle the tubule becomes impermeable to water, and sodium is actively transported out of the tubule with a decrease in the osmolality of the filtrate and an increase in the osmolality of the interstitial fluid. The sodium transport in the ascending limb is necessary for maintenance of the interstitial fluid concentration gradient. An additional 5 percent of the glomerular filtrate fluid is reabsorbed in the loop of Henle, making a total of 80 percent of the total water reabsorbed at this point. Urine Formation Once the tubular fluid enters the distal convoluted tubule and collecting duct, it is hypotonic (low salt concentration) in comparison to blood plasma because of the active transport of sodium out of the tubule at the loop of Henle. In the presence of vasopressin, the antidiuretic hormone, the collecting duct becomes permeable to water, and the water moves from the tubular fluid in order to maintain isoosmolality. However, in the absence of vasopressin, the collecting duct is impermeable to water, which results in excretion of a large volume of hypotonic urine. Normally, another 19 percent of the original glomerular filtrate fluid is reabsorbed in the last portion of the nephron, so that a total of 99 percent of the fluid filtered at the glomerulus is reabsorbed—only 1 percent of the fluid entering the nephron is excreted in the urine. Thus, the normal flow of urine is only about 1 mL/min, while in the absence of vasopressin it can be increased to 16 mL/min. The kidney’s ability to concentrate urine is determined by the measurement of urine osmolality. Urine osmolality can vary between 50 and 1400 mOsm/L. Certain nephrotoxicants compromise the kidney’s ability to concentrate the urine. These changes occur early after the exposure to the nephrotoxicant and frequently foreshadow graver consequences. The excretion of urea, a metabolic breakdown product of protein, is a special case. Urea passively diffuses out of the glomerular filtrate of the tubules as fluid volume decreases. At low urine flow, more urea has the opportunity to leave the tubule. Under these conditions only 10–20 percent of the urea is excreted. At conditions where the urine flow is high, the urea has less time to diffuse through



membranes with the water; this results in a 50–70 percent excretion of urea. A second factor in urea excretion is that it accumulates in the medullary interstitial fluid along a concentration gradient. Since the walls of the collecting ducts are permeable to urea fluid where they pass through the medulla, the urea content of the urine is higher than it would be if they passed only through regions with low urea concentration. Passive reabsorption occurs for all nonionic compounds, while ionic chemicals are not passively reabsorbed. For organic acids, a basic urine is desirable to maximize excretion since more of the acid will be ionized at higher pH (Haldane equation, Chapter 4). For organic bases, an acidic urine is desirable for maximal excretion, because more of the basic compound will be ionized.

Bladder The urine that flows from the collecting ducts is deposited in the bladder. Little of the literature is devoted to the bladder and its functioning. However, some compounds are toxic to the bladder. Bladder cancer is thought to be caused by occupational exposure to bicyclic aromatic amines. The bladder epithelium contains high levels of an enzyme, prostaglandin H synthase (PHS), which can activate certain aromatic amines, such as benidine, 4-aminobiphenyl, and 2-aminonaphthalene, to compounds that can react with DNA. The normal metabolism of these compounds involves acetylation, and there are several genetic polymorphisms of the enzymes (N-acetyltransferases) responsible for acetylating them. Individuals with slow acetylating enzymes are more likely to develop bladder cancer after exposure.

Important Kidney Functions Seldom Considered as Toxic Endpoints Renal Erythropoietic Factor The kidney synthesizes hormones essential for certain metabolic functions. For example, hypoxia stimulates the kidneys to secrete renal erythropoietic factor, which acts on a blood globulin (proerythropoietin) released from the liver to form erythropoietin, a circulating glycoprotein with a molecular weight of 60,000 daltons. The erythropoietin acts on erythropoietinsensitive stem cells in the bone marrow, stimulating them to increase hemoglobin synthesis, produce more red blood cells, and release them into the circulating blood. The increased oxygen-carrying capacity of the blood reduces the effects of hypoxia. Thus, in chronic renal failure, anemia usually develops, in large part caused by decreased synthesis of erythropoietic factor because of damage to the kidney tissues responsible for its synthesis. In addition to hypoxia, androgens and cobalt salts also increase production of renal erythropoietic factor by the kidneys. In fact, administration of cobalt salts produces an overabundance of red cells in the blood (i.e., polycythemia) by this mechanism. Polycythemia has been observed in heavy drinkers of cobalt-contaminated beer. Regulation of Blood Pressure The kidney is involved in regulating blood pressure in several ways. The kidney produces renin, a proteolytic enzyme, which cleaves a plasma protein globulin to form angiotensin I. Angiotensin I is converted to angiotensin II, a potent vasoconstrictor. The angiotensin II stimulates release of aldosterone from the adrenal cortex, and aldosterone increases reabsorption of sodium in the kidney, leading to an increase in blood plasma osmolality and an increase in extracellular volume. A decrease in the mean renal arterial pressure is the stimulus controlling kidney renin production and the compensatory increase in arterial pressure by the abovementioned mechanisms. In addition, renal disease and narrowing of the renal arteries are known to cause sustained hypertension in humans. It appears that the kidney produces vasodepressor substances that are thought to be important in the regulation of blood pressure. Thus, changes in the kidney that disturb the renin– angiotensin–aldosterone system and/or secretion of the vasodepressor substances are suspected of playing a key role in the etiology of certain forms of hypertension.



Metabolism of Vitamin D The kidney also plays a key role in the metabolism of vitamin D, thus performing a vital function in the hormonal regulation of calcium in the body. Vitamin D3 (cholecalciferol) is relatively inactive. The liver hydroxylates vitamin D3 to 25-hydroxycalciferol, and then, the kidney hydroxylates the 25-hydroxycalciferol to 1,25-dihydroxycalciferol, the most potent active form of vitamin D. The kidney is also the key to the metabolism of parathyroid hormone, another hormone important to calcium regulation. If the kidney is damaged, thereby disrupting its role in vitamin D and parathyroid hormone metabolism, the development of a renal osteodystrophy can occur, which is characterized by skeletal disease and hyperplasia of the parathyroid gland.

6.2 FUNCTIONAL MEASUREMENTS TO EVALUATE KIDNEY INJURY From the preceding paragraphs it should be clear that the kidney plays an essential role in maintaining a number of vital body functions. Therefore, if a disruption of normal kidney function is caused by the action of a toxic agent, a number of serious sequelae can occur besides a disruption in blood waste elimination. However, for clinical purposes, alterations in the excretion of wastes are the principal endpoints for determining the action of nephrotoxicants. Nevertheless, it must be remembered that changes in the other functions may also be present, even if they are not conveniently or routinely measured as toxic endpoints. Determining the excretion rate of certain drugs from the kidney is a useful clinical procedure for diagnosing the functional status of the kidney. This rate of elimination in the urine is the net result of three renal processes:

• Glomerular filtration • Tubular reabsorption • Tubular secretion The rates of glomerular filtration and tubular secretion are dependent on the concentration of the drug in the plasma, and the rate of reabsorption by the tubules is dependent on the concentration of drug in the urine. The Glomerular Filtration Rate The glomerular filtration rate (GFR) can be measured in intact animals and humans by measuring both the excretion and plasma levels of those chemicals that are freely filtered through the glomeruli and neither secreted nor reabsorbed by the kidney tubules. The substance used should ideally be one that is freely filtered, not metabolized, not stored in the kidney, and not protein bound. Inulin, a polymer of fructose with a molecular weight of 5200 daltons, meets these criteria. For measuring the glomerular filtration rate the inulin is allowed to equilibrate within the body, and then accurately timed urine specimens and plasma samples are collected. The following general formula is used to determine the clearance in this procedure: Ua × V = Cl Pa where

Ua V Pa Cl

= = = =

concentration of substance a per milliliter urine urine volume excreted per unit time concentration of substance a per milliliter of plasma clearance of substance per unit of time

For clearance of inulin (in), the following values can be used to demonstrate a sample calculation: Uin = 31 mg/mL



V = 1.2 mL/min Pin = 0.30 mg/mL Thus, (31 mg / ml) × (1.2 ml / min) = 124 ml / min. 0.30 mg / ml The normal human glomerular filtration rate in adult humans is about 125 mL/min and inulin clearance is routinely used as a measure of glomerular function. The GFR is not only a measure of the functional capacity of the glomeruli but also indicates the kidney’s ability to concentrate urine by removal of water. By comparing the amount (milliliters) of urine voided in one minute to the amount (milliliters) of plasma cleared, information can be gained about the amount of water reabsorbed during passage through the tubules. Diseases or nephrotoxicants that affect the glomerulus or those that produce renal vascular disease have a profound effect on the glomerular filtration rate. Indeed, any significant renal disease or nephrotoxic compromise can reduce the glomerular filtration rate. It should also be realized that any agent inducing severe hypotension or shock will likewise reduce the glomerular filtration rate. Measurement of certain natural endogenous substances in the blood can be used to assess glomerular function as well. The measurement of blood urea nitrogen (BUN) and plasma creatinine are two endogenous compounds routinely measured for the clinical assessment of glomerular function. As glomerular filtration decreases, BUN and plasma creatinine become more elevated. Normal BUN ranges from 5 to 25 mg/100 mL, while serum creatinine ranges from 0.5 to 0.95 mg/mL of serum. Nephrotoxicants may also disrupt the selective permeability of the glomerular apparatus. Normally, the result is an increase in porosity in the glomerulus; protein enters the glomerular filtrate and subsequently the urine. Therefore, if a compound causes excretion of large amounts of protein into the urine it must be suspected as a nephrotoxicant, and measurement of protein in urine, particularly those of high molecular weight, is used to determine which chemicals produce toxic changes to the glomerulus. The normal excretion of protein in humans is no more than 150 mg in 24 h. Renal Plasma Flow Some organic acids, such as p-aminohippuric acid (PAH), can be used in clearance studies to obtain information about the total amount of plasma flowing through the kidneys. PAH is transported so effectively that it is almost completely removed from the plasma in a single passage through the kidney (i.e., 80–90 percent). Any chemically induced reduction in the PAH clearance may be caused by either a disruption of the active secretory process or by an alteration of the renal blood flow. In a clinical setting, measurements can be made of the concentration of PAH per milliliter of plasma (PPAH), of the concentration of PAH per milliliter of urine (UPAH), and of the volume of urine excreted per minute (V). Using the formula that was previously discussed, the clearance of PAH in mL/min can be calculated. This calculation represents the rate of plasma flow through the kidneys (average renal plasma flow in the normal, healthy adult male is about 650 mL/min). Excretion Ratio Another useful calculation for evaluating kidney injury is the excretion ratio: Excretion ratio =

Renal plasma clearance of drugs (ml / min) Normal GFR (ml / min)

If the ratio is less than 1.0, it indicates that a drug has been partially filtered, perhaps also secreted, and then partially reabsorbed. A value greater than 1.0 indicates that secretion, in addition to filtration, is involved in the excretion. A substance that is completely reabsorbed, such as glucose, would have an excretion ratio of 0, and a substance such as PAH that is completely cleared can have a ratio of about 5.



Additional Clinical Test Alterations in renal function can be determined by a variety of other tests. A battery of such tests includes urinary pH, measurement of urine volume, and a determination of the excretion of sodium and potassium. An excess of protein or the appearance of sugar in the urine indicates abnormalities in renal function as would changes in urine sediments. These are all general tests, but they can provide information about the changes in total kidney function.

6.3 ADVERSE EFFECTS OF CHEMICALS ON THE KIDNEY Frequently, exposure to large amounts of a chemical can cause kidney effects that are not observed at lesser exposures. Effects of kidney damage are frequently assessed in nonspecific terms such as changes in kidney weight (both increases and decreases) or increases in protein content of the urine (proteinuria) or changes in volume of urine (polyuria, oliguria, or anuria). Acute renal failure (ARF) is one of the more common responses of the kidney to toxicants. ARF is characterized by a rapid decline in glomerular filtration rate and an increase in the concentration of nitrogenous compounds in the blood. Numerous mechanisms have been identified that lead to ARF. Compounds that cause renal vasoconstriction reduce the amount of blood that reaches the glomerulus and cause hypoperfusion, a reduction in the amount of blood filtered. When toxicants cause glomerular injury, they can reduce the amount of filtrate that enters the tubules, called hypofiltration. When the tubular cells are injured by toxicants, the permeability of the tubule is increased and the filtrate is allowed to backleak into the interstitium and into the circulation, producing an apparent reduction of the GFR. Some toxicants may reduce the adhesion of tubular cells to each other, causing them to obstruct the pathway for filtrate to be reabsorbed and thus increasing the pressure within the tubule leading to a resistance of movement of filtrate into the tubule. The kidney is capable of overcoming substantial loss of function. If a single kidney is lost, the remaining kidney can increase its GFR by 40–60 percent. Individual nephrons can increase the reabsorption of water and solutes so that the osmotic balance is maintained and there is no apparent difference in tests of kidney function. Although the compensatory mechanisms protect the whole organism in the short term, the compensatory responses may lead to chronic renal failure in the long term. The increase in glomerular pressure leads to sclerosis of the glomerulus and the degeneration of the capillary loops, among other changes in the nephron whose roles in compensatory nephron damage are not as well documented. The loss of additional nephrons and the capacity to remove wastes by this mechanism leads to additional compensation by other nephrons, which are subsequently damaged by similar mechanisms, eventually leading to chronic renal failure. Other means of protecting the kidney from damage include the induction of metallothionein and heat-shock proteins. Heat-shock proteins play a housekeeping role to maintain normal protein structure and/or degrade damaged proteins. Metallothionein is a low-molecular-weight protein that binds heavy metals and prevents them from inducing toxic responses. The production of metallothionein is induced by the presence of heavy metals, and, when low doses of the heavy metal are given, the metallothionein is produced and can provide protection against larger doses given at a later time. If no exposure has occurred previously, no protection is provided because metallothionein is not present to bind the heavy metal. In addition to the organ-level response of the kidney, many toxicants affect specific regions of the nephron. They may damage the glomerulus, the proximal tubule, or the further tubule elements such as the loop of Henle, distal tubule, or collecting duct. The most common site of injury for toxicants is the proximal tubule. Nephrotoxic Agents Many compounds are known to adversely affect kidney tissues at some exposure level, but the kidney is the tissue affected at the least lowest observed adverse effect levels for only a few compounds. The chemicals for which the American Conference of Governmental Industrial Hygienists (ACGIH) has



established Threshold Limit Values (trademark) (TLVs) that are intended to protect against affects on the kidney are given in Table 6.1. For these compounds, however, the renal system may not be the only system the TLV is intended to protect. Two classes of environmentally or occupationally relevant chemicals that damage the kidney are the heavy metals and halogenated hydrocarbons. The adverse effects of representative chemicals from each group are discussed below. Some occupations that have exposure to nephrotoxicants are given in Table 6.2. Cadmium The kidney is the organ most sensitive to the toxic effects of cadmium. Numerous factors have been used as indicators of kidney damage by cadmium. One of the early indicators is the presence of 2-microglobulin, a low-molecular-weight protein that is usually reabsorbed by the proximal tubules. Proximal tubule damage of the nephrons caused by cadmium is also evidenced by glycosuria, aminoaciduria, and the diminished ability of the kidney to secrete PAH. As damage increases, there is an increase in urinary excretion of low- and high-molecular-weight proteins, which predicts an acceleration of the decline in glomerular filtration rate. Workers in factories where nickel/cadmium batteries are manufactured and who are exposed to excessive amounts of cadmium oxide exhibit

TABLE 6-1. Chemicals with ACGIH TLVs Specifically Set to Prevent Renal Effects Arsine Cadmium Chloroform 1-chloro-1-nitropropane o-Chlorostyrene Hexavalent Chromium Compounds (water soluble) Chromyl chloride †Cyclohexane p-dichlorobenzene 1,1-dichloroethane Diethanolamine 1,4-dioxane Diphenylamine Dipropylketone Epichlorohydrin Ethyl bromide Ethylene chlorohydrin Ethylene dibromide Ethylene oxide Ethyl silicate Hexachlorobutadiene Hexachloroethane Hexafluoroacetone Indene Iodoform Lead Lead Arsenate Mercury, aryl, inorganic, elemental Mesityl oxide

Methyl tert-butyl ether Methyl Chloride Methyl Chloroform Methylcyclohexanol 2-methyl cyclopentadienyl manganese tricarbanol 4,4′-methylene bis(2-chloroaniline) Methyl ethyl ketone peroxide Methyl isoamyl ketone Methyl isobutyl ketone Nickel, elemental Nitrogen trifluoride Oxygen difluoride Paraquat Phenothiazine Phosphorous (yellow) Picloram Pindone Propargyl alcohol Propylene dichloride Pyridine Stoddard Solvents 4,4′-thiobis(6-tert-butyl-m-cresol) o-Tolidine o-,m-, and p-toluidene 1,2,3-trichloropropane Uranium Vinylidene chloride Xylidene (mixed isomers)

†1998 Notice of Intended Changes includes kidney effects which were not listed previously



TABLE 6.2 Industrial Operation with Exposure to Nephrotoxicantsa Industrial operation Amalgam manufacturers Chemists Chloralkali Dry cleaning Manufacturing batteries Manufacturing cellulose acetate Metal degreasing Paint manufacturers Plumbers

Nephrotoxicant Mercury Chloroform Mercury Perchloroethylene Mercury, Lead, Cadmium Dioxane Perchloroethylene Lead, Cadmium Lead


List in alphabetical order.

consistent proteinuria, and cadmium-induced kidney damage may appear years after workers are removed from exposure. In Japan excessive cadmium intake was also linked to a peculiar form of renal osteodystrophy known as “ ouch-ouch disease” or “ itai-itai byo.” It has been proposed that this disease is caused by excessive loss of cadmium and phosphorus in the urine, combined with dietary calcium deficiency. The kidney naturally accumulates cadmium. Normally cadmium accumulates in the kidney over the lifetime of the individual until the age of 50. About 50 percent of the total burden of cadmium in the body is borne by the liver and kidney, with the kidney having 10 times the concentration of the liver. Cadmium induces synthesis in the liver of metallothionein, a protein with a high binding affinity for cadmium. While metallothionein acts to protect certain organs, such as the testes, from cadmium toxicity, it may play a role in cadmium toxicity in the kidney. After the available metallothionein in proximal tubule cells is overcome by high cadmium concentrations, the free cadmium exerts toxic effects on the cells in the proximal tubule. Chronic cadmium exposure has also been implicated as a factor in hypertension. However, while the development of hypertension may involve the kidney, the role of cadmium in the etiology of hypertension in humans is far from conclusive. Mercury Inorganic mercury (Hg2+) is a classical nephrotoxicant. It is used as a model compound for producing kidney failure in animals, and massive doses of mercuric ion can damage the proximal tubule and cause acute renal failure. A brief polyuria is followed by oliguria or even anuria. The anuria (kidney failure) leads, of course, to a life-threatening accumulation of bodily wastes and may last many days. If recovery occurs, a polyuria follows, which is probably caused by a decreased sodium absorption in the proximal tubule. Such disturbances in tubular function may last several months. Acute exposure to high concentrations of mercury is rare; usually mercury exposure occurs at lower dose rates. The part of the nephron most sensitive to mercuric ion toxicity is the pars recta or straight portion of the proximal tubule (Figure 6.3). Early damage is characterized by the presence of enzymes in the urine that are normally found in the brush border portion of the cells lining the tubule. Further damage results in the presence of intracellular enzymes from these cells in the urine. Longer-term exposure and damage can lead to the presence of glucose, amino acids, and proteins in the urine. Also associated with long-term exposure to mercury is a reduction in the GFR caused by vasoconstriction, tubular damage, and damage to the glomerulus. Chloralkali workers exposed to mercury have increased glomerular dysfunction and elevated excretion of high-molecular-weight proteins. 2-Microglobulin has been found at elevated levels in the blood plasma of these workers, but levels in urine were not increased. Lead Lead is a known nephrotoxicant in humans. Lead causes damage principally to the proximal tubule of the nephron. Reabsorption of glucose, phosphate, and amino acids is depressed in the



proximal tubule. This leads to glycosuria, aminoaciduria, and a hyperphosphaturia with hypophosphatemia. These changes are reversible on treatment with a chelating agent such as ethylene–diamine tetraacetic acid (EDTA), but only when the lead exposure has been relatively short. Long-term, prolonged exposure to lead may cause irreversible dysfunction and morphologic changes. This is manifested by intense interstitial fibrosis accompanied by tubular atrophy and dilation. The glomeruli are involved in later stages of the disease. Eventually, long-term lead exposure syndrome results in renal failure and death. There has been linkage of the chronic renal damage to saturnine (lead-induced) gout, in which uric acid is increased in the kidney.

Other Toxic Metals Table 6.3 lists those metals known to be toxic to the kidneys. As with most nephrotoxicities, the proximal tubule appears to be the most sensitive to toxic effects, with more extensive nephron involvement at higher dosages. In all animal and human exposures to uranium that are acutely injurious, the kidney is the main target of concern. Necrosis in the pars recta of proximal tubules and ascending limb of the loop of Henle and collecting tubules occurs, with accompanying loss of function. In the urine, there are increased casts, protein, glucose, catalase and other enzymes, amino acids, and 2-microglobulin; assays of urinary amino acid and 2-microglobulin appear to be the most sensitive in this regard. With recovery, areas of renal necrosis are at least partially replaced; however, the new cells may not be functionally equivalent to the original cells.

Halogenated Hydrocarbons Carbon tetrachloride (CCl4) and chloroform (CHCl3) are nephrotoxicants. Again, the proximal tubule appears to be the portion of the nephron most sensitive to damage by these agents. However, lesions are seen in other parts of the nephrons as well. It should be noted that carbon tetrachloride causes severe blood hepatic necrosis in humans, but the ultimate cause of death is kidney failure. The mechanism of the injury to the kidney is not known. However, it has been reported that backdiffusion of glomerular filtrate was important in the early stages of oliguria, and decreased renal blood flow contributed in the later stages of oliguria following carbon tetrachloride inhalation in humans. It appears that chloroform and carbon tetrachloride are activated to a toxic chemical species in the kidney by a mixed-function oxidase system similar to that found in the liver. The toxic metabolite covalently binds to tissue macromolecules in the kidney, and this leads to nephrotoxicity. The exposure levels leading to renal damage in humans have not been well defined. An increased incidence of proteinuria was reported in workers exposed to vapor concentrations of around 200 ppm, while the urine protein content was not changed after inhalation exposure to 50 ppm for 70 min or 10 ppm for 3 h. Bromobenzene, tetrachloroethylene, and 1,1,2-trichloroethylene also produce toxic effects to the kidney similar to those of chloroform and carbon tetrachloride. TABLE 6.3 Metal Nephrotoxic Agents Metals of Principal Concern Cadmium Lead Mercury

Other Metals Having Nephrotoxicity Arsenic Bismuth Chromium Platinum Thallium Uranium



Methoxyflurane (1,1-difluoro-2,2-dichloromethyl ether) is a halogenated surgical anesthetic that causes renal failure in humans and animals. Its causes a polyuria and an increase in serum osmolality, serum sodium, and blood urea nitrogen. Methoxyflurane is metabolized to inorganic fluoride anion and oxalate. The fluoride anion has been shown to be responsible for acting on the collecting tubules, which results in vasopressin resistance and causes polyuria. Bromomethane Humans exposed to high levels of bromomethane vapor commonly suffer from renal congestion, anuria or oliguria, and proteinuria; however, renal effects after exposure are frequently minimal or absent. Animal studies report similar signs of renal injury such as swelling, edema, nephrosis, and tubular necrosis. Hexachloroethane Hexachlorethane has been found to cause tubular atrophy, degeneration, hypertrophy, and/or dilation in rats. Other Nephrotoxic Compounds Methyl Isobutyl Ketone In rats exposed to as low as 100 ppm of MIBK, microscopic examination showed toxic nephrosis of the proximal tubules. The exposed rats also showed hyaline droplet degeneration of the proximal renal tubules and occasionally tubular necrosis. The tubular damage was considered transient and reversible. Dioxane Dioxane exposure can be significant from both an oral and inhalation exposure route. It has a selective action on the convoluted tubules of the kidneys, and causes renal obstruction. In several cases of fatal industrial exposure, injuries to the kidney were identified as causing the deaths. Signs of severe hemorrhagic nephritis and central hepatic necrosis occurred after about 2 months of what were considered to be heavy exposures to dioxane vapor. No cases of jaundice were observed, and death occurred within 1 week after onset of illness from acute renal failure. Phenol In subchronic toxicity tests with guinea pigs phenol causes renal proximal tubule swelling and edema and glomerular degeneration. Agents Causing Obstructive Uropathies A number of agents cause nephrotoxicity through physical deposition in the tubular sections of the nephron. Certain chemical agents can be concentrated in the tubular fluid to levels well above their solubility limit in water. The result is that crystals are deposited in the kidney tubules, causing physical damage. Methotrexate and sulfonamide drugs can cause nephrotoxicity by this mechanism. Acute renal failure of this type is also associated with the ingestion of ethylene glycol. Ethylene glycol is metabolized to oxalic acid by the body; the acid, in turn, is deposited in the lumen of the tubule of the nephron as well as within the cell of the tubule as insoluble calcium oxalate salt. However, ethylene glycol additionally appears to cause a nephrotoxicity to the proximal tubule, which is independent of oxalate deposition. The deposition of large quantities of oxalate crystals in the tubular elements of the nephrons probably contributes to the nephrotoxicity observed. Oxalate found in the leaves of rhubarb is of a sufficient quantity that it can cause deposition of oxalate crystals in the tubular elements of the nephron, and can lead to nephrotoxicity. Part of the nephrotoxicity caused by methoxyflurane is likewise believed to be caused by deposition of calcium oxalate crystals in the tubular elements of the nephrons. Agent Producing Pigment-Induced Nephropathies A number of chemicals can cause the release of certain pigments such as methemoglobin, hemoglobin, and myoglobin into the blood. When this occurs, an associated acute renal failure may develop. Arsine


NEPHROTOXICITY: TOXIC RESPONSES OF THE KIDNEY TABLE 6.4 Therapeutic Agents Known to Cause Nephrotoxicity Acetaminophen (analgesic) Aminoglycoside antibiotics Amphotericin B (antibiotic) Cephalosporadine (antibiotic) Colistimethate (antibiotic) Gentamycin Kanamycin Neomycin Polymyxin B (antibiotic) Streptomycin Tetracyclines (particularly outdated formulations) (antibiotics)

gas causes massive hemolysis of red blood cells, which results in hemoglobinuria and associated renal failure (see Chapter 4 for a general listing of hemolytic agents, methemoglobin formers, etc.). Heroin overdosage can result in a prolonged pressure on dependent muscles and a lysis of the muscle cell, leading to a release of myoglobin into the blood. Heroin may also cause some direct lysis of the muscle cells. The result can be myoglobinuria and ultimately acute renal failure. Aniline dyes are another group of chemicals that have been shown to release methemoglobin, with an associated renal failure.

Therapeutic Agents Table 6.4 lists a number of therapeutic agents known to cause nephrotoxicity. Acetaminophen is oxidized by the microsomal P450 oxygenase system in the renal cortex to a toxic metabolite. The microsomal P450 oxygenase system of the kidney is similar to that of the liver (see Chapter 5). Cephalosporadine reaches high toxic concentrations in the nephron because the organic ion transport system of the proximal tubule secretes it into the tubule. The nephrotoxicity of cephalosporadine can be diminished by compounds that compete with the organic anion secretion system in the proximal tubule, such as probenicid. The resulting decrease in tubular concentration of cephalosporidine in tubular fluid results in elimination of toxicity. Other therapeutic agents can, in certain individuals, elicit a nephrotoxicity by an allergic type of reaction. However, such nephrotoxicities are usually only rarely encountered.

6.4 SUMMARY The kidney performs a number of functions essential for the maintenance of life:

• Elimination of waste products (particularly nitrogen-containing wastes from the metabolism of proteins) from the blood

• Regulation of acid-base balance, extracellular volume, and electrolyte balance Toxic agents that disrupt these key functions can be life-threatening. The kidney is a highly metabolic organ sensitive to deprivation of oxygen, and any agent that significantly impedes renal flow will cause two adverse sequelae; acute renal failure that can result in death:

• First, less blood plasma will reach the kidney, resulting in a decrease in removal of blood wastes with a resulting increase of wastes in the blood (i.e., uremia).



• Second, if blood flow is compromised long enough, tissue ischemia will result in irreversible organ damage. Nephrotoxicants, agents toxic to the nephron, the principal excretory unit of the kidney, also disrupt key life-preserving functions.

• The glomerulus normally filters out the high-molecular-weight proteins from the blood. •

However, toxic agents will increase its permeability, allowing these proteins to appear in urine. Agents that damage the tubular element of the nephron will compromise its ability to reabsorb solutes such as glucose and amino acids, which are necessary for normal maintenance of the body, or disrupt sodium transport out of the nephron tubule, which could result in diuresis or excess urine formation or an unbalancing of the body’s ionic (salt) homeostasis. If damage to the nephron is excessive, renal failure can decrease or completely stop urine flow, and cause death by poisoning from the body’s own products.

Many agents directly toxic to the nephron are commercially or industrially important.

• Mercury, lead, and cadmium are industrially the most important nephrotoxic metals. • Halogenated hydrocarbons, particularly carbon tetrachloride and chloroform, are nephrotoxic.

• Certain therapeutic agents, such as phenacetin, aspirin, and the aminoglycoside antibiotics are directly nephrotoxic. Chemicals can cause nephrotoxicity indirectly:

• Some agents deposit crystals in the tubular element of the nephron, resulting in physical damage.

• Hemolytic agents such as arsine gas are capable of pigment neuropathy by releasing hemoglobin into the blood.

REFERENCES AND SUGGESTED READING American Conference of Governmental Industrial Hygienists, TLVs and Other Occupational Exposure Values— 1998, ACGIH, Cincinnati, OH, 1997. Agency for Toxic Substances and Disease Registry, Toxicological Profiles on CD-ROM, CRC Press, Boca Raton, FL, 1997. Berndt, W. O., “ Renal Methods of Toxicology,” in Principles and Methods of Toxicology, A. W. Hayes, eds., Raven Press, New York, 1982, pp. 447–474. Brenner, B., and F. Rector, The Kidney, Saunders, Philadelphia, 1976. Ganong, F., Review of Medical Physiology, Lange Medical Publications, Los Altos, CA, 1973, pp. 510–532. Goldstein, R., and R. Schnellman, “ Toxic responses of the kidney,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 5th ed., C. D. Klaassen, ed., McGraw-Hill, New York, 1996, pp. 417–442. National Institute for Occupational Safety and Health, NIOSH Criteria Documents on CD-ROM, CDC-NIOSH, Cincinnati, 1996. Pitts, R., The Physiology of the Kidney and Body Fluids, 2nd ed., Year Book Medical Publications, Chicago, 1968. Porter, G. A., and W. A. Bennett, “ Toxic nephropathies,” in B. M. Brenner, and F. C. Rector, eds., The Kidney, 2nd ed., Vol. II, Saunders, Philadelphia, 1981. Ullrich, K., and D. Marsh, “ Kidney, water, and electrolyte metabolism,” Ann. Rev. Physiol., 25: 91 (1963). Weiner, I., “ Mechanisms of drug absorption and excretion: The renal excretion of drugs and related compounds,” Ann. Rev. Pharmacol. 7: 39 (1967).

7 Neurotoxicity: Toxic Responses of the Nervous System NEUROTOXICITY: TOXIC RESPONSES OF THE NERVOUS SYSTEM


Neurotoxic chemicals are significant contributors to the human health effects that result from environmental and workplace chemical exposures. The National Institute for Occupational Safety and Health (NIOSH) reports that exposure to neurotoxic chemicals is one of the 10 leading causes of work-related disease and injury and that over 25 percent of the chemicals for which the American Conference of Governmental Industrial Hygienists (ACGIH) has established Threshold Limit Values (TLV) (trademarks) have demonstrated nervous system effects. Other sources have estimated that of the 400 or so commonly used chemicals (primarily solvents and various pesticides), 42 percent are neurotoxic. Thus neurotoxicity is an important consequence of human exposure to industrial chemicals. Studying toxic effects in the nervous system presents many challenges not usually encountered when working with other systems. Foremost is the sheer complexity of the human nervous system. Indeed, it is this complexity that in large part distinguishes us from other organisms and accounts for the exceedingly diverse spectrum of human behavior. As a result of its complexity, the nervous system displays a variety of responses to toxicant exposure. These may include changes in heart rate, breathing rate, sensory perception, coordination, mood, and many other physiological, behavioral, cognitive, and emotional effects. Quantitating these effects is sometimes difficult enough, but even when it is possible, the significance for human health may not be clear. For example, are feelings of euphoria or drowsiness toxic effects? Also, while a temporary decrease in reaction time may not in itself be life-threatening, in an industrial setting where the worker is surrounded by other hazards, a loss in the ability to react may result in disaster. Considerations such as these must become a part of the overall neurotoxicity assessment. Not all industrial chemicals are neurotoxicants, but for those that are, neurotoxic effects are often extremely sensitive indicators of low-level exposure. This, of course, depends on developing appropriate methods for monitoring such subtle effects. Changes in behavior are commonly used as sensitive and easily measured neurotoxic end points, although they may present some difficulty in terms of their objective quantification and baseline variability among individuals. Some of the standard tests for neurotoxicity are described later in this chapter. Besides being complex, the nervous system is ubiquitous, its network extending throughout the body. We can conveniently divide this network into the central nervous system, comprising mainly the brain and spinal cord, and the peripheral nervous system, comprising all other components, including sensory and motor nerves. This distinction is important for the purposes of neurotoxicity because, as described later, some neurotoxicants appear to target only the central or peripheral nervous systems, but not both. The brief overview of the nervous system in the next section will suggest several different ways in which neurotoxic chemicals may impair nervous system function:

• Outright neuronal destruction may result in permanent damage since neurons do not usually regenerate. Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




• Chemicals may disrupt the electrical impulse along the axon, either by harming the myelin • •

sheath or membrane integrity, or by impairing the synthesis or functioning of proteins essential to axonal transport. Chemicals may also inhibit the neurotransmitters by blocking their synthesis, release, or binding to receptors. General protein synthesis impairment may have an effect not only on neurotransmitter production, but also the production of important enzymes which break down neurotransmitters when they are no longer needed.

We will next consider the mechanisms of electrical and chemical signal transmission through the nervous system in more detail. The reader should keep in mind that proper nervous system function depends on all steps in signal transmission working properly, and a disruption in any step may result in what would be described as a neurotoxic effect.

7.1 MECHANISMS OF NEURONAL TRANSMISSION In one sense, the nervous system is little more than an enormous network of interconnected nerve cells, or neurons, supported by various other auxilliary cell types. However, this description is deceptive in its simplicity. Neurons come in many shapes, sizes, and functions, but may be generically described as having dendrites, a cell body, and an axon. The dendrites receive chemical signals from an adjacent neuron. These signals then trigger electrical impulses along the axon and in turn stimulate the release of more chemical signals at the terminal boutons. In this way, a stimulus may travel the entire length of the human body. The electrical impulse is often maintained along the length of the neuron with the aid of the myelin sheath, which acts as an insulator surrounding the axon. Successive neurons meet at a gap called the synapse, and it is across this gap that the chemical signals, or neurotransmitters, diffuse from one neuron to the dendrites of the next. Alternatively, neurons may terminate at muscles or glands, releasing neurotransmitters to specialized receptors at these sites. It has been found that these basic features of the nervous system are similar throughout a wide taxonomic range. Most multicellular organisms possess some form of nervous system which includes neurons, neurotransmitters, and electrical signal conduction. This similarity provides us with substantial confidence in using neurotoxicity test results in animals to predict neurotoxic effects in humans.

The Action Potential Electrical signals are initiated and propagated along the axon by what is called an action potential. The source of this potential is a charge difference across the nerve membrane, created by the movement of sodium (Na+), potassium (K+), and chloride (Cl–) ions. This charge difference is determined by the selective permeability of the membrane, as well as concentration and potential gradients, and active transport. When the membrane is at rest, the concentration of K+ ions is greater inside the cell than outside, while the concentrations of Na+ and Cl– are greater outside the cell. The concentrations of K+ and Cl– ions counterbalance each other, and this balance is maintained against their concentration gradients by the resulting potential gradient. Thus, in this equilibrium state, the tendency of either ion to diffuse across the membrane and down its concentration gradient is controlled by the imbalance caused by the potential gradient. Meanwhile, the membrane is relatively impermeable to Na+, and therefore a net positive charge exists on the outside of the cell relative to the inside (Figure 7.1a). When the cell is stimulated and an action potential is created, the membrane becomes locally permeable to sodium, and an influx of positive charge occurs. The result is a depolarization of the membrane that is propagated down the axon as current flows ahead of the action potential, depolarizing the membrane further (Figure 7.1b). Behind the action potential, the membrane permeability again shifts to favor K+ movement and to decrease Na+ movement. The resulting repolarization from the K+



Figure 7.1 Propagation of an action potential along a nerve fiber. (a) The resting electrochemical potential of a nerve. (b) Stimulation alters the sodium permeability of the nerve. (c) As sodium ions rush in, the adjacent gradient begins to depolarize, which increases sodium permeability and allows sodium to enter this part of the nerve as well. This action propagates down the nerve as a small, local current. (d) Repolarization begins in the same place the impulse started. The high positive charge inside the cell increases the potassium permeability. Potassium ions flow out of the cell and reestablish the resting potential. (e) Repolarization travels down the nerve until it is complete.

ions moving out of the cell brings the membrane state back to its original resting potential (Figure 7.1c).

Neurotransmitter Activity Communication between neurons and other cells occurs by both electrical and chemical signals. Electrical signals, the fastest means of communication, are transmitted between tightly packed neurons through membrane pores called gap junctions. The slightly slower chemical signals consist of neurotransmitters released at the synapse, which then bind to receptors on the postsynaptic cell,



triggering a response in that cell. The neurotransmitters include several types of molecules whose release from the neurotransmitter vesicles is stimulated by the action potential in the presynaptic cell. This process is illustrated in Figure 7.2. The most common neurotransmitter at synapses in the human body is acetylcholine (ACh), a component primarily of the peripheral nervous system. This chemical usually functions in an excitatory fashion, meaning that it initiates an action potential in the postsynaptic neuron, although it may also serve to inhibit signal propagation at some synapses. Whatever its role, ACh is hydrolyzed and its activity terminated by the enzyme acetylcholinesterase (AChE) after it has triggered a response in the postsynaptic cell, and in this way the signal is sent only once. Future signals are transmitted by newly synthesized ACh molecules released by the presynaptic neuron. Failure to hydrolyze older molecules,

Figure 7.2 Schematic model of a nerve synapse and the use of neurotransmitters to pass the stimulus from nerve to nerve. Neurotransmitters are released and stimulate receptors. Receptor stimulation increases cAMP levels, which affect sodium/potassium ATPase activity and hence the electrochemical gradient, membrane permeabilities, etc. Stimulus is ended either by (1) the breakdown of acetylcholinesterase, or (2) the reuptake of epinephrine in adrenergic nerves.



an effect seen with some neurotoxicants described later, will result in the inappropriate and continuous stimulation of the postsynaptic cell by the accumulated ACh molecules. Other neurotransmitters are known and include γ-aminobutyric acid (GABA—a component of the central nervous system), amines (epinephrine, norepinephrine, dopamine, serotonin), amino acids (glycine, glutamate), and peptides (enkephalins, endorphins). Their actions on the postsynaptic cell may be either excitatory or inhibitory, and may be directed toward another neuron, a muscle fiber, or a glandular cell.

7.2 AGENTS THAT ACT ON THE NEURON Membrane Disruption Effective transmission of neuronal signal depends on intact membranes, both along the length of the axon and at the terminus where neurotransmitter is released. In addition, the axons of many neurons are typically wrapped with myelin, which aids in the propagation of the action potential by minimizing any loss of potential across the membrane to the outside of the cell. Chemicals that disrupt the integrity of this membrane system can seriously impair nervous system function. Many commonly used industrial solvents, such as methanol, trichloroethylene, and tetrachloroethylene, are excellent cleaning agents and degreasers because of their lipophilicity. However, this same property also makes them destructive to the lipids in cell membranes. Coupled with the potential for substantial inhalation exposure resulting from their volatility, these membrane disrupters may pose a serious threat to the nervous system. Some metals are disruptive of the myelin sheath that surrounds neurons of the central nervous system and some of the peripheral nervous system. In an industrial setting, some of these metal compounds, such as lead, thallium, and triethyltin, may be readily inhaled in the course of smelting or soldering operations. They may then directly attack the myelin sheath, or disrupt the functioning of the accessory cells that myelinate neurons, namely, the Schwann cells and oligodendrocytes. The results of such damage may vary from the vision loss commonly seen with thallium poisoning to the impaired cognition associated with lead exposure. Many insecticides impair membrane function by interfering with the ion channels responsible for maintaining the proper balance of sodium and potassium ions across the membrane. An example is the organochlorine insecticide DDT (dichlorodiphenyltrichloroethane), which blocks ion channels and inhibits active transport, thus impeding the repolarization of the membrane after propagation of the action potential. The resulting symptoms, which include tremors, seizures, and increased sensitivity to stimuli, are seen in both the poisoned target insect and the accidentally exposed human. The use of DDT has been banned in the United States (though more because of its environmental persistence than its neurotoxicity), but many organochlorine insecticides, as well as some similarly acting pyrethroid esters, remain in use worldwide and are thus a potential hazard for workers involved in both their manufacture and application. Peripheral Sensory and Motor Nerves Effects on peripheral sensorimotor nerves are manifested as abnormal sensation and impaired motor control in the distal regions of the body, primarily hands and feet. Peripheral nerves often have the capacity to regenerate if damage is limited to the axonal region, so that quick action to limit exposure to the toxicant may result in complete reversal of the toxic effects. This is a fortunate contrast to CNS neuropathy, which usually results in permanent and irreversible damage. Some peripheral neurotoxicants, such as the solvents methyl n-butyl ketone and n-hexane, are thought to exert their effects mainly through the formation of toxic metabolites formed in the liver. The observed pathology is distal axonal swelling in both motor and sensory neurons, and is thought to be caused mainly by the metabolite 2, 5-hexanedione. A similar effect is seen as a result



of exposure to carbon disulfide, another common industrial solvent, although the role of its metabolites is less clear. Other chemicals may be transformed outside of the body into products that remain closely associated with the parent compound, leading to confusion about what is actually causing the observed effect. An interesting example is the effect on the cranial nerves that has long been seen with trichloroethylene inhalation. Trichloroethylene is known to target the central nervous system, while the cranial nerves are part of the peripheral nervous system controlling sensory and motor functions in the face and head. Recent studies suggest that the cranial neuropathy previously attributed to trichloroethylene inhalation may be caused by dichloroacetylene, an abiologically formed breakdown product of trichloroethylene that may occur in some industrial settings. Dichloroacetylene has clearly been shown to target the cranial nerves, whereas this has not yet been demonstrated with pure trichloroethylene.

Permanent Brain Lesions Damage to the neurons of the brain may produce varying results, depending on the affected area. Sensory, cognitive, or motor skills may be impaired, and the degree of effect may range from slightly debilitating to severe or even fatal. As with other CNS effects, those resulting from lesions in the brain are likely to be irreversible. Due to the high metabolic rate of neuronal tissue, the brain is by necessity highly perfused with blood vessels, and this presents a problem when a neurotoxic chemical finds its way into the bloodstream. Humans have evolved an important protection against potential brain toxicants known as the blood–brain barrier. This consists of several anatomical adaptations, such as tightly joined cells with few transport vesicles, which serve to decrease the permeability of membranes to many bloodborne chemicals. While it is quite effective at minimizing brain exposure to most large or hydrophilic molecules, the blood–brain barrier may still be traversed by some highly lipophilic molecules, as discussed below. The classic example is the neurotoxic metal mercury. When it exists in its ionized form or as an inorganic salt, mercury is water-soluble and, although it may circulate in the bloodstream of an exposed individual, it is not likely to cross the blood–brain barrier and cause damage to the brain. The neurotoxic symptoms of mercury poisoning, which may include tremors, mood disorders, psychosis, and possibly death, are manifested when the lipophilic elemental mercury or an organic mercury species is formed. The transformation of inorganic mercury to organic mercury is commonly performed by bacteria in the environment, but may also occur as a result of bacterial activity within the human gut. Since elemental mercury and organic mercury can cross membranes more easily than the ionized form, the blood–brain barrier presents a less formidable obstacle. Once inside the brain, elemental or organic form of mercury may be transformed into the ionized mercury, and thus remain there for a long time, producing severe brain lesions.

Anoxia In addition to those mentioned above, the nervous system depends on other important physiological functions that, if impaired by toxic agents, may result in symptoms of neurotoxicity. Of particular importance is the relationship between the nervous system and the respiratory system. The high metabolic rate of neurons requires that they be well supplied with oxygen and a rapid waste transport system. Compounds like carbon monoxide, which compete with oxygen for hemoglobin binding sites, may severely reduce the oxygen supply to neurons and eventually cause their deaths by anoxia. The most serious case would be destruction of neurons in the brain, leading to functional damage or death of the individual. Other compounds may produce the same result but through a different mechanism, as in the case of cyanide or hydrogen sulfide, which irreversibly bind to cytochrome oxidase, an essential component of the respiratory electron transfer chain.



When the affected neurons are those of the brain, the results are usually serious, as described in the previous section. However, life-threatening effects may also result from damage to the autonomic nerves, such as those controlling breathing and heart rate. This is often the effect of cyanide or hydrogen sulfide, which leads to death.

7.3 AGENTS THAT ACT ON THE SYNAPSE Anticholinesterase Agents When the enzyme acetylcholinesterase (AChE) is prevented from hydrolyzing acetylcholine (ACh), overstimulation of the postsynaptic cell results. This is an important mode of action for a variety of insecticides, two groups of which are the organophosphorus and the carbamate esters. However, as is the case with the organochlorines, what makes these anticholinesterase compounds effective insecticides also makes them potentially hazardous to humans who come into contact with them. Organophosphorus esters, or organophosphates, include malathion and parathion. These compounds bind to acetylcholinesterase (AChE), rendering it inactive in what is generally considered an irreversible reaction. Carbamate esters, such as sevin and aldicarb, also inactivate AChE by binding to it, although this reaction is considered reversible. Thus, carbamate poisoning is usually less severe than organophosphate poisoning. Since ACh is such a ubiquitous neurotransmitter in the body, the symptoms of anticholinesterase toxicity may take on a variety of forms. These may include decreased cognitive and motor skills and loss of autonomic nervous function, resulting in vomiting and diarrhea, seizures, tremors, and fatigue. Neurotransmitter Inhibitors and Receptor Antagonists Although many neurotoxic insecticides target AChE and thus the function of the neurotransmitter ACh, other insecticides exert their effects on other neurotransmitters. An example is the chlorinated cyclodienes, such as chlordane and endosulfan, which block the binding of the inhibitory neurotransmitter γ-aminobutyric acid (GABA) to its postsynaptic receptor. The indicators of toxicity are again the generalized symptoms of seizures, nausea, dizziness, and mood swings. Other chemicals that are implicated in neurotransmitter inhibition include carbon disulfide and DDT, which inhibit norepinephrine function, and manganese, which inhibits serotonin, norepinephrine, and dopamine function. Another commonly encountered chemical, nicotine, which is found in tobacco products and some insecticides, binds to a subset of ACh receptors that bear its name. These “ nicotinic receptors” are found throughout the central and autonomic nervous systems and at neuromuscular junctions. Their increased stimulation can lead to the well-known and often contrary symptoms of nicotine poisoning, such as excitability, nausea, and increased heart rate, followed by muscle relaxation, decreased heart rate, and sometimes coma or death.

7.4 INTERACTIONS OF INDUSTRIAL CHEMICAL WITH OTHER SUBSTANCES The effects of nicotine on the nervous system have already been discussed. While truly severe effects are expected only at high acute exposures, the lower but chronic exposure of a smoker to nicotine may combine with workplace exposure to other chemicals to produce an additive or even synergistic effect on the nervous system. The same is true for other neurotoxic chemicals to which a worker may be exposed outside the workplace, but that nonetheless serve to exacerbate the symptoms of neurotoxicity resulting from workplace chemical exposure. Perhaps the most common neurotoxicant of this sort is ethanol, which impairs axonal signal transmission by disrupting the sodium and potassium channels. This results in general CNS depression and uncoordination. It may also have serious consequences when combined with other



chemicals with depressive effects. Prescribed sedatives, such as barbiturates, may produce similar results. Caffeine is another commonly encountered substance with mild stimulatory effects resulting from the inhibition of cyclic AMP metabolism, a second messenger in the response cascade of postsynaptic nerve cells. However, the effects of caffeine on the nervous system are relatively mild, and are not thought to be particularly hazardous. More potent, and often illegal, stimulatory drugs, such as amphetamines and cocaine, may produce additive effects when taken in conjunction with workplace exposure to other neurotoxicants. The resulting symptoms may be life-threatening alterations in breathing and heart rates, or violent mood swings.

7.5 GENERAL POPULATION EXPOSURE TO ENVIRONMENTAL NEUROTOXICANTS Besides being workplace hazards, many neurotoxicants find their way into the environment through either deliberate or inadvertent release. Thus the general population may become exposed to these chemicals through the air, food, or drinking water. The infamous Minamata incident of the 1950s, in which residents of a coastal Japanese town suffered severe and occasionally fatal neurotoxicity from methyl mercury poisoning is only one example. In this case, exposure was due to ingestion of fish contaminated by discharge from a local acetaldehyde plant. Because humans are exposed to a variety of environmental chemicals, pinpointing the culprit causing a particular toxic effect can be difficult. Also, environmental exposure levels are usually, except in extreme cases, not so high as to cause blatant symptomology in all individuals. More often, the situation is such that only some members of an exposed population (presumably those that are most highly exposed or sensitive) show symptoms of varying degrees. In the case of neurotoxic effects, these often appear as changes in behavior that may serve as early warning signs of further toxicity possibly occurring if exposure is continued. Neurotoxic effects may be found among individuals exposed to dissolved metals such as lead and arsenic in their drinking water. Likewise, organic solvents, such as trichloroethylene and carbon tetrachloride, are now seen at various concentrations in most groundwater and surface water supplies. Often, these levels are not considered high enough to pose a serious risk, although we still do not know enough about the effects of low-level, chronic exposure to these chemicals on the developing nervous systems of fetuses and infants. Several epidemiological studies of populations living near water supplies contaminated with industrial solvents have reported elevated symptoms of neurotoxicity (e.g., decreased reaction times, reduced cognitive skills, mood disorders), although usually not at levels high enough as to be incontrovertible. Exposure to neurotoxicants through inhalation is seldom considered to be a serious threat outside an industrial setting. This is because volatile chemicals are easily dispersed in the atmosphere, so that only people living near industrial sources of air pollution may be subject to breathing elevated levels of neurotoxic chemicals. Usually, the primary concern is exposure within a confined space, such as the potential inhalation of volatile organic compounds from showering with contaminated water. Substantially increased levels of tetrachloroethylene have been measured in houses or automobiles in which newly dry-cleaned clothes are being stored or transported. However, the correlation between exposure to such levels and possible neurotoxic effects is as yet unknown.

7.6 EVALUATION OF INJURY TO THE NERVOUS SYSTEM Clinical Signs Taken as a whole, the evaluation of an individual for possible neurotoxicant exposure can generally be described as a series of steps, such as



• A study of the patient’s history, including diseases, chronic health problems, drug use, and exposure to other industrial or environmental chemicals.

• An evaluation of the patient’s mental status, as determined by various intelligence, memory, or mood tests.

• An evaluation of the patient’s sensory, motor, reflex, and cranial nerve function. These are assessed by simple, noninvasive tests, some of which are described below.

• An inspection of the patient’s work environment, which includes monitoring for the neurotoxic chemicals that may be implicated on the basis of results from the patient’s evaluative tests. Many clinical symptoms can often be indicators of CNS disturbance. These may include dizziness, vertigo, headache, mood swings, fatigue, memory loss, and various other cognitive disorders. Effects on the peripheral nervous system, specifically the peripheral sensory and motor neurons, are manifested by changes in breathing rate, heart rate, tendon reflex, perspiration, and gastrointestinal function. Standardized tests for cognition include IQ tests and performance with discriminatory tasks. An example is the Wechsler Adult Intelligence Scale (WAIS), in which the subject is presented with a list of words of increasing difficulty and is asked to provide a definition. Since the results obtained often depend strongly on the way in which the test is administered, and may be particularly vulnerable to hidden biases, evaluations of cognitive skills are not without controversy. Similar tests, which rely on the subject’s answers to certain questions, may be used to measure mood and memory effects. A difficulty with using such tests to evaluate possible neurotoxicant exposure in the workplace is that, in order to measure a change in cognitive skills, the individual’s preexposure level of skill must be available for comparison. This is rarely the case. Less subjective than cognitive tests are the standard physiological measurements, such as heart and breathing rate for autonomic nervous system effects, and sensory effects such as impaired hearing or vision. Decreased reaction time in response to a stimulus may indicate peripheral nervous system effects as well, and electroencephalograph (EEG), or “ brain wave,” measurements present a noninvasive method for monitoring the central nervous system. With some large, easily accessible neurons, as exist in the legs or arms, changes in the conduction velocity along the axon may be measured directly.

Behavioral Tests There is a vast array of available behavioral tests, which may be performed on workers to indicate potential neurotoxicity. These include measurements of reaction time to a stimulus, changes in dexterity as measured while performing various tasks, perception, motor steadiness, and general coordination. Sometimes cognitive and mood factors will be involved in determining the outcome of these tests as well. Several batteries of standard neurobehavioral tests have been developed, such as the World Health Organization Neurobehavioral Core Test Battery and the Finnish Test Battery, and are routinely used in industry around the world. An example of the type of neurobehavioral tests that may be administered is the Luria test for acoustic-motor function. The subject listens to a sequence of high and low tones, then repeats the pattern by knocking on a table with a fist for high tones and a flat hand for low tones. This test measures both acoustic perception and motor skills. Visual perception and hand dexterity may be measured with the Santa Ana dexterity test, in which the subject must rotate, within a given time, a pattern created with moveable colored pegs. Because behavior is determined by many factors, neurobehavioral tests are useful only as a first-step screening procedure for neurotoxicity. Taking preventative steps on behalf of workers requires knowledge of the neurotoxicants present, as well as their mechanisms of action which may result in the observed behavioral effects.



7.7 SUMMARY Evaluating neurotoxicity can be problematic for several reasons:

• The proper functioning of the nervous system depends upon many complicated steps occurring in a precisely controlled fashion, and thus an agent that is disruptive of any of these steps may be potentially neurotoxic.

• The operation of the nervous system is intimately involved with that of other systems, such as the respiratory, gastrointestinal, and endocrine systems, so that impairment of respiration or hormone function may occur in conjunction with, or as a result of, neurotoxic symptoms.

TABLE 7.1 Some Common Neurotoxic Chemicals Chemical Metals Arsenic Barium Lead Manganese Mercury (organic) Thallium Tin (organic) Organic solvents Acetone Benzene Carbon tetrachloride Carbon disulfide n-Hexane Methanol Tetrachloroethylene Toluene Trichloroethylene Organic pesticides Chlordane Cyanide 2,4-D DDT Endosulfan Lindane Malathion Parathion Rotenone Other Carbon monoxide Ethanol Hydrogen sulfide Nicotine Petroleum distillates


Site(s) of Action

Seizures, tremors Muscle spasms Insomnia, tremors Insomnia, confusion Ataxia, tremors, confusion Seizures, psychosis Headache, psychosis

Peripheral motor neurons Ion channels Myelin, synapse, axon Synapse Peripheral motor neurons, axon Myelin, axon Myelin

CNS depression Giddiness, ataxia CNS depression, giddiness Dizziness, psychosis

Blindness, mild inebriation Dizziness, ataxia Dizziness, euphoria Giddiness, tremors

Neuronal membrane Neuronal membrane Neuronal membrane Peripheral sensory and motor neurons, synapse Peripheral sensory and motor neurons, axon Neuronal membrane, axon Neuronal membrane Neuronal membrane Neuronal membrane, axon

Blurred vision, ataxia Confusion, labored breathing Muscle spasms, convulsions Dizziness, convulsions CNS depression, convulsions Headache, convulsions Blurred vision, headache Ataxia, convulsions Tremors, convulsions

Synapse Cytochrome oxidase Neuronal membrane Ion channels, synapse Synapse Synapse Synapse, axon Synapse, axon Electron transport chain

Headache, dizziness CNS depression, giddiness Convulsions, coma Excitability, nausea CNS depression, dizziness

Hemoglobin Ion channels, neuronal membrane Cytochrome oxidase Synapse Neuronal membrane

Numbness, giddiness



• Although much of the damage which may occur in the nervous system is irreversible, some protective adaptations and redundancy of function exist, which sometimes makes predicting the effects of neurotoxicant exposure less than straightforward. The chemicals cited in the sections above serve only as examples and are by no means a comprehensive list of neurotoxic hazards in the workplace. In addition, their symptoms of exposure are often varied and may be attributed to several mechanisms of action, which makes categorizing them by effect somewhat difficult. Recommended exposure limits for chemicals commonly encountered in an occupational setting are published by various agencies in the United States, such as the Occupational Safety and Health Administration (OSHA), the National Institute for Occupational Safety and Health (NIOSH), and the American Conference of Governmental Industrial Hygienists (ACGIH). In addition, several other countries have organizations with similar purposes. The publications of these groups should be consulted for specific recommended exposure levels of neurotoxic chemicals. Table 7.1 presents a few of the more common industrial neurotoxicants, along with some general symptoms of exposure and their primary sites of action. It should be clear to the reader by now that, as the result of multiple and overlapping effects of many neurotoxic chemicals, any listing of effects must necessarily be a simplified representation. Also, the science of neurotoxicity is continually evolving, so that revisions of such lists are to be expected as new information is obtained. Nonetheless, considerable progress has been made in recent years toward developing reliable methods of neurotoxicity evaluation and minimizing exposure to potential neurotoxicants.

REFERENCES AND SUGGESTED READING Annau, Z., ed., Neurobehavioral Toxicology, Johns Hopkins Univ. Press, Baltimore, 1986. Anthony, D. C., T. J. Montine, and D. G. Graham. “ Toxic responses of the nervous system,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 5th ed., C. D. Klaassen, ed., McGraw-Hill, New York, 1966, pp. 463–486. Anthony, D. C., and D. G. Graham. “ Toxic responses of the nervous system,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 4th ed., M. O. Amdur, J. Doull, and C. D. Klaassen, eds., McGraw-Hill, New York, 1991. Araki, S., ed., Neurobehavioral Methods and Effects in Occupational and Environmental Health, Academic Press, London, 1995. Baker, E. L. Jr., “ Neurologic and behavioral disorders,” in Occupational Health: Recognizing and Preventing Work-Related Disease, 2nd ed., B. S. Levy and D. H. Wegman, eds., Little, Brown, Boston, 1988. Chang, L. W., and W. Slikker, Jr., eds., Neurotoxicology: Approaches and Methods, Academic Press, London, 1995. Feldman, R. G., Occupational and Environmental Neurotoxicology, Lippincott, Williams and Wilkins Publishers, Philadelphia, 1998. Johnson, B. L., ed., Advances in Neurobehavioral Toxicology: Applications in Environmental and Occupational Health, Lewis Publishers, Chelsea, MI, 1990. Kilburn, K. H., Chemical Brain Injury, Van Nostrand-Reinhold, New York, 1998. Norton, S., “ Toxic responses of the central nervous system,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 3rd ed., C. D. Klassen, M. O. Amdur, and J. Doull, eds., Macmillan, New York, 1986. Office of Technology Assessment, Congressional Board of the 101st Congress, Neurotoxicity: Identifying and Controlling Poisons of the Nervous System, Van Nostrand-, New York, 1990. Tilson, H. A., and G. J. Harry, eds., Neurotoxicology (Target Organ Toxicology Series), Taylor and Francis, London, 1999. Tilson, H. A., and C. L. Mitchell, Neurotoxicology, Raven Press, New York, 1992. Weiss, B., and J. L. O’Donoghue, eds., Neurobehavioral Toxicity: Analysis and Interpretation, Raven Press, New York, 1994.

8 Dermal and Ocular Toxicology: Toxic Effects of the Skin and Eyes DERMAL AND OCULAR TOXICOLOGY


The skin is the body’s first line of defense against external toxicant exposure. Normal skin is an excellent barrier to many substances, but because of its enormous surface area (1.5–2.0 m2), it can act as a portal of entry for many diverse chemicals that come into contact with it, causing local and/or systemic effects. Understanding the composition of the skin and factors that influence the migration of chemicals across it are prerequisites to understanding the various manifestations of toxicant injury of the skin. Ocular toxicity will also be touched on in this chapter since many aspects pertaining to skin toxicity are relevant to ocular toxicity; the main difference is that the eye seldom serves as a significant portal of entry because of its small surface area. In this chapter you will learn about the

• • • • •

Composition of the skin Ability of the skin to defend against toxicants Types of skin maladies Commonly used tests to determine skin disorders Composition of the eye and exposures pertaining to ocular toxicity

8.1 SKIN HISTOLOGY The skin is composed of two layers: the outer epidermis and the underlying dermis. The two layers are firmly associated and together form a barrier that ranges in thickness from 0.5–4 mm or more in different parts of the body. The epidermis and dermis are separated by a basement membrane, which has an undulating appearance. The uneven interface gives rise to dermal ridges and provides the basis for the fingerprints used in personal identification since the patterns of ridges are unique for each individual. Hair follicles, sebaceous glands, and eccrine (i.e., sweat) glands span the epidermis and are embedded in the dermis. A third subcutaneous layer lays below the dermis and is composed mainly of adipocytes. Even though this layer is not technically part of the skin it plays an integral role by acting as a heat insulator and shock absorber. (See Figure 8.1.) The epidermis is composed of several layers of cells—some living and some dead. The majority of the epidermis is composed of keratinocytes, which undergo keratinization, a process during which they migrate upward from the lowest layers of the epidermis and accumulate large amounts of keratin (80 percent once fully mature and nonviable). By the time they reach the outer layer of the epidermis, the stratum corneum, the cells are no longer viable. They have become flattened and have lost their aqueous environment, which is replaced by lipids. The superficial cells of the stratum corneum are continuously lost and must be replaced by new cells migrating from the lower layers of the epidermis. The lowest layers of the epidermis immediately adjacent to the dermis (stratum germinativum and stratum spinosum) are responsible for the continual supply of new keratinocytes and initiation of the keratinization process. The migration and differentiation of keratinocytes from the lower viable layers Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




Figure 8.1 Diagram of a cross-section of skin. (Based on Doull, J., et al., eds., Casarett and Doull’s Toxicology: The Basic Science of Poisons, 2nd Ed. New York: Macmillan Publishing Company, 1980). Reprinted with permission.

to the upper stratum corneum take approximately 2 weeks, with another 2 weeks elapsing before the keratinocytes are shed from the surface. The lowest two layers of the epidermis also contain melanocytes, which produce the pigment melanin. Melanocytes extrude melanin, which is taken up by the surrounding epidermal cells, giving them their characteristic brown color. Langerhans’ cells are also found in these layers and play a role in the skin’s immune response to foreign agents. The dermis has a largely supportive function and represents about 90 percent of the skin in thickness. The predominant cells of the dermis are fibroblasts, macrophages, and adipocytes. Fibroblasts secrete collagen and elastin, thereby providing the skin with elastic properties. The dermis is well supplied with lymph and blood capillaries. The capillaries terminate in the dermis without extending into the epidermis. A toxicant must penetrate the epidermis and dermis in order to enter the systemic circulation; however, once the stratum corneum is breached, the remaining layers pose little resistance to toxicant penetration. Hair follicles are embedded within the dermis and have a capillary at the bulb of the follicle. In some instances, hair can enhance toxicant absorption across skin by providing a shunt to the blood supply at the base of the follicle. Eccrine glands are embedded deep within the dermis, and coiled sweat ducts wind upward through the epidermis and out through the stratum corneum.

8.2 FUNCTIONS The skin is an effective barrier to many substances, but it is a perfect barrier to very few. This is an important concept, since even though relatively small amounts of chemicals cross the skin, it can be sufficient to cause local and/or systemic toxicity. The passage of chemicals through the skin appears to be by passive diffusion with no evidence of active transport of compounds. The outer stratum



corneum is the primary layer governing the rate of diffusion, which is very slow for most chemicals. This layer also prevents water loss by diffusion and evaporation from the body except, of course, at the sweat glands, which helps regulate body temperature. The viable layers of the epidermis and the dermis are poor barriers to toxicants, since hydrophilic agents readily diffuse into the intercellular water and hydrophobic agents can embed in cell membranes, eventually reaching the blood supply in the dermis. Several factors influence the rate of diffusion of chemicals across the stratum corneum. In general, hydrophobic agents of low molecular weight can permeate the skin better than can those that are hydrophilic and of high molecular weight. This is due to the low water and high lipid content of the stratum corneum, which allows hydrophobic agents to penetrate more readily. However, if the skin becomes hydrated on prolonged exposure to water, its effectiveness as a barrier to hydrophilic substances is reduced. Often the skin of lab animals is covered with plastic wrap to enhance the hydration of the skin and increase the rate of uptake of agents applied to the surface of the skin. For compounds with the same hydrophobicity, the smaller compound will diffuse across the skin fastest since its rate of diffusion is quickest. A good example of the diffusion of a class of toxicants across the skin that can cause systemic toxicity is the organophosphate pesticides (e.g., parathion) used in agriculture. These compounds are hydrophobic, are very potent, and can lead to systemic effects such as peripheral neuropathy (i.e., nerve damage) and lethality after exposure to only the skin. The property of diffusion of agents across the skin and the reservoir capacity of the skin can be useful in delivering drugs to the systemic circulation over a prolonged period (typically 1–7 days). Transdermal drug delivery using specially designed skin patches is used to deliver nicotine, estradiol, and nitroglycerin. This approach provides a steady dose, avoids large peak plasma concentrations from loading doses, and prevents first-pass metabolism by the liver for agents that are sensitive to metabolism such as nitroglycerin. The rate of diffusion through the epidermis varies among anatomical sites and is not solely a function of skin thickness. In fact, the skin on the sole of the foot has a higher rate of diffusion than the skin of the forehead or abdomen, even though it is much thicker. Therefore, skin thickness is not a useful indicator of how much chemical will reach the systemic circulation in a given amount of time. If the skin is wounded, the barrier to chemicals is compromised and a shorter or direct route to the systemic circulation is available since the skin can no longer repel the chemicals. In addition, diseases (e.g., psoriasis) can compromise the ability of skin to repel chemicals. The skin also provides protection against microorganisms and ultraviolet (UV) radiation. Hydrated skin has a greater risk of becoming infected by microorganisms than does dry skin, which is why soldiers in Vietnam often suffered from foot infections. The stratum corneum and epidermis, but primarily melanin pigmentation, provide protection against UV radiation by absorbing the energy before it reaches more sensitive cells and causes adverse effects such as DNA damage. (See Table 8.1.) Another important aspect of the skin’s barrier function is its ability to metabolize chemicals that cross the stratum corneum and enter the viable layers of the skin. Even though the metabolic activity of the skin on a body weight basis is not nearly as great as that of the liver, it does play a crucial role in determining the ultimate effects of some chemicals. The epidermis and pilosebaceous units of the skin contain the highest levels of metabolic activity, which includes phase I (e.g., cytochrome

TABLE 8.1 Defense Roles of the Skin Prevent water loss Act as a barrier for physical trauma Retard chemical penetration Prevent ultraviolet light penetration and damage Inhibit microorganism growth and penetration Regulate body temperature and electrolyte homeostasis



P450-mediated) and phase II enzymes (e.g., epoxide hydrolase, UDP glucuronosyl transferase, glutathione transferase). Some chemicals that cross the skin are simply degraded and eliminated as innocuous metabolites. For others such as benzo(a)pyrene or crude coal tar (the latter is often used in dermatological therapy), metabolism of the parent compound can produce a metabolite that is a skin sensitizer or carcinogen. In addition to metabolizing foreign agents, the skin also has anabolic and catabolic metabolic activity important to its maintenance.


Irritants Irritant contact dermatitis is one of the most common occupational diseases. The highest incidence of chronic irritant dermatitis of the hands occurs in food handlers, janitorial workers, construction workers, mechanics, metal workers, horticulturists, and those exposed to wet working environments, such as hairdressers, nurses, and domestic workers. Contact irritant dermatitis is confined to the area of irritant exposure, and since it is not immunity-related, it can occur in anyone given a sufficient exposure to a chemical. Previous exposure to the chemical is not required to elicit a response as is needed for allergic contact dermatitis, since contact irritant dermatitis is not a hypersensitivity reaction (discussed below). A range of responses can occur after exposure to an irritant, including, but not limited to, hives (wheals), reddening of the skin (erythema), blistering, eczemas or rashes that weep and ooze, hyperkeratosis (thickening of the skin), pustules, and dryness and roughness of the skin. Unlike corrosive chemicals (e.g., strong acids and bases), the ultimate skin damage from irritant contact dermatitis is not due to the primary actions of the chemicals but to the secondary inflammatory response elicited by the chemical. It is important to note that even though the ultimate inflammatory response elicited by different chemicals may appear the same, they often occur through different mechanisms. A wide array of factors influence the ability of an irritant to elicit an inflammatory response. As discussed in Section 8.2, factors affecting skin permeability and chemical composition of the irritant determine the rate of percutaneous penetration and how much chemical reaches the viable layers of the skin. A variety of other factors determine whether irritant dermatitis occurs and to what magnitude. Higher concentrations and greater amounts of a given agent contacting the skin surface are more likely to elicit a response than lower concentrations and smaller quantities. The surface area of skin exposed to an irritant can also be important. For some irritants, a certain area of skin exposure is required to trigger a response, and below that threshold dermatitis does not occur. The genetic makeup and age of the individual plays a critical role in the sensitivity to a particular agent since the same chemical can cause no response in one individual and a dramatic response in another. The genetic factors influencing sensitivity are unknown, however. In general, children appear to be more, and the elderly less, susceptible to irritants. Concomitant disease may increase or decrease sensitivity to an irritant, and certain medications such as corticosteroids can suppress the irritant response to some agents. Extremes in temperature, humidity, sweating, and occlusion can lower the threshold of irritation for a given compound. The range of agents that can cause irritant dermatitis is extensive and diverse, and all cannot be touched on in this section. Table 8.2 lists some of the most commonly encountered classes of irritants. All of these agents have the potential of causing irritation on primary exposure; however, in the workplace, exposure to a potential irritant often occurs repeatedly and to relatively low quantities. Since the response is dependent on the amount of irritant to which the individual is exposed, repeated exposure may be required before clinical signs of dermatitis appear. Management of contact irritant dermatitis is based on reducing or avoiding the amount of exposure to the irritant. Wearing gloves to provide protection against wetness or chemicals and minimizing wet working conditions and hand washing can be very helpful. Complete healing of lesions may take several weeks, and the likelihood of a flare-up is often increased for months.



TABLE 8.2 Potential Inducers of Irritant Contact Dermatitis Agent Water Cleansers Bases Acids Organic solvents Oxidants Reducing agents Plants

Examples — Soaps and detergents Epoxy resin hardeners, lime, cement, and ammonium Hydrochloric acid and citric acid Many petroleum-based products Peroxides, benzoyl peroxide, and cyclohexanone Thioglycolates Orange peel, asparagus, and cucumbers

Source: Adapted from Rietschel (1985).

Extremely corrosive and reactive chemicals can cause immediate coagulative necrosis at the site of contact resulting in substantial tissue damage. These chemicals, called primary irritants, differ from those that cause irritant contact dermatitis in that they cause nonselective damage at the site of contact, which is not a result of the secondary inflammatory response. Primary irritants cause damage resulting from their reactivity, such as acids precipitating proteins and solvents dissolving cell membranes, both resulting in cell damage, death, or disruption of the keratin ultrastructure. The resulting damage is in direct proportion to the concentration of chemical in contact with the tissue. It is important to realize that primary irritants are not always in a liquid form. Many primary irritants are solid chemicals that become hydrated on contact with the skin, and gaseous agents are often converted to acids on contact with water available on the skin and mucous membranes. Ammonia, hydrogen chloride, hydrogen peroxide, phenol, chlorine, sodium hydroxide, and a variety of antiseptic or germicidal agents (e.g., cresol, iodine, boric acid, hexachlorophene, thimerosal) are some of the many commonly encountered primary irritants that can cause skin burns.

Allergic Contact Dermatitis Allergic contact dermatitis is a delayed type IV hypersensitivity reaction that is mediated by a triggered immune response. Typical of a true immune reaction, minute quantities of the allergenic agent can trigger a response. This differentiates it from irritant dermatitis, which is proportional to the dose applied. Allergic contact dermatitis can be very similar to irritant contact dermatitis clinically, but allergic contact dermatitis tends to be more severe and is not always restricted to the part of the body exposed to the chemical. On first exposure to the allergenic chemical, little or no response occurs. After this first exposure, the individual becomes sensitized to the chemical, and subsequent exposures elicit the typical delayed type IV hypersensitivity reaction. The allergenic agents (haptens) are typically low-molecular-weight chemicals that are electrophilic or hydrophilic. These agents are seldom allergenic alone and must be linked with a carrier protein to form a complete allergen. Some chemicals must be metabolically activated in order to form an allergen, which can occur within the skin as a result of the skin’s phase I and phase II metabolic activities. Sensitization occurs when the hapten/carrier protein (antigen) is engulfed by an antigen-presenting cell (e.g., macrophages and Langerhans cells) and the processed antigen is presented to a helper T cell (CD4+). The T cell produces cytokines that activate and cause the proliferation of additional T cells that specifically recognize the antigen. The secretion of cytokines also causes inflammation of the contact area and activation of monocytes into macrophages. The active macrophages are the ultimate effector cells of the reaction. They act to eliminate the foreign antigen and, through secretion of additional chemical mediators, enhance the inflammation of the contact site. Keratinocytes also play a role in the hypersensitivity reaction. They are capable of producing many different cytokines and



can act as antigen-presenting cells under certain circumstances. After the sensitization process occurs, subsequent exposure to the allergenic chemical triggers the same cascade of events as described above. However, the prior sensitization reaction resulting in a population of T cells specific for the antigen allows the cascade of events to proceed much faster. Table 8.3 lists some of the most common agents that trigger contact dermatitis. The actual number of potential allergenic agents is almost limitless. Individual sensitivity to a particular allergen varies greatly and is dependent on many factors, as discussed for irritant contact dermatitis. The genetic makeup of the person probably plays the greatest role in determining whether a response occurs. This is similar to the variability noticed among individuals for their sensitivity to IgE-mediated allergies, such as hay fever, in which some people respond while others do not. Patch testing is used to try to determine to which agent a person with suspected allergic contact dermatitis may be sensitive. Unfortunately, the test is usually limited to agents that are the most frequent causes of allergic contact dermatitis. As such, identifying sensitivity to an agent unique to a given occupation may be impossible. Patch testing should be performed by physicians trained and experienced in the technique, its pitfalls, and the subtleties of interpretation. If a compound is identified as allergenic, the sensitive individual can attempt to avoid exposure to that agent. The distribution of the allergic response on the body can also provide clues as to what the allergenic compound is. For example, linear stripes may indicate plant-induced dermatitis while a rash on the lower abdomen may indicate an allergy to a nickel-containing pants button. A variety of treatments are used to help alleviate contact dermatitis. The best treatment, however, is avoidance of the allergen or irritant. Baths and wet compresses, antibiotics, antihistamines, and corticosteroids are used in various combinations to treat contact dermatitis. A unique situation arises when a contact allergen is ingested or enters the systemic circulation. The most serious effects include generalized skin eruption, headache, malaise, and arthralgia. Flaring of a previous contact dermatitis, vesicular hand eruptions, and eczema in flexor areas of the body may be less dramatic disturbances. Systemic exposure can trigger a delayed type IV hypersensitivity reaction with subsequent deposition of immunoglobulins and complement in the skin, which are potent inducers of the secondary inflammatory response. Therefore, systemic exposure to a contact allergen may induce a widespread delayed type IV hypersensitivity reaction that is not localized to one area of the body.

Ulcers Some chemicals can cause ulceration of the skin. This involves sloughing of the epidermis and damage to the exposed dermis. Ulcers are commonly triggered by acids, burns, and trauma and can occur on

TABLE 8.3 Commonly Encountered Contact Allergens Source Plants and trees Metals Glues and bonding agents Hygiene products and topical medications Antiseptics Leather Rubber products



Rhus Nickel and chromium Bisphenol A, formaldehyde, acrylic monomers Bacitracin, neomycin, benzalkonium chloride, lanolin, benzocaine, and propylene glycol Chloramine, glutaraldehyde, thimerosal, and mercurials Formaldehyde and glutaraldehyde Hydroquinone, diphenylguanidine, and p-phenylenediamine

Poison oak and ivy Earrings, coins, and watches Glues, building materials, and paints Creams, shampoos, and topical medications Betadine

Rubber gloves and boots



mucous membranes and the skin. Two commonly encountered compounds that induce ulcers are cement and chrome. Urticaria Like allergic contact dermatitis, urticaria can be triggered by immunity-related mechanisms, and minute quantities of allergen can therefore trigger the reaction. Urticaria results in the typical hives, which are pruritic red wheals that erupt on the skin. Asthma is also a common occurrence after exposure to an inducer of urticaria. The symptoms often last less than 24 h. In severe cases, however, anaphylaxis and/or death may occur. The reaction is an immediate type I hypersensitivity reaction that is mediated by activated mast cells. The mast cells may be activated directly by the chemical (nonimmune), or activation may occur when the chemical acts as an allergen (immunity-mediated) and binds to the IgE immunoglobulins located on mast cells. When sufficient quantities of IgE become bound by the allergen or the mast cell is directly activated, the mast cell releases vasoactive peptides and histamine that cause the ultimate hive through activation of additional cellular components of the reaction. Most compounds that induce urticaria must enter the systemic circulation. Often urticaria is triggered by compounds to which the responder has a specific allergy, but induction by completely idiopathic mechanisms is also seen. Some potential nonimmune inducers of urticaria (i.e., direct activators of mast cells) are curare, aspirin, azo dyes, and toxins from plants and animals. A smaller number of agents may cause contact urticaria on exposure only to the epidermis. Cobalt chloride, benzoic acid, butylhydroxyanisol (BHA), and methanol have been reported to cause this form of urticaria. One of the most common inducers of contact urticaria seen in the medical community is caused by latex rubber products such as gloves. Natural latex rubber contains a protein that is capable of inducing an immediate type I hypersensitivity reaction. Simple contact with latex rubber products, such as gloves, can trigger the hypersensitivity reaction and cause hiving, asthma, anaphylaxis, and sometimes death. Toxic Epidermal Necrolysis Toxic epidermal necrolysis (TEN) is one of the most immediate life-threatening skin diseases caused by chemicals or drugs. Mortality is usually 25–30 percent, but can be as high as 75 percent in the elderly. Luckily, the incidence of TEN is fairly low, with approximately one person per million per year becoming affected. The disease is characterized by a sudden onset of large, red, tender areas involving a large percentage of the total body surface area. As the disease progresses, necrosis of the epidermis with widespread detachment occurs at the affected areas. Once the epidermis is lost, only the dermis remains, severely compromising the ability of the skin to regulate temperature, fluid, and electrolyte homeostasis. Since the epidermis is lost, the remaining dermis posses little resistance to chemicals entering the systemic circulation and to infection from microorganisms. The ultimate mechanism of drug or chemical induction of the disease has remained elusive. Recent evidence has implicated immunologic and metabolic mechanisms, but they are far from conclusive. TEN has been associated with graft–host disease, and, even though it is a controversial area, TEN is believed to be part of the same spectrum of disease as the Stevens–Johnson syndrome (erythema multiforme major), which is another serious reaction to drugs and infection. Acneiform Dermatoses Acne is a very disfiguring ailment, but fairly innocuous in terms of producing long-lasting damage to the skin. In the workplace, the most common causes of acne are petroleum, coal tar, and cutting oil products. They are termed comedogenic since they induce the characteristic comedo, which is either open (blackhead) or closed (whitehead). The black color of open comedones is due to pigmentary changes resulting in accumulation of melanin. The comedogenic agents produce biochemical and physiological alterations in the hair follicle and cell structure that cause accumulation of compacted



keratinocytes in the hair follicles and sebaceous glands. The keratinocytes clog the hair follicles and sebaceous glands and become bathed in sebum (released from the sebaceous glands). Halogenated chemicals—especially polyhalogenated naphthalenes, biphenyls, dibenzofurans, and contaminants of herbicides such as polychlorophenol and dichloroaniline—cause a very disfiguring and recalcitrant form of acne called chloracne. Chloracne is typically characterized by the presence of many comedones and straw-colored cysts behind the ears, around the eyes, and on the shoulders, back, and genitalia. Other symptoms that may or may not occur include conjunctivitis and eye discharge due to hypersecretion of the Meibomian glands around the eyelids, hyperpigmentation, and increased hair in atypical locations. Since chloracne is a very persistent disease, the best method of treatment is to prevent exposure to the halogenated chemicals. This could involve putting up splash guards and other devices to prevent the chemicals from coming into contact with the skin along with changing chemical soaked clothing frequently. Pigment Disturbances Some chemicals can cause either an increase or decrease in pigmentation. These compounds often cause hyperpigmentation (darkening of the skin) by enhancing the production of melanin or by causing deposition of endogenous or exogenous pigment in the upper epidermis. Hypopigmentation (loss of pigment from the skin) can be caused by decreased melanin production and/or loss, melanocyte damage, or vascular abnormalities. Some common hyperpigment inducers are coal tar compounds, metals (e.g., mercury, lead, arsenic), petroleum oils, and a variety of drugs. Phenols and catechols are potent depigmentors that act by killing melanocytes. Photosensitivity Photosensitivity is an abnormal sensitivity to ultraviolet (UV) and visible light and can be caused by endogenous and exogenous factors. Wavelengths outside the UV and visible light ranges are seldom involved, since the earth’s atmosphere significantly filters those wavelengths or they are not sufficiently energetic to cause skin damage. In order for any form of electromagnetic radiation to produce an effect, it must first be absorbed. Chromophores, epidermal thickness, and water content all affect the ability of light to penetrate the skin, and those parameters vary from region to region on the body. Melanin is the most significant chromophore, since it can absorb a wide range of radiation from UVB (290–320 nm) through the visible spectrum. Exposure to intense sunlight causes erythema (redness or sunburn) due to vasodilation of the exposed areas. Inflammatory mediators may be released at these areas and have been implicated in the systemic symptoms of sunburn such as fever, chills, and malaise. UVB is the most important radiation band in causing erythema. Sunlight has up to 100-fold greater UVA (320–400 nm), but UVA is 1000 times less potent than UVB in causing erythema. UVB exposure causes darkening of the skin through enhanced melanin production or through oxidation of melanin. Oxidation of melanin occurs immediately, but offers no additional protection against sun damage. Enhanced melanin production is noticeable within 3 days of exposure. UV exposure also enhances thickening of the skin, primarily in the stratum corneum, which further retards subsequent UV absorption. Chronic exposure to UV light can induce a number of skin changes such as freckling, wrinkling, and precancerous and malignant skin lesions. UV light is not the only type of radiation that can induce skin changes. Depending on the dose delivered, ionizing radiation can cause acute changes such as redness, blistering, swelling, ulceration, and pain. Following a latent period or chronic exposure, epidermal thickening, freckling, nonhealing ulcerations, and malignancies may occur. Phototoxicity results from systemic or topical exposure to exogenous chemicals. The symptoms are very similar to severe sunburn and include reddening and blistering of the skin. Chronic exposure can result in hyperpigmentation and thickening of the affected areas. Unlike sunburn, phototoxicity often results from exposure to the UVA band, but the UVB band is sometimes involved. Phototoxic chemicals are protoxicants (i.e., they are not toxic in their native form) and must be activated by UV



light to a toxic form. Phototoxic chemicals readily absorb UV light and become excited to a higher-energy state. Once the excited chemical returns to the ground state, it releases its energy, which can lead to production of reactive oxygen species and other reactive products that damage cellular components and macromolecules, ultimately causing cell death. The resulting damage is similar to that caused by irritant chemicals (discussed in Section 8.3) that cause cell death. Phototoxicant-induced cell death triggers an inflammatory response that produces the clinical signs of phototoxicity. Dyes (eosin, acridine orange), polycyclic aromatic hydrocarbons (anthracene, fluoranthene), tetracyclines, sulfonamides, and furocoumarins (trimethoxypsoralen, 8-methoxypsoralen) are commonly encountered phototoxic drugs and chemicals. Photoallergy is very similar to contact allergic dermatitis and is a delayed type IV hypersensitivity reaction. The difference between an allergenic chemical and a photoallergenic chemical is that the photoallergenic chemical must be activated by exposure to light—most often UVA. Once activated, the photoallergen complexes with cellular protein to form a complete allergen that triggers the delayed type IV hypersensitivity reaction. Since it is a hypersensitivity reaction, previous exposure to the phototoxic chemical is required for a response. Subsequent topical or systemic exposure to the photoallergen may induce the hypersensitivity reaction, which has clinical manifestations similar to allergic contact dermatitis (see the subsection on allergic contact dermatitis). Testing for photoallergy is similar to the patch testing used for regular allergens, but the potential allergens are tested in duplicate. One set of the patches is removed during the test and irradiated with UV light. By comparing duplicate samples, the physician can determine whether the compound is allergenic and is also a photoallergen. Skin Cancer Skin cancer is the most common neoplasm in humans with half a million new cases occurring per year in the United States. Even though exposure to UV light is the primary cause of skin cancer, chemicals can also induce malignancies. UV light and carcinogenic agents induce alterations in epidermal cell DNA. These alterations can lead to permanent mutations in critical genes that cause uncontrolled proliferation of the affected cells, ultimately leading to a cancerous lesion. UVB rays are the most potent inducers of DNA damage and work by inducing pyrimidine dimers. In addition to inducing DNA damage, UV light also has an immunosuppressive effect that may reduce the surveillance and elimination of cancerous cells by the immune system. Since UVB light is the most potent inducer of DNA damage, utilization of a sunscreen that blocks UVB radiation is critical in preventing skin cancer along with the other skin effects associated with UV light exposure. Ionizing radiation is also a potent inducer of skin cancer. Fortunately, ionizing radiation is no longer used for treatment of skin ailments such as acne and psoriasis, as was done in the recent past. The best characterized chemical inducers of skin cancer are the polycyclic aromatic hydrocarbons (PAH). In the 1700s, scrotal cancer was found to be prevalent among chimney sweeps in England. The compounds that induced the cancer were later determined to be PAHs present in high concentrations in coal tar, creosote, pitch, and soot. PAHs must be bioactivated within the skin, often to a reactive epoxide, by cytochrome P450 metabolism (discussed in Section 8.2) in order to cause DNA damage. The epoxides are electrophilic and can form DNA adducts that may produce gene mutations. Other carcinogenic agents may cause DNA damage through different mechanisms, but the ultimate lesion is a gene mutation that leads to a cancerous lesion. Eye Toxicity The eye is a very complex organ composed of many different types of cells. Disease, drugs, and chemicals can injure various parts of the eye with many different manifestations of injury. The most common cause of injury in an occupational setting is exposure of the cornea and conjunctiva to agents that are splashed onto the eye. Many other effects can occur to other parts of the eye such as the retina and optic nerve (see Figure 8.2), but they are usually limited to effects caused by drugs and various diseases. This section therefore focuses on external exposure of the eye to chemicals.



Figure 8.2 Diagrammatic cross section of the eye, with enlargement of details in cornea, chamber angle, lens, and retina. Casarett & Doull’s Toxicology: The Basic Science of Poisons, 5th Ed. McGraw-Hill, 1996. Reprinted with permission.



The structure of the eye is shown in Figure 8.2. The cornea and conjunctiva are the first line of defense against chemicals that contact the eye. Acids and alkalis are the more commonly encountered agents that cause eye damage. Acids cause protein damage, which leads to eye injury that can range in severity from burns that heal completely to those that perforate the globe. Alkalis such as ammonia can also cause serious eye burns. Alkali burns differ from acid burns in that they may lead to additional damage as time elapses, even if the burn was relatively mild at the time of injury. The best treatment for both types of substances is irrigation with large volumes of water, which reduces the acid or alkali concentration. Some compounds such as unslaked lime, which is found in many commercial wall plasters, may stick to the eye and form clumps that are not readily diluted or washed away with water irrigation. In these cases, irrigation followed by debridement of any remaining particles is required to remove as much contamination as possible. Two other agents that are frequent causes of eye damage are organic solvents and detergents. Organic solvents cause damage by dissolving fats in the eye. The damage is seldom extensive or long-lasting; however, if the solvent is hot, thermal burn may complicate the picture. Detergents act by disrupting proteins in the eye and lowering the surface tension of aqueous solutions. Detergents contain a nonpolar section and a polar section on the same molecule, allowing them to emulsify compounds with widely different hydrophobicities. They are commonly found in wetting agents, antifoaming agents, emulsifying agents, and solubilizers. Other parts of the eye can be affected by chemicals, either directly or as a result of the ensuing immune response that follows chemical burns. Many corrosive chemicals can cause lid damage and scarring of the puncta or canaliculi. Normal tear flow enters the lacrimal canaliculi in the lid margin via the puncta. The tears flow through the common canaliculus, lacrimal sac, and nasolacrimal duct into the nasopharynx. Damage of the puncta or canaliculi obstructs tear flow and can cause the tears to run down the cheeks. If a corrosive chemical penetrates the cornea and reaches the anterior chamber, it may cause damage to the iris. Damage to the iris increases vascular permeability with ensuing liberation of protein into the normally low-protein aqueous humor. These proteins can clog the outflow of fluid from the interior of the eye and lead to pressure buildup and glaucoma. Leukocytes may also infiltrate the aqueous humor from the inflamed iris vasculature and contribute to the blockage of the outflow system. Methanol is a unique eye toxicant since it affects the nerves of the eye and retinal and photoreceptor cells. Ingested methanol is metabolized to formaldehyde, then formate, then CO2 and water, with formate considered the toxic metabolite. Methanol intoxication can lead to appreciable, and sometimes permanent, loss of vision. Since methanol is first metabolized by alcohol dehydrogenase, ethanol can be used to prevent the formation of formate. The ethanol successfully competes for the alcohol dehydrogenase enzyme preventing the metabolism of methanol. Ethanol must be administered for a sufficient length of time so that all the methanol is eliminated from the body.

8.4 SUMMARY Toxicity of the skin and eye can occur after exposure to many different substances that cause injury through a variety of mechanisms. This chapter covered the major problems caused by chemicals encountered at home and at work, but a variety of other skin and eye diseases, including the ones mentioned in this chapter, can occur in reaction to systemically administered drugs. Whether a chemical can produce an effect after it comes into contact with the skin or eye depends on many factors, including genetic makeup, status of health, and efficiency of the skin’s barrier function. The following are some important points about eye and skin toxicity.

• The most common skin disease is irritant and allergic contact dermatitis, with allergic contact dermatitis usually being more severe.



• A person must first be sensitized to a chemical before allergic contact dermatitis can occur. • • • •

Since allergic contact dermatitis is an immune reaction, minute quantities of allergen can trigger the reaction, which makes management of future flare-ups difficult. Urticaria may or may not occur through immunity related mechanisms. The ultimate trigger of the hives associated with urticaria is due to the release of histamine and vasoactive agents from mast cells that are activated after chemical exposure. Phototoxicity and photoallergy are similar to irritant and allergic contact dermatitis, respectively. The difference is that the phototoxicant or photoallergen must be activated by exposure to UV light. Skin cancer is the most prevalent form of cancer. Its main cause is exposure to UV light, but many chemicals can induce cancerous lesions, too, such as polycyclic aromatic hydrocarbons and arsenic. The main cause of eye toxicity in the workplace is due to chemicals that are splashed onto the eye and cause corneal burns. Secondary events triggered by the burn can lead to further complications such as glaucoma.

REFERENCES AND SUGGESTED READING Bradley, T., R. E. Brown, J. O. Kucan, E. C. Smoot, and J. Hussmann. “ Toxic epidermal necrolysis: A review and report of the successful use of biobrane for early wound coverage,” Ann. Plastic Surg. 35: 124–132 (1995). Goldstein, S. M., and B. U. Wintroub, Adverse Cutaneous Reactions to Medication, Williams & Wilkins, Baltimore, 1996. Grandjean, P., Skin Penetration: Hazardous Chemicals at Work, Taylor & Francis, New York, 1990. Haschek, W. M., and C. G. Rousseaux, Handbook of Toxicologic Pathology, Academic Press, San Diego, 1991. Hogan, D. J., “ Review of contact dermatitis for non-dermatologists,” J. Florida Med. Assoc. 77: 663–666 (1990). Marzulli, F. N., and H. I. Maibach, Dermatotoxicology, Hemisphere Publishing, Washington, DC, 1987. Potts, A. M., “ Toxic responses of the eye,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 5th ed., C. D. Klaassen, ed., McGraw-Hill, New York, 1996, pp. 583–615. Rice, R. H., and D. E. Cohen, “ Toxic responses of the skin,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 5th ed., C. D. Klaassen ed., McGraw-Hill, New York, 1996, pp. 529–546. Rietschel, R. L., “ Dermatotoxicity: Toxic effects in the skin,” in Industrial Toxicology: Safety and Health Applications in the Workplace, P. L. Williams and J. L. Burson, eds., Van Nostrand-Reinhold, New York, 1985, pp. 138–161. Taylor, J. S., and P. Praditsuwan, “ Latex allergy: Review of 44 cases including outcome and frequent association with allergic hand eczema,” Arch. Dermatol. 132: 265–271 (1996). Wang, R. G. M., J. B. Knaak, and H. I. Maiback, Health Risk Assessment: Dermal and Inhalation Exposure and Absorption of Toxicants, CRC Press, Ann Arbor, MI, 1993. Zug, K. A., and M. McKay, “ Eczematous dermatitis: A practical review,” Am. Family Phys 54: 1243–1250 (1996).

9 Pulmonotoxicity: Toxic Effects in the Lung PULMONOTOXICITY: TOXIC EFFECTS IN THE LUNG


In this chapter the anatomy and physiology of the lung will be related to the most prevalent mechanisms of lung toxicity resulting from exposure to commonly encountered industrial toxins. Specifically, the chapter will discuss

• • • • •

Lung anatomy and physiology Defense mechanisms of the lung Different classes of chemicals that are known to damage the lung Four basic mechanisms by which industrial chemicals exert toxic effects on the lung Clinical evaluation of occupational lung injuries

When considering toxicology and the lung, it is important to note that the lung is both a target organ for many toxins and a major port of entry into the body, providing toxins the opportunity to exert toxic effects in other organs.

9.1 LUNG ANATOMY AND PHYSIOLOGY The lung and the rest of the respiratory system provide all the cells in the body with the ability to exchange oxygen and carbon dioxide. The same system can also provide many industrial toxins with entry to (and in some cases exit from) the body. Essentially, the respiratory system is an air pump, just as the heart is a blood pump for the circulatory system. Changes in the anatomy and physiology of the lung due to toxin exposure can often result in very severe health consequences for the exposed individual. An understanding of the structure and function of the respiratory tract is essential to understanding why so many individuals in occupational exposures suffer these toxicologic outcomes.

Upper Airway The entry-level area into the respiratory system is known as the nasopharyngeal region. The upper airway is generally considered to extend from the nose down to approximately the area of the vocal cords. Air that is inhaled into the nose enters the nasal openings and goes initially upward, then takes an abrupt turn and goes downward into the throat. Of course, humans can also choose to breathe through the mouth, in which the nasopharyngeal portion of the “ respiratory tree” is bypassed. In most instances, mouth breathing entails a calculated effort on the part of the individual and has been observed when the nasal pathway is blocked or obstructed and when the individual needs to dramatically increase the volume of breathing. Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




Inhaled air is highly “ conditioned” before it leaves the upper airway system. Relatively cold air, for instance, will be warmed to body temperature (37 °C) before it reaches the lung. In like manner, air that is at an elevated temperature will be cooled to body temperature within the nasopharyngeal system. The lung and the portion of the respiratory system below the upper airway is a very moist physiological system and is quite sensitive to humidity. The inhaled air is therefore highly humidified during its passage from the nares to the lung, and the air that enters the nares is cleared of the larger particles. The nose hairs function to some extent in this process, and the turbulent nature of the air passages in the nares also contributes to the deposition of the larger particles, preventing them from being inhaled into the lower passages of the respiratory system. The lining of the nasal wall is known as the mucosa and is highly inundated with blood capillaries. Therefore, air that is inhaled through the nose comes immediately into contact with mucosal surfaces, which only thinly separate the air from these blood vessels. Deposition of toxic chemicals in the upper airway system can therefore result in both toxicity to the mucosal tissue and absorption of the agent into the systemic circulation by way of these capillaries.

Sinus Cavities There are four pairs of hollow cavities within the skull that are lined with a mucosal lining that is similar to the lining of the nasopharyngeal region. In order to view these sinuses from different angles, Figure 9.1 shows a frontal view of the skull, while Figure 9.2 represents a sagittal view. Since the sinuses are connected to the nasopharyngeal airways through a number of small openings, inhaled air also enters the sinuses. Acute sinusitis can occur when inhaled airborne toxins irritate the surfaces of sinus mucosa. As in other parts of the respiratory system, irritation of the mucosal lining leads to an inflammatory response in these tissues. As a result of the inflammation, there is an accumulation of

Figure 9.1 Frontal view of the skull, showing frontal, maxillary, and ethmoid sinuses. (Reproduced with permission from W. O. Fenn and H. Rahn, Handbook of Physiology. American Physiology Society, Washington, DC, 1964.)



Figure 9.2 Sagittal view of the skull, showing nasal turbinates and sphenoid sinuses. [Reproduced with permission from Fenn and Rahn (1964) (see Figure 9.1 source note).]

mucous, and the poor drainage characteristics of the sinuses lead to the growth of bacteria. Some individuals who suffer some sinusitis have severe headaches while others may experience only a continuous “ postnasal drip.” Many factors can contribute to sinusitis, in addition to or in conjunction with inhaled toxins, such as allergic hypersensitivity, individual characteristics of the sinuses in each person, and climatic conditions.

Tracheobronchiolar Region The trachea is a tube surrounded with cartilaginous rings that connects the nasopharyngeal region with the bronchioles. This region is essentially a conducting airway system to the lungs. The bronchi are a sequence of bifurcating branches of tubes. Each tube divides into two or three smaller tubes, and each successive branch then divides into smaller tubes, and so on. (see Figure 9.3). The bronchi themselves do not allow for the absorption of oxygen or carbon dioxide across their surfaces; they are merely



Figure 9.3 Schematic representation of the subdivisions of the conducting airways and terminal respiratory units. (Reproduced with permission from E. R. Weibel, Morphometry of the Human Lung, Springer-Verlag, New York, 1963.)

conducting airway tubes. In the bronchi near the lung itself, very small air sacs, or alveoli, begin to appear (at about the nineteenth or twentieth division) and increase in frequency with proximity to the lung. The bronchi in this region are known as respiratory bronchioles. It is in these alveoli that gas exchange between the inhaled air and the blood circulatory system occurs.

Pulmonary System and Gas Exchange The number of alveoli in the lungs number in the hundreds of millions, although the size of each individual alveolus is quite small. The total surface area of the human lung, which results from the summation of these alveoli, approximates that of about one-third of the square footage of an average American home. In each alveolus, a thin wall separates the blood in the capillary vessels from the inhaled air in the alveolus. In Figure 9.4, the terminal bronchiole and the many surrounding alveoli can be seen in relationship to the pulmonary blood supply. The wall between the blood vessel and the alveolus is a combination of the capillary endothelium, a basement membrane adjacent to the capillary, the space between the capillary and the alveolus (known as the interstitial space), a basement membrane adjacent



Figure 9.4 Photomicrograph of lung tissue, showing the relationship of a terminal bronchiole (TB) and its accompanying blood vessel, the pulmonary artery (PA), to the alveoli. (Reproduced with permission from J. F. Murray, The Normal Lung. The Basis for Diagnosis and Treatment of Pulmonary Disease, Saunders, Philadelphia, 1976.)

to the alveolus, and the alveolar epithelium. In many instances, the red blood cells are just barely able to fit through the small capillaries, so the blood cell wall is often in very close proximity to this membrane complex with the alveolus. Figure 9.5 illustrates how the remarkable design discussed above facilitates gas exchange. Carbon dioxide and oxygen readily cross this membrane complex in a process of simple diffusion. Many inhaled airborne industrial chemicals will also readily cross this membrane and will enter the bloodstream. These potential toxins thus enter the blood circulatory system in a manner analogous to someone receiving an intravenous infusion of a drug. A unique view of the alveoli is provided in Figure 9.6. The small holes, called pores of Kohn, provide for some ventilation between adjacent alveoli. Toxicologic insult to the lung as well as various disease states can result in a functional derangement of this membrane system. Exposure to some chemicals may result in an increase in fluid in the interstitial space. If sufficient fluid accumulates, a condition known as pulmonary edema, gas exchange can be hindered sufficiently to result in severe difficulty in breathing and even in death. Damage to the membrane itself can result in scarring, which may increase the thickness of the membrane or decrease the elasticity of the lung tissue, or both. As with pulmonary edema, an increase in the thickness of the membrane can deleteriously affect pulmonary gas exchange. Alterations in elasticity make the work of breathing harder, which can decrease the volume of respiration as the individual tires from the increased effort required. Of course, whenever gas exchange or the volume of respiration is sufficiently decreased, the amount of oxygen pressure in the circulatory system will also decline. If this decline proceeds to a sufficient extent, affected individuals can become seriously compromised in their health status.



Figure 9.5 Electron micrograph of an alveolar septum, showing the various tissue layers through which oxygen and carbon dioxide must move during the process of diffusion. The surface of the alveolar spaces (AS) is lined by continuous epithelium (EP). The capillary containing red blood cells (RBC) is lined by endothelium (E). Both layers rest on basement membranes (BM) that appear fused over the “ thin” portion of the membrane and that are separated by an interstitial space (IS) over the “ thick” portion of the membrane. [Reproduced with permission from Murray (1976) (see Figure 9.4 source note).]

Physiologic Differences between Inhalation and Ingestion Following inhalation, the chemical goes directly into the bloodstream without being first processed by the gastrointestinal system. This can result in an extremely rapid uptake of an industrial chemical from the air. For some chemicals, this also results in an extremely rapid onset of toxicity following inhalation of the agent. Inhalation of a chemical might also result in a higher degree of toxicity than if the compound were ingested. This is because a chemical absorbed from the gut will go first to the liver, which is the primary metabolizing organ of the body. The liver thus has the opportunity to eliminate the compound before it exerts its effect in some other target organ. This is called the first-pass effect. When the chemical is inhaled, it bypasses the liver and the toxin has the opportunity to reach a specific target organ and exert some degree of toxicity before the liver has the opportunity to eliminate it. Particulates Many chemical and radionuclide agents are deposited in the respiratory tract in the form of solid particles or droplets, also referred to as aerosols, meaning a population of particles that remain



Figure 9.6 Scanning electron micrograph showing interior of an alveolus and its pores of Kohn. (Reproduced with permission from D. V. Bates et al., Respiratory Function in Disease. An Introduction to the Integrated Study of the Lung. Saunders, Philadelphia, 1971.)

suspended in air over time. Some terms that are also used are dusts, fumes, smokes, mists, and smog. Dusts, which result from industrial processes such as sandblasting and grinding, are considered to be identical to the compounds from which they originated. In contrast, fumes usually result from a chemical change in compounds during processes such as welding, in which combustion or sublimation occurs. Smokes result when organic materials are burned; mists are aerosols composed of water condensing on other particles; and smog is a conglomerate mixture of particles and gases that is prevalent in certain environments such as areas with mountains, plenty of sunlight, and periodic temperature inversions. The toxicity of inhaled particulates has been known for a long time, especially in relation to occupational exposure. The early (1493–1541) famous toxicologist Paracelsus described the relationship between mining occupations and pulmonary toxicity in the sixteenth century.

Particle Size In the case of particulates, size is the primary critical determinant of how much of and where the agents will be deposited. The range in particle size for various aerosols is generally as follows: dusts, up to 100 µm; fumes, from 10 Å to 0.1 µm; smokes, less than 0.5 µm. The pattern of airflow in the respiratory system and anatomic features of the exposed individual are also important. Most inhaled particles are not spherical, but highly irregular in shape. In order to categorize the highly heterogeneous nature of inhaled particles, the aerodynamic diameter is calculated for the population of particulates of interest. This value is based on the settling velocity of the population of particles and roughly approximates what the particles’ diameter would be if it were compared to a



spherical particle in the time it takes the particles to settle in the air. This calculation is also referred to as the mass median aerodynamic diameter. If the number of particles is of primary interest (and not necessarily particle shape), the count median diameter is determined. Of course, the size of particles may change during the course of traversing the respiratory tract. Since the respiratory tract is highly humidified, particles that absorb water could be expected to undergo chemical reactions and increase in size as they descend. Lung Deposition Mechanisms Particles tend to deposit in the lung according to size, air velocity, and regional characteristics of the respiratory system. In the nares, nose hairs tend to block out the very large particles that enter the nose. Once inside the nares, the abrupt turn in the nasopharyngeal system of humans (from going up to going down) results in the impact of many of the larger particles on the walls of this region of the respiratory system. This mechanism, referred to as impaction, results from the aerodynamic tendency of particles to travel in a linear direction, even when the respiratory system is turning and branching. An analogy would be a bifurcating freeway system, in which the safety department will often place barrels at the point of bifurcation since cars are most likely to strike this location. In a similar manner, particles are more likely to strike the points of bifurcation in the respiratory system. A related mechanism of deposition is known as interception. This process occurs when a particle comes close enough to contact a respiratory surface and, subsequently, deposits there. Interception does not have to occur at the bifurcations or turns and is mostly a factor in the deposition of fibers, which are much longer than other forms of particles. It is not uncommon for a fiber to be only a few µms in diameter and several hundred µms in length, so the probability of contact with the respiratory surfaces is enhanced. In the tracheobronchiolar region, the declining airflow allows gravitational influences to result in the deposition of particles in the 1–5 µm range. This process, referred to as sedimentation, increases in frequency as the particles in this size range descend lower into the bronchiolar tree. Sedimentation can also occur in the alveolar region, but the simple process of diffusion will result in the deposition of particles in the 1-µm range. Clearance Mechanisms The respiratory system has an extraordinary design for the clearance of particles and other toxins. Generally, the clearance mechanism is related to the site of deposition. This respiratory clearance should not be confused with total body clearance or systemic clearance in the pharmacokinetic sense. Respiratory clearance removes particles and other toxins from the respiratory tree; ultimate removal from the body is achieved through the gastrointestinal system, the lymphatics, and the pulmonary blood. In the nasopharyngeal and tracheobronchial regions, there is a mucociliary escalator mechanism. In the respiratory wall, there are pseudostratified columnar epithelial cells together with specialized goblet cells, which produce a layer of mucous along the wall of epithelial cells. Hundreds of cilia, which resemble small hairs, protrude from the epithelial cells (Figure 9.7). The mucous itself is in two layers: the lower layer, known as sol, contains the cilia and is thin and watery so that cilia movement is not impeded; the upper layer, the gel, is thick and viscous. The cilia beat in unison and move the gel layer along like a continuous sheet (Figure 9.8). Inhaled particles and other toxins become trapped on the gel layer. In the tracheobronchial region, the cilia beat upward, and the entrapped particles in the gel are propelled up toward the mouth. Typically, an individual will solubilize the material in saliva, which is then eliminated via the gastrointestinal tract. Occasionally the material may be coughed out of the body. In the nasopharyngeal region, the cilia beat downward toward the mouth and rely on the same mechanisms of removal. Typically, mucociliary clearance will occur within hours of the deposition of most particles, and in healthy individuals, the process is usually completed within 48 h.



Figure 9.7 Scanning electron micrograph of the luminal surface of a bronchiole, showing the cilia. The mucous layer has been removed. [Reproduced with permission from Ebert and Terracio, “ The Bronchiolar Epithelium in Cigarette Smokers,” Am. Rev. Resp. Disease 111, 6 (1975).]

In the alveolar region, macrophages provide a mobile and effective defense against particles, bacteria, and other offensive agents that reach the lower respiratory tree. Chemotactic factors are released when these inhaled agents deposit in the lung, and these factors alert the phagocytic cells to the location of the agents. The macrophages then engulf them and attempt to ingest them with proteolytic enzymes. An example of a macrophage moving from one alveolus to another through a connecting pore is shown in Figure 9.9. A very wide variety of potentially toxic agents, including viruses, bacteria, chemicals, and particles of many sizes, can be successfully broken down by macrophages. However, in certain situations, such as in unhealthy individuals (e.g., long-term smokers), the macrophages might be inefficient or in lower numbers, and this defense might be abrogated to a significant extent. Additionally, some particles are not particularly digestible by the macrophages. In such cases, as with tuberculosis infections and with some fibers, the macrophage may

Figure 9.8 Schematic representation of the mucociliary blanket, showing the wavelike motion of the cilia within the sol layer. [Reproduced with permission from A. C. Hilding, “ Experimental studies on some little-understood aspects of the physiology of the respiratory tract,” Trans. Am. Acad. Ophthalmol. And Otol. (July–Aug. 1961).



Figure 9.9 Scanning electron micrograph of the interior of an alveolus showing pores of Kohn (P) and a macrophage (arrow). [Reproduced with permission from Murray (1976) (see Figure 9.4 source note).]

rupture and spill the proteolytic enzymes into the lung tissues and damage them. If successful phagocytosis has occurred, the phagocytized material is then removed by either the mucociliary escalator or by lymphatic drainage. The action by the macrophages is initially very rapid, with inhaled particles engulfed by some macrophages within minutes of inhalation.

Gases and Vapors Many injuries to the lung and to distant organs have been known to occur following inhalation exposure to gases and vapors, especially in the workplace. Most industrial chemicals can exist in the gas or vapor state under certain situations, and various industrial processes can create even the fairly extreme physicochemical conditions necessary to vaporize potentially toxic agents. Everyday in the workplace, millions of workers are exposed to countless potentially toxic chemicals in the form of gases and vapors. The potential for highly toxic outcomes from inhalation exposures to gases and vapors is related to the fact that once they are inhaled into the lung, they can pass directly into the bloodstream. In a pharmacokinetic sense, inhaled gases and vapors are injected into the bloodstream as a patient would receive a drug through an intravenous (or intraarterial) infusion. Once a gaseous chemical enters the alveolar spaces of the lung, it can cross the relatively permeable alveocapillary membrane complex and enter the pulmonary blood. This complex consists primarily of the capillary and alveolar membranes, separated by an interstitial space (sometimes with fluid in it). The lining of the alveolar membrane also has a lining of surfactant (dipalmitoyl lecithin), which serves to equalize the inflation pressures of the heterogeneously sized alveolar sacs. The passage of the inhaled gases and vapors across the alveocapillary membrane complex, or the diffusion efficiency, is influenced by several factors. The solubility of the inhaled compound is important, as highly water-soluble compounds are often more likely to deposit in the upper respiratory



system, before reaching the alveolar regions of the lung. The condition of the alveocapillary membrane is also important. Poor health conditions in a patient might lead to the engorgement of the interstitial space with fluid, which would impair the diffusion of toxic chemicals across the alveocapillary membrane. While this protects the affected individual from the toxic effects of the inhaled chemical, it also prevents the free exchange of oxygen and carbon dioxide, which can have obvious life-threatening outcomes. The degree of uptake of inhaled gases and vapors can be quite significant in workers in many occupations. Following the initiation of inhalation, rapid uptake of perchloroethylene, a commonly used dry cleaning solvent for which there are thousands of potential exposures, can be observed in many different tissues (Figure 9.10). In this case, the uptake of perchloroethylene in circulating blood and seven tissues was remarkably rapid, and for many industrial chemicals, it is often within minutes of exposure. It is often interesting to note that the levels of the inhaled solvent remained fairly constant throughout the inhalation exposure period. This can have important ramifications in occupational exposures, as workers who enter an environment with a potentially toxic gas can experience systemic toxic effects almost immediately, and these effects can persist for long periods of time (while the inhalation exposure period continues). For instance, many industrial solvents cause neurobehavioral depression following inhalation exposure, and workers have been known to be injured as a result of falls or mishaps with industrial machinery almost immediately after breathing the chemicals. Obviously, the length of exposure affects the amount of chemical inhaled. However, for many gases and vapors a steady-state equilibrium can be established after a certain period of inhalation exposure. In this way, the level of chemical in the blood does not continue to increase, despite the continued inhalation exposure to the compound (Figure 9.10). This has important ramifications in industrial exposures because it helps explain why workers sometimes do not experience toxic effects to certain chemicals despite long-term exposure.

Figure 9.10 The uptake and disposition of perchloroethylene (PER) in the blood and seven tissues of laboratory rats is shown. The animals inhaled 2500 ppm of perchloroethylene for 120 min in dynamic inhalation exposure chambers, and blood and tissues were analyzed for the solvent by electron capture-gas chromatography. (Supported by US Air Force Grants AFOSR 870248 and 910356.)



Air-Pollutant Gases Many of the air pollutants are inhaled as gases, such as carbon monoxide, sulfur dioxide, and the various oxides of nitrogen. By far, the number one killer as far as toxic gases are concerned is carbon monoxide. The incomplete burning of various fuels results in the emission of carbon monoxide, and every year there are many deaths and injuries from individuals who breathe this gas in an enclosed space. While some of these are suicides, there are also many industrial exposures to carbon monoxide and other combustion pollutants. A number of air pollutant gases are produced by a complex interaction of sunlight, humidity, temperature, hydrocarbons, and the oxides of nitrogen. These interactions generate smog, as well as other gases such as ozone and the aldehydes.

Tobacco Smoke Toxicity resulting from the intentional and unintentional inhalation of tobacco smoke is an important consideration given its enormous magnitude of incidence, its interaction with the toxicity of other inhaled industrial pollutants, and its representation of the toxicity of both particulates and gases. The number of people who die and are significantly injured each year in the United States due to inhalation exposures to industrial chemicals cannot be stated with certainty; however, it is definitely much smaller than the number of people who die and are experiencing diminished health status as a result of tobacco smoke inhalation. The smoking of tobacco products causes pulmonary emphysema, chronic bronchitis, and lung cancer in many thousands of Americans each year. Interference with Pulmonary Defense Tobacco smoke inhalation results in the derangement of the pulmonary defense mechanisms necessary to protect against the inhalation of industrial toxins. It has been shown that, following chronic cigarette smoking, the cilia in the mucociliary escalator become increasingly paralyzed. The decrease in ciliary activity slows or prevents the removal of deposited toxins from the nasopharyngeal and tracheobronchial regions, as the gel layer becomes more sedentary. Many of the more than 2000 components of tobacco smoke are known to be respiratory irritants, and these irritating properties lead to an increased production of mucous in the respiratory system. Therefore, there is a decreased movement (and removal) of mucous simultaneously with an increase in mucous production. Eventually, some of the airways can become impeded and even blocked, severely limiting the respiratory volume of the affected individual. Sometimes the overworked mucous glands will increase in size sufficiently to block the airways themselves, further impeding airflow and increasing resistance. It has been shown that the cellular defense mechanisms of the lung, particularly the alveolar macrophages and the alveolar polymorphonuclear leukocytes, are significantly impacted by tobacco smoke inhalation. In many cases, these cells may be killed, causing the release of proteolytic enzymes, which come in contact with the respiratory membrane surfaces. Pulmonary emphysema can result, if this process is extensive, from the severe rupturing of the septa walls. Even short of cell death, these cells become less efficient in the removal of particulates and other toxins. Therefore, the inhalation of toxic agents in industrial environments has the potential to exert greater toxicity in smokers than in equally exposed nonsmokers. This has been shown repeatedly for many exposures to toxic chemicals in occupational studies, such as with asbestos. For this reason, occupational epidemiologists and physicians will often look for correlations between toxicity in an industrial worker population and tobacco use. Lung Cancer and Tobacco Smoke Bronchogenic carcinoma data from the 1980s estimated that approximately 90 percent of the more than 100,000 lung cancer cases each year in the United States are due to tobacco smoke inhalation. A very distressing aspect of this unpleasant data is that the incidence of lung cancer, previously occurring more often in men, is growing rapidly in the female population. The increasing incidence of tobacco smoke inhalation by women has been followed in an appropriate timeframe by an explosion in lung cancer cases in women. Whereas breast cancer was



previously the number one cause of cancer deaths in women, now this dubious honor is being replaced by lung cancer, as is the case in men. Women are also entering the industrial environment in increasing numbers, pursuing occupations previously held predominantly by men. This now incites the question of whether there will be a correlation between this increased smoking incidence among women and the incidence of cancer from industrial chemicals.


Irritation of Respiratory Airways One of the most common toxicity manifestations from inhaled agents in industrial exposures is the irritation of the airways, resulting in breathing difficulties and even death for the exposed individual. Often, this response results from bronchoconstriction, as the airways react to diminish the extent of the unwanted exposure. This can be a protective mechanism, if the affected person can quickly remove himself/herself or be removed from the offending agent. Of course, diminished inhalation over any extended period of time has obvious deleterious effects for the worker. The chemical warfare agents, chlorine and phosgene, exert immediate toxicity by airway irritation. If the level of exposure is sufficient, the exposed individual can die within minutes of the initiation of exposure. Often a high dose exposure is accompanied by dyspnea (difficulty in breathing, either real or perceived), cough, lacrimation (tears), nasopharyngeal irritation, dizziness, and headache. The dose response for chlorine exposures is summarized in Table 9.1. An interesting aspect of most industrial inhalation exposures involving the irritation of the airways is that the symptoms appear very serious at first, but seldom result in permanent respiratory damage. The coughing and choking are very alarming to both the affected individual and onlookers (including medical personnel), and at least should result in the injury being taken seriously (which is often a problem in industrial toxicity episodes). Chest X rays and pulmonary function tests should be conducted on these individuals, in case there are permanent or late onset toxicity manifestations such as pulmonary edema. Although most of these individuals will recover completely, many people have died from irritation of the airways following industrial chemical inhalation, and every incident must be treated as a serious episode. It is highly recommended that workers have a baseline pulmonary function test on file with which to compare after an irritant exposure.

Fibrosis and Pneumoconiosis A variety of lung diseases resulting from the inhalation of dusts has been encountered in occupational environments. The disease mechanism, known as fibrosis, results when the lung gradually loses elasticity as a result of the pulmonary response to long-term dust inhalation. The disease condition is referred to as pneumoconiosis, derived from the Latin and Greek root words pneumo, which means breath or spirit, and coniosis, which means dust.

TABLE 9.1 Chlorine Dose–Response Relationships <4 ppm 15 ppm 30 ppm >40 ppm >1000 ppm

Can be tolerated up to 30 min Severe respiratory symptoms begin Coughing, choking, chest pain Pulmonary edema Immediate death



Silicosis Following long-term inhalation of silica-containing dusts, many workers have developed irreversible lung damage known as silicosis. One-half to two-thirds of the rocks in the crust of the planet contain silica, so it is to be expected that many industrial processes result in the production of silica-containing dusts. While some of the inhaled silica dioxide crystals will deposit in the nares and on the mucociliary escalator, a certain number will reach the alveolar regions of the respiratory system. Unfortunately, the alveolar macrophages that ingest the silica particles will be damaged by the silicic acid produced following phagocytosis. Damaged and killed macrophages will release phagocytic enzymes into the alveolar sacs, which will result in their progressive destruction over time. This eventually results in a “ stiffening” of the lung tissues, which makes breathing more difficult for the affected patient. Over a long period of time, the body will try to wall off the area, resulting in the development of a silicotic nodule. Patients with advanced silicosis often have greater susceptibility to respiratory infections such as tuberculosis. In any one patient, one might find each of these stages located in the same lung. Even after an individual has been removed from the further inhalation of silica dust, this progressive deterioration will continue. Another negative aspect of the disease is that it is very difficult to treat, and currently, clinicians can do little more than alleviate symptomatic suffering.

Asbestosis The highly effective flame retardant asbestosis has been used for centuries, and in the past few decades, it has been used in industry for a variety of purposes. Many thousands of workers have received very high doses of asbestos in the shipbuilding industry. Usually, insulation workers were exposed to asbestos dust in very enclosed spaces, which tended to increase the concentration of the inhaled fibers. Countless individuals have been exposed to asbestosis fibers while working with the brake linings of cars. Chrysotile, or “ white” asbestos, accounts for about 90 percent of the asbestos in industrial applications; the amphiboles account for most of the other potential exposures, in which crocidolite, or “ blue” asbestos, is the most important (and was the first form found to be carcinogenic). The insidious nature of asbestosis is that major symptoms seldom appear until 5–10 years (or longer) after the inhalation of the asbestos fibers. As with silicosis, the inability of macrophages to digest the fibers leads to a progressive fibrosis of the lung tissue. However, with asbestosis there is also pleural thickening and calcification, which can be picked up by X-ray examination in the relatively early stages of the disease. Pleural calcification may exist in patients when there are no other symptoms present. Pulmonary function tests are often useful, in that decreases in compliance and total lung capacity are observed. A pathologic finding in asbestosis is the appearance of “ asbestos bodies,” which are structures formed by the protein encapsulation of asbestos fibers that resemble a “ barbell” in weight lifting (the protein is thicker on the ends). Asbestosis eventually leads to the development of malignant neoplasms in the respiratory tract. One form of cancer, mesothelioma, is so rare in situations outside of asbestos exposure that many physicians consider it a “ marker” disease for asbestosis. A higher incidence (up to an 80-fold increase) of bronchogenic carcinoma is distinctly correlated with tobacco smoke inhalation and asbestos exposure. These asbestos related cancer deaths generally occur from 25–40 years after the asbestos inhalation.

Excess Lung Collagen Most types of pulmonary fibrosis involve distinct changes in the proportion of the types of lung collagen that is produced in the affected lung. Such information is used by pathologists today in determining the degree of pulmonary fibrosis that has occurred. In most normal lungs, the two most common collagen types, type I and type III, are observed at a ratio of approximately 2:1. When pulmonary fibrosis occurs, there is generally an increase in type I collagen in relation to type III collagen. Mechanistically, the presence of the fibers causes macrophages to release lymphokines and various growth factors, which leads to an increase in the production of certain collagen types. Since type III



is considered to be more compliant than type I, this might be the cause of the “ stiffening” of the lung tissue, but this is not known for certain. Emphysema Whenever inhaled toxins result in the progressive destruction of the alveolar walls of the lung tissue, there is an enlargement of the lung air spaces accompanied by a decrease in the surface area of the lung available for gas exchange. This is commonly referred to as emphysema, and it is a relatively common pulmonary disease condition in the United States. Although emphysema is due primarily to tobacco smoke inhalation, a number of inhaled industrial toxins may also be responsible for the development of emphysematic conditions. For instance, the inhalation of coal dust by miners over extended periods has been shown to result in both pulmonary fibrosis and emphysema. Recent research has indicated that a genetically related deficiency in α-1-antiprotease, of a biochemical inhibitor of elastase, is clinically related to the relatively early onset of emphysema. It is believed that the breakdown of the alveolar walls is modulated by elastases, which are released by neutrophils and perhaps alveolar macrophages, and if the α-1-antiprotease enzyme is genetically absent or decreased, this results in a higher incidence of emphysema. In this scenario, if an inhaled toxin causes increased migration of the normally protective cells (neutrophils and macrophages) to the site of the inhaled toxin deposition, then these cells may end up damaging the lung tissue in addition to eliminating the toxins. Pulmonary Edema Many inhaled agents produce sufficient cellular toxicity to cause an increase in the membrane permeability of the alveocapillary membrane complex of the lung and other airway linings. This results in an increase in fluid, either in the interstitial space of the alveocapillary membrane complex or on the surface of the airways or alveolar sacs. This increase in fluid is called edema, and its presence impedes the exchange of oxygen and carbon dioxide between the alveolar air and the pulmonary blood. If the decrease in gas exchange proceeds sufficiently, the affected individual can die, literally in their own fluids. Among the many agents that result in pulmonary edema are the air pollutant gases, such as nitrogen dioxide and ozone. These agents typically exert their lung toxicity at relatively low levels of exposure in air-pollution episodes, but in industrial exposures, workers may be exposed to considerably higher concentrations. Chlorine and phosgene, two of the more potent inducers of pulmonary edema, were shown to induce thousands of deaths when used as chemical warfare gases in World War I. Recently, it was reported that the Iraqi military has used one or both of these agents against the Kurdish minority in that country. Since chlorine is now the primary chemical used to keep water supplies clean, its industrial use has soared. Municipalities use chlorine for their drinking water treatment; therefore, its geographic distribution is widespread. Large-scale releases of chlorine have occurred during transport to these disparate localities, and there have been a number of fatalities from pulmonary edema following chlorine inhalation. Phosgene is also used frequently in industry; however, strict industrial hygiene controls, due to the extreme toxicity of the chemical, has resulted in a low frequency of worker injury. Other agents known to cause pulmonary edema include nickel oxide, paraquat, cadmium oxide, and some industrial solvents. The delayed onset of pulmonary edema in most cases of chemical inhalation results in a significant hazard for exposed workers. Usually, the edema fluid is not readily detected by the exposed individual or by clinical examination for at least several hours after the termination of exposure. In a typical occupational exposure, the worker may experience short-term symptoms involving irritation of the airway, which influences them to seek immediate medical assistance. Since the short-term symptoms usually have no immediate cytotoxic sequelae, the medical examination will result in no revelation of significant morbidity, and the patient will be released. Then, 4–24 h later, the pulmonary edema rapidly develops, usually while the patient is asleep. Often, when patients awake



with difficulty breathing, they are already in an advanced stage of pulmonary decline, and the condition is difficult to treat. It is critical that individuals who have been exposed (or potentially exposed) to agents known to cause pulmonary edema, be kept overnight (or at least 24 h following the exposure) at a medical facility where they can be closely monitored. A series of chest X rays during the “ critical period,” when pulmonary edema could be initiated, should be taken and examined for the appearance of fluid in the lung.

Respiratory Allergic Responses Among the potential allergic reactions of the respiratory system in industrial exposures, there are many well-characterized conditions, as well as somewhat mysterious and hard-to-define personnel histories. Many of the characterized diseases have historically involved certain occupations and are often named after the occupations in which they were first observed. The allergic reactions involve antibody formation against certain inhaled toxins or to dusts and organic particles. Subsequent exposure to the same agent then often results in a more severe reaction, which is understandably a real problem in the workplace where individuals often work in the same environment and receive repeated exposures. In the less characterized occurrences, it often appears that exposure to one agent might result in a nonspecific reaction to a multitude of other compounds inhaled at some later time.

Occupation-Related Inhaled Allergic Disorders A very old disease, known as “ farmers’ lung,” involves the allergic reaction to the Actinomycetes spores found in hay. Hay that is collected in the field is often damp, and the high temperatures that can arise inside damp hay over time may give rise to large numbers of the thermophilic Actinomycetes spores. When the farmers inhale these spores, IgG antibodies are produced (against the spores), and subsequent exposures result in potentially severe allergic reactions. An interesting aspect of the disease is that the time interval between the initial exposure and the expressed toxicity can be highly variable. Various aches and pains, fever, chills, cough, weight loss, and malaise accompany the condition, which is often confused with pneumonia. Over the long term, fibrosis can also materialize. “ Malt worker’s lung,” contracted from the dust of bird droppings, presents with similar allergic alveolitis and has been reported in individuals in the whiskey industry. “ Cheese washer’s lung” has been reported in the widespread cheese industry. Ironically, this condition is due to Penicillium spores. In the lumber industry, “ maple bark stripper’s disease” results from the inhalation of fungus particles, particularly Cryptostroma. Bagassosis results from the inhalation of the bagasse dust left behind after the moisture has been removed from sugar cane stalks. Once the disease is in progress, the worker must be removed from any further contact with the bagasse dust, or the symptoms are likely to return and will usually get progressively worse. In the textile industry, the inhalation of cotton dust and other organic fibers has long been associated with reactive airway disturbances known as byssinosis. Individuals with this condition complain of chest tightness, wheezing, and other respiratory difficulties. It should be noted that these symptoms might appear after a short, or even an extended, absence from the industrial setting. A particular pattern seems to be that the first day back at work after a break, such as a weekend, is the most likely time for an episode. Unlike the previously cited occupational diseases, byssinosis does not appear to be necessarily related to the presence of bacteria, fungus, or some other living organism; the cotton or textile dust is the only requirement. Bronchoconstriction results from the release of histamine and 5-hydroxytryptamine following inhalation of the cotton dust. If the affected workers are removed from the environment containing the offending dusts relatively early in the process (i.e., months or very few years), then the patients appear to recover without permanent lung decrements. Long-term development of the disease, however, has been shown to result in permanent injury. In addition, the symptoms associated with byssinosis are usually more severe in smokers than in nonsmokers.



Industrially Related or Occupational Asthma Many individuals develop asthma following workplace exposure, and some asthmatics suffer additional provocation following the inhalation of certain industrial toxins. The inhalation of wood dusts, for instance, has been implicated in both situations. Some grocery workers have developed an asthmatic condition following the wrapping of meats with plastic film. Apparently, heating the plastic to seal it releases toluene diisocyanate, which is then inhaled. Subsequent exposure to even very low levels of the plastic, or its component, may result in a severe reoccurrence of symptoms. It has been shown that the bronchiolar muscles of asthmatics will undergo constriction at a lower concentration of inhaled industrial chemicals than will those of nonasthmatics. Not surprisingly, these individuals often find themselves reacting in situations in which their co-workers do not respond. A further complication for these workers is that exercise tends to exacerbate the asthma symptoms. Physical exertion, obviously required in many industrial situations, along with the simultaneously chemical exposure can lead to severe complications for the affected worker.

Lung Cancer Until the twentieth century, lung cancer was relatively rare. The rapid promotion of lung cancer to the number one cancer killer is directly related to the inhalation of tobacco smoke (probably 80–90 percent of all lung cancers) and industrial/atmospheric chemicals. The relationship between tobacco smoke inhalation and lung cancer was discussed previously. Many industrial chemicals have also been linked to lung cancer in workers and laboratory animal studies. The dusts and fumes of many metals have been demonstrated to be carcinogenic in lung tissue. Epidemiologic studies conducted on worker populations in smelting operations have long shown definitive relationships between metal inhalation and lung cancer. Industrial metal carcinogens include nickel, arsenic, cadmium, chromium, and beryllium. Workers in mining operations, including metal recovery from ores, are at risk for developing lung cancers because of exposure to certain metals such as chromium and uranium. The inhalation of benzo(a)pyrene and other polycyclic aromatic hydrocarbons, from coke oven emissions, has also been linked to the development of lung cancer. Radioactive materials have long been recognized as inducers of lung cancer. Uranium miners have an elevated incidence of lung cancers, as did the victims of the atomic bomb explosions at Hiroshima and Nagasaki. Recently, the potential for inhalation of radon gas has become a concern, due to the large population with the possibility for long-term exposure. Smoking has been shown to exacerbate the incidence of lung cancer when in conjunction with exposure to radioactive materials. An important feature regarding the development of lung cancer in humans is the generally long latent period. Normally it takes 20–40 years following the inhalation of most toxins before lung tumors appear. For this reason, it is often difficult to establish the definitive etiology of the lung cancer. Cancer of the upper respiratory tract does occur and is associated with some professions, such as chromate and nickel industry workers. By far, though, the majority of respiratory system cancers occur in the bronchioles and the lung tissues.

9.3 SUMMARY The lungs provide a unique pathway for industrial toxins and tobacco smoke to enter the body, since the interface between the alveolar air and the pulmonary blood can facilitate the diffusion of both life-giving air and life-threatening toxins. The beautiful design of the respiratory system provides a number of highly efficient methods of protection from commonly encountered potential toxins, including

• Humidification and temperature control • The mucociliary escalator



• Alveolar macrophages Many industrial toxins are encountered as particulates, which undergo characteristic deposition in certain regions of the respiratory system according to various physicochemical processes. The speed and mechanism by which particulates are cleared from the various respiratory regions vary significantly. Industrial chemicals that are inhaled as gases and vapors are often taken up very rapidly, and the effects in workers can be substantial, both in the lung and at distant sites. Inhaled industrial toxins exert toxicity by several distinct physiological mechanisms, which have historically led to many deleterious disease states in workers. Specific mechanisms of respiratoryrelated toxicity include

• • • • •

Irritation of respiratory airways Fibrosis and pneumoconiosis Pulmonary edema Respiratory allergic responses Lung cancer

Some inhaled agents exert toxic effects by more than one mechanism, and many workers may suffer from more than one lung-related disease condition. Potential interactions between different inhaled toxins, especially tobacco smoke and various industrial chemicals, pose an additional threat. There is a tremendous potential for inhalation exposure to toxic chemicals in the workplace; therefore, workers must be monitored thoroughly by vigorous programs in industrial hygiene, environmental monitoring, occupational physicals, and toxicology.

REFERENCES AND SUGGESTED READING Church, D. F., and W. A. Pryor, “ The oxidative stress placed on the lung by cigarette smoke,” in The Lung, Vol II, R. G. Crystal, J. B. West, P. J. Barres, et al., eds., Raven Press, New York, 1991, pp. 1975–1979. Dosman, J. A., and D. J. Cotton, eds., Occupational Pulmonary Disease. Focus on Grain Dust and Health, Academic Press, New York, 1980. Duffell, G. M., “ Pulmonotoxicity: Toxic effects in the lung,” in Industrial Toxicology, 1st ed., P. L. Williams, and J. L. Burson, eds., Van Nostrand-Reinhold, New York, 1985. Ebert, R. V., and M. J. Terracio, “ The bronchiolar epithelium in cigarette smokers,” Am. Rev. Resp. Disease 111: 6 (1975). Fenn, W. O., and H. Rahn, Handbook of Physiology, American Physiology Society, Washington, D.C., 1964. Frazier, C. A., ed., Occupational Asthma, Van Nostrand-Reinhold, New York, 1980. Guyton, A. V., Textbook of Medical Physiology, 8th ed. Saunders, Philadelphia, 1991. Hahn, F. F., “ Carcinogenic responses of the lung to inhaled materials,” in Concepts in Inhalation Toxicology, R. O. McClellan, R. F. Henderson, eds., Hemisphere, New York, 1989, pp. 313–346. Hatch, T., and P. Gross, Pulmonary Deposition and Retention of Inhaled Aerosols, Academic Press, New York, 1964. Lippmann, M., “ Biophysical factors affecting fiber toxicity,” in Fiber Toxicology, D. B. Wahrheit, ed., Academic Press, San Diego, 1993, pp. 259–303. Mauderly, J. L., “ Effects of Inhaled Toxicants on Pulmonary Function,” in Concepts in Inhalation Toxicology, R. O. McClellan, and R. F. Henderson, eds., Hemisphere, New York, 1989, pp. 347–402. McClellan, R. O., and R. F. Henderson, eds., Concepts in Inhalation Toxicology, Hemisphere, New York, 1989. Menzel, D. B., and M. O. Amdur, “ Toxic responses of the respiratory system,” in Doull’s Toxicology: The Basic Science of Poisons, 3rd ed., Macmillan, New York, 1986. Morgan, W. K. C., and A. Seaton, eds., Occupational Lung Diseases. Saunders, Philadelphia, 1975. Morrow, P. E., “ Dust overloading in the lungs: Update and appraisal,” Toxicol. Appl. Pharmacol. 113: 1–12 (1992).



Muir, D., ed., Clinical Aspects of Inhaled Particles, Davis, Philadelphia, 1972. Parent, R. A., Treatise on Pulmonary Toxicology, Vol. I, Comparative Biology of the Normal Lung. CRC Press, Boca Raton, FL, 1991. Parkes, W. R., Occupational Lung Disorders, 2nd ed., Butterworths, Woburn, MA, 1982. Samet, J. M., “ Epidemiology of lung cancer,” in Lung Biology in Health and Disease, C. Lenfant, ed., Marcel Dekker, New York, 1994. Shami, S. G., and M. J. Evans, “ Kinetics of pulmonary cells,” in Comparative Biology of the Normal Lung, Vol. 1. Treatise on Pulmonary Toxicology, R. A. Parent, ed., CRC Press, Boca Raton, FL, 1991, pp. 145–155. Steele, R, “ The pathology of silicosis,” in Medicine in the Mining Industries, J. M. Rogan, ed., Davis, Philadelphia, 1972. Tager, I. B., S. T. Weiss, A. Muñoz, B. Rosener, and F. E. Speizer, “ Longitudinal study of the effects of maternal smoking on pulmonary function in children,” NEJM, 309: 699–703 (1983). USEPA, Respiratory Health Effects of Passive Smoking: Lung Cancer and Other Disorders, USEPA/600/6-90/006, 1992. Witschi, H. R., and J. A. Last, “ Toxic responses of the respiratory system,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 5th ed., C. D. Klaassen, ed., McGraw-Hill, New York, 1996.

10 Immunotoxicity: Toxic Effects on the Immune System IMMUNOTOXICITY: TOXIC EFFECTS ON THE IMMUNE SYSTEM


This chapter discusses

• • • • • •

Basic elements and functioning of the immune system Types of immune reactions and disorders Clinical tests to detect immunotoxicity Tests to detect immunotoxicity in animal models Specific chemicals that adversely affect the immune system Multiple chemical sensitivity

10.1 OVERVIEW OF IMMUNOTOXICITY Exposure to a variety of chemicals and biological agents has been implicated in the onset of symptoms of immune origin, including acute and chronic respiratory distress, dermal reactions, and manifestations of autoimmune disease. The types of substances associated with immune system effects is extraordinarily diverse, and include chemicals found in occupational and environmental settings, infectious materials, certain foods and dietary supplements, and therapeutic agents. As discussed in this chapter, dysregulation of the immune system by toxicants can lead directly to adverse health effects, as well as rendering the body more susceptible to infectious disease and cancer. The immune system is highly complex, with many facets poorly understood. Because of this, assessment of potential immunotoxic effects of drugs, chemicals, and other agents is not a simple task. Often, measurement of a variety of components of the immune system and/or their functionality is required to gain an appreciation of the likelihood of immune dysfunction from drug or chemical exposure. Increasingly, there is realization that the immune system may be among the most sensitive target organs for toxicity for many chemicals and, as a result, merits special attention.

10.2 BIOLOGY OF THE IMMUNE RESPONSE The immune system has evolved primarily to defend the body against the invasion of microorganisms, although normal immune function is important in regulating and sustaining the internal environment as well, such as recognition and removal of malignant cells. There are two types of immunity: natural immunity (also termed innate immunity) and acquired immunity (also termed specific immunity). Natural immunity is nonspecific in that it is directed to a wide variety of foreign substances, and is rarely enhanced by prior exposure to these substances. Natural immunity arises from several mechanisms, including complement, natural-killer (NK) cells, mucosal barriers, and the unique activity of Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




polymorphonuclear and mononuclear phagocytic cells. Parts of this nonspecific immune system may contribute to the pathogenesis of an inflammatory response, and certain aspects of this system may be important in the etiology of autoimmunity. Acquired immunity, in contrast, is highly specific and increases in magnitude with successive exposure to foreign substances. Substances that trigger these specific immune responses are termed immunogens, and may be either foreign or endogenous. In many cases, immunogens are proteins, although a variety of macromolecules can be immunogenic under appropriate circumstances, including polysaccharides, nucleic acids, and ribonucleic acids. There are two types of acquired immune responses: humoral immunity and cell-mediated immunity. Humoral immunity involves the production of proteins capable of binding to foreign substances. These belong to a special class of proteins called immunoglobulins, and the proteins themselves are called antibodies. The substances to which the antibodies bind are called antigens. Antibody binding can neutralize toxins, cause agglutination of bacteria and other microorganisms, and lead to precipitation of soluble foreign proteins. Each of these is important in defense of the host. In cell-mediated immunity, specialized cells rather than antibodies are responsible for the destruction of foreign cells. A critical function of the immune system is to effectively distinguish between macromolecules that belong, or do not belong, in the body. The specific immune response is believed to be highly individualistic, a process which defines “ self” while also defending the organism against “ nonself.” This is evident by the response to certain environmental toxicants, to allergens or antigens, and the specific rejection of allografts. Recognition of “ self” is known to be guided, in part, by genetic variations in proteins of the class I and II major histocompatibility complex (MHC). Initially, the ability of the immune system to differentiate “ self” from “ nonself” is an educational process. During maturation, the system must ignore an infinite variety of self-molecules and yet be primed and ready to respond to an array of exogenous antigens. Immunomodulatory control mechanisms lead to immune tolerance of self and carefully orchestrate the immune response to targets and removal of foreign macromolecules and cells. These control mechanisms arise from interactions among the several different cell types with roles in proper immune function. Lymphocytes are considered to be the major cells involved in a specific immune response in humans. They are derived from pluripotent stem cells and undergo an orderly differentiation and maturation process to become T cells or B cells (see Figure 4.1 in the chapter on hematotoxicity), with critical functional roles in the host defense. T-cell development occurs primarily in the thymus, where cell surface protein markers are acquired during the selection and differentiation process. These protein markers are called CD antigens (for cluster of differentiation), and at least 78 different CD antigens have been identified in humans. The presence of certain CD antigens, detectable by immunofluorescence, has been used to positively identify immunocytes. In general, mature T cells are characterized by the presence of CD3+ and CD4+ or CD8+ surface markers and are devoid of surface or cytoplasmic immunoglobulin. There are various subtypes of T cells, such as T-helper (TH) cells, T-suppressor (TS) cells, and cytotoxic cells (TC). TH lymphocytes carry the CD4+ marker, while TS and TC lymphocytes have the CD8+ marker. Together, these T-lymphocyte populations play a vital role in initiating and regulating the immune response. Human B cells develop from stem cells in the fetal liver and, after birth, B-cell development occurs principally in the bone marrow. B-cell development and maturation are characterized by class-specific immunoglobulin (Ig) expression on the cell surface. Monoclonal reagents can identify the Ig expressed on the surface of B cells. Immunophenotypic characterization of cells via these markers has proved to be invaluable in certain clinical situations and a useful research tool. B cells play an important role in recognition of antigens and are responsible for antibody production. Another important cell in the specific immune response is the antigen-presenting cell (APC). These cells make first contact with the antigen and may also process the antigen; that is, modify it in such a way as to enable its recognition by T cells. This category of cells is defined more by function than cell type. In general, the most important APCs are tissue macrophages and peripheral blood monocytes, although cells of other types (e.g., Langerhans cells in the skin, dendritic cells in lymphoid tissue) may also perform this function.



In the specific immune response, antigen may be taken up by APCs and presented to T or B cells. In order to present the antigen to T cells, the antigen must be processed, or partially digested by the APC and then presented on its cell surface bound to an MHC class II molecule. Presentation of antigen to B cells does not require this processing, and in fact B cells are capable of recognizing antigens directly, without APC presentation. Antigens, either presented by APCs or encountered independently, interact with immunoglobulins on the cell surface of B-cell clones. Different B-cell clones vary in the immunoglobulins expressed on their cell surface, and these immunoglobulins can be quite specific in terms of the antigens with which they will interact. Thus, a particular antigen may interact with only one or a few B cell clones, a critical aspect in creating a specific immune response. When the antigen binds to an immunoglobulin receptor on the B cell surface, the antigen–receptor complex migrates to one pole of the cell and is internalized within the cell. The B cell becomes activated, and the antigen is processed leading to display of antigenic peptides on the cell surface in conjunction with an MHC class II protein. T-cell activation is postulated to require at least two signals. The first signal is thought to be an interaction between the CD4+ T-cell receptor of T-helper (TH) lymphocytes and antigenic peptides and MHC class II proteins presented by APCs or B cells. The second signal may be under the influence of other receptor–ligand pairs on the T cell and cognate interactions through adhesion molecules of APCs, MHC complex, and the various cytokines produced by T-cell subsets and accessory cells, such as macrophages. When activated, TH cells proliferate, creating more cells for interaction with APC and B cells. An effective immune response requires the activation of specific subsets of TH cells (TH1 and TH2 cells) which secrete different cytokines. Cytokines are low-molecular-weight proteins that mediate communication between cell populations. A list and functional classification of cytokines is shown in Table 10.1. The TH1 cells are involved in the activation of macrophages by INF-γ, secrete tumor necrosis factor (TNF), and mediate delayed-type hypersensitivity responses. The most critical function of TH2 cells is to regulate B cells, but they also secrete cytokines (specifically, interleukins, designated IL) that may regulate mast cells (IL3, IL4, and IL10), eosinophils (IL5) and IgE (IL4) responses in allergic diseases. Of the several factors known to participate in immunomodulation, IL4 and IL10 are particularly noted to upregulate the humoral response while suppressing the cell-mediated response (see below for more discussion of humoral versus cell-mediated immunity). IL13, which is produced by activation of T cells (Table 10.1) and shares many of the properties of IL4, also suppresses cell-mediated immune responses and the production of proinflammatory cytokines (IL1, IL6, IL8, IL10, IL12, and TNF). When an activated TH cell binds to the antigenic peptide-MHC complex of a B cell, the B cell is stimulated to replicate and differentiate into an antibody secreting plasma cell. This B cell clonal expansion leads to increased production of antibody specific to that B cell, and this antibody, in turn, has reactivity directed rather specifically to the antigen initiating the response. Through this mechanism, the immune system is able to produce the necessary quantities of antibodies targeting specific molecules (antigens) regarded as foreign. The synthesis of the antibody is tightly regulated, however, and the proliferation of plasma cells and antibody synthesis are controlled by cytokines and interactions with T cells. T-amplifier cells (TA) and T-suppressor cells (TS), as their names imply, function to enhance or suppress the immune response, respectively. Control of the immune response is achieved by balancing the stimulatory and inhibitory effects of T cells and various cytokines. After an encounter with an antigen, the immune system appears to retain “ memory” of that antigen and is able to mount a more rapid and greater antibody response on subsequent contact, even if the period between exposures to the antigen span several years. The basis for this memory is still not well understood. Initial (primary) immune responses to T-dependent antigens require a proliferative response by naive T and B cells. As these cells mature, they differentiate and become effector cells. The elimination of effector T cells and the factors controlling the survival of memory cells is still controversial. Because immune responses to viruses or immunization encountered in childhood generally result in lifelong immunity, it has been presumed that memory



TABLE 10.1 Cytokines and their Functions Cytokine IL1 (IL1-α and IL1-β) IL2 IL3

IL4 IL5 IL6 IL7 IL8 IL9 IL10 TNF-α TGF-β TNF-β Interferons

Produced by


Several cell types, including neutrophils and macrophages T cells T cells

Variety of effects, including neutrophil and macrophage activation, T- and B-cell chemotaxis, and increased IL2 and IL6 production Stimulates replication of T cells, NK cells, and B cells Involved in regulation of progenitor cells for several different cell types, including granulocytes, macrophages, T cells, and B cells Activated T cells Activates T and B cells; suppresses synthesis of IL1 and TNF T cells and activated B cells Increases secretion of immune globulins by B cells Several cell types, including T Important in inflammatory reactions and in differentiation and B cells of B cells into Ig-secreting cells Bone marrow stromal cells Important in regulating lymphocyte growth and differentiation Activated monocytes and Activates neutrophils; important for chemotaxis of macrophages neutrophils and lymphocytes TH cells Stimulates growth of TH cells B cells Stimulates growth of T cells in the presence of IL2 and IL4 Variety of cells, primarily Important in inflammatory responses; effects similar to IL1 activated macrophages Variety of cells Inhibits T-cell proliferation and suppresses inflammatory responses Important in mediating cytotoxic immune responses, cell Activated CD4+ cells (TH) lysis Leukocytes (INF-α), (INF): Neoplastic growth inhibitor; activates macrophages; fibroblasts (INF-β), and protects against viral infections by interfering with viral lymphocytes and NK cells protein synthesis (INF-γ)

is afforded by long-lived cells that become activated only following repeat exposure to the antigen or immunization. While it has been assumed that “ memory cells” last indefinitely following a single antigen contact, recent evidence suggests the life-span of memory cells may be related to repeat contact with antigen. In order to be recognized by the immune system, antigens must be of appreciable size. Some of the smallest antigens, for example, are natural substances with molecular weights in the low thousands. There are circumstances where much smaller molecules can elicit an immune response, but this requires the participation of a large molecule to serve as a carrier. For example, some metals, drugs, and organic environmental and occupational chemicals too small to be recognized by the immune system can become antigenic when bound to a macromolecule such as a protein. Once the immune response has been initiated, antibodies will recognize and bind the small molecule even when it is not bound to the carrier molecule. In situations such as this, the small molecule is called a hapten. The antibodies themselves are glycoproteins, the basic unit of which consists of two pairs of peptide chains (see Figure 10.1) connected by disulfide bonds. The longer peptide chain is termed the heavy (or H) chain and the shorter is the light (or L) chain. There are five main types of antibodies, or immunoglobulins (Ig): IgG, IgM, IgA, IgE, and IgD. They differ both in structure and function. IgG is present in the greatest concentration in serum, has a molecular weight of around 150,000 (there are four subtypes of somewhat different sizes), and is important in secondary immune responses. IgM is a primary response antibody, meaning that it is increased



Figure 10.1 Light and Heavy Chain Structure of IgG. IgG illustrates the basic structure of antibody proteins, which consists of two long, heavy chains and two shorter, light chains held together by disulfide bonds. Composition of C domains is relatively constant, while V domain varies, creating the binding specificity characteristic of antibodies.

very early in an immune response. IgM is much larger than the other Igs, consisting of five sets of heavy/light-chain pairs bound together at a single point with another peptide (the J chain). Its molecular weight is about 970,000. IgA may exist as a monomer (one basic unit of two pairs of H and L chains) or as a dimer—two basic units bound together with a J chain. The monomeric IgA has a molecular weight of about 160,000 and is the predominant form of IgA found in serum. IgA is the primary Ig found in secretions (e.g., tears and saliva), mostly in the dimeric form with a molecular weight of 385,000. IgD has a molecular weight of about 184,000, and is present in very low concentrations in serum. Its function is unclear, but it may play a role in B-cell differentiation. IgE is slightly larger than IgG (molecular weight of 188,000), and is normally present in low concentrations in serum. It can attach itself to leukocytes and mast cells, and is the primary antibody involved in hypersensitivity reactions. In cell-mediated immunity, cells carrying the antigen on their surface are attacked directly by cytotoxic T cells (TC) or other cell types such as natural-killer (NK) cells. In the case of TC cells, recognition of cells to be destroyed is through interaction between processed antigen in conjunction with MHC class I molecules on the target cell surface and an antigen receptor on the TC. In order to be active, the TC must also receive stimulation from CD4+ cells, principally in the form of IL2. Mechanisms of target cell recognition by NK cells are not well understood.



10.3 TYPES OF IMMUNE REACTIONS AND DISORDERS Interactions of toxicants with the immune system may result in undesirable effects of three principal types—those manifested as (1) a hypersensitivity reaction, (2) immunosuppression, or (3) autoimmunity. Each is discussed below. Allergic Reactions Allergic reactions are divided into four classes: Type I. Type I immune response is limited to IgE-mediated hypersensitivity (allergic) reaction. This reaction involves an initial exposure in which immune symptoms are generally absent (sensitization), followed by reexposure that can elicit a strong allergic reaction. In type I immune responses, antigen interacts with IgE antibodies passively bound to mast cells. On binding of antigen to the IgE, the mast cells release histamine and serotonin, which are responsible for many of the immediate symptoms of an allergic reaction such as upper respiratory tract congestion and hives. In a severe reaction, termed anaphylaxis, histamine and serotonin release can cause vasodilation leading to vasomotor collapse, and bronchiolar constriction making breathing difficult. This type of reaction has occurred following the administration of a number of different drugs and diagnostic agents, hormones, and a variety of sulfiting agents (e.g., sodium bisulfite, sodium metabisulfite, etc.). Type II. Type II reaction is believed to be the result of the binding of a drug or chemical to a cell surface, followed by a specific antibody-mediated cytotoxicity that is directed at the agent (drug or chemical) or at the cell membrane that has been altered by the compound. Under some circumstances, immune complexes may become adsorbed to a cell surface (erythrocytes, thrombocytes or granulocytes) resulting in a complement-mediated cytotoxic response, leading to induction of immune hemolytic anemia, thrombocytopenia or granulocytopenia. Type III. Soluble immune complexes consisting of a drug or chemical hapten (plus carrier molecule) and its specific antibody plus complement components are primarily responsible for immune complex disease. A particular form of immune complex disease arising from injection of an antigen is called serum sickness syndrome. Clinically, a type III reaction may be characterized by the onset of fever and the occurrence of a rash that may include purpura and/or urticaria. The immunopathology includes the activation of complement and the deposition of immune complexes in areas such as blood vessel walls, joints, and renal glomeruli. Some of the signs and symptoms associated with drug-related lupus may be included under type III reactions. Type IV. These reactions involve cell-mediated and/or delayed-type hypersensitivity responses. The expression of type IV reactions requires prior exposure to the agent and T-cell sensitization. A special subpopulation of T cells (TD) appear to be responsible for this reaction. The TD cells react with antigens in tissues and release lymphokines, attracting macrophages to the site and leading to an inflammatory response. The reaction is termed delayed because the inflammatory reaction may not peak for 24–48 h, as opposed to responses occurring within a few minutes to a few hours with other reaction types. These reactions are usually seen as contact dermatitis occurring after the use of certain drugs or exposure to some chemicals. Immunosuppression Impairment of one or more components of the immune system from drug or chemical exposure can lead to loss of immune function, or immunosuppression. Clinically, this is manifested primarily as increased susceptibility to infectious disease, although diminished immune function could conceivably increase vulnerability to cancer by impairing immune surveillance and removal of malignant cells. In



certain situations immunosuppression is intentionally induced via drug therapy to prevent rejection of transplants. Agents employed for this purpose are diverse, and several potential mechanisms are involved, including inhibition of cytokine production (e.g., corticosteroids, cyclosporin) and lymphocyte proliferation (e.g., azothioprine). Most of the evidence that environmental and occupational chemicals suppress immune responses is derived from animal studies, and while the same principles likely apply to humans as well, there are few clear examples in the clinical literature of immunosuppression from chemical exposure other than that from intentional treatment with immunosuppressive drugs. The opposite reaction, immunological enhancement, is also possible, and several natural and synthetic agents have been shown to increase immune responsiveness under experimental conditions. Examples of agents that increase immune reactivity include the bacillus Calmette-Guerin (BCG), alum (aluminum potassium sulfate or aluminum hydroxide), bacterial lipopolysaccharides and peptidoglycans, a variety of synthetic polymers, and the antiparasitic drug Levamisole (phenylimidazolethioazole). Difficulty in producing a controlled stimulation of the immune system and the enormous potential for undesirable side effects limit the therapeutic use of these agents. To date, there are no examples of environmental or occupational chemicals shown to produce immune stimulation in humans, other than in the context of allergic reactions.

Autoimmunity Autoimmunity is defined as the induction and expression of antibodies to self-tissue, including nuclear macromolecules. Studies of drug-related autoimmunity in humans have provided some of the best examples of this type of reaction. Although there are many types of autoimmune disease, the most common autoimmune syndrome produced by drugs is one resembling systemic lupus erythematosus (SLE). Clinical signs and symptoms of so-called drug lupus are not identical to idiopathic SLE, however. Both are characterized by arthralgia and the appearance of antinuclear antibodies in the blood, but the pattern of antinuclear antibodies is somewhat different, and renal and CNS complications dominate idiopathic SLE while these are typically absent in drug-lupus. Symptoms of drug lupus generally subside after the drug is withdrawn. Demonstration of autoimmune responses from environmental exposure to chemicals (other than drugs) has been difficult, in part because of problems identifying etiologic agents in retrospective studies of patients developing autoimmune disease. One concern is that some chemicals may exacerbate underlying autoimmune disease (e.g., SLE), rendering symptomatic a patient with subclinical disease or increasing the duration or severity of symptoms in those with active disease. Unfortunately, differentiating the effects of chemical exposure from progression of the underlying disease is difficult or impossible in practice. Understanding of autoimmune consequences of chemical exposure is further hampered by the general lack of satisfactory animal models—the results obtained in laboratory animals seldom correspond exactly to observations in humans.

10.4 CLINICAL TESTS FOR DETECTING IMMUNOTOXICITY In the clinical setting, the use and proper interpretation of immunologic laboratory tests can be important in establishing a differential diagnosis in a patient who has been exposed to an immunotoxic agent. Immune system testing for diagnostic purposes can be challenging, however, because of the complexity of the immune system and difficulty in establishing normal values for many of the tests. When immune dysfunction from chemical exposure is suspected, it is important to be sure that the patient is free from infectious disease and not taking medications that can influence immune function—obvious confounders to interpretation of any immune tests. Also, it is important to recognize that many immune parameters, such as lymphocyte subpopulation counts, can vary normally by age and gender, making the use of appropriate controls essential for proper interpretation of results. Finally,



temporal variations in most tests are common. In order to demonstrate that an abnormality exists, it is usually advisable to repeat the test one or more times to insure that a consistent result is obtained. Some of the laboratory tests available provide information relevant to assessing humoral immunity, others are useful in evaluating cellular immunity, and some can provide insight regarding both. Described below are examples of assays commonly used in the evaluation of individuals exposed to chemicals in the environment or workplace. Immunoglobulin Concentrations The concentrations in serum of each immunoglobulin can be determined with the exception of IgD, which exists primarily on cell surfaces. Single-radial diffusion is commonly employed for most immunoglobulins, although enzyme linked immunosorbent assay (ELISA) or radioimmunoassay (RIA) is often needed to measure the low concentrations of IgE typically present. Diminished immunoglobulin concentrations, either in total or of specific classes, may suggest immunodeficiency, but is not sufficient to establish a diagnosis. Conversely, immunoglobulins within normal limits do not necessarily indicate immunocompetence. There may be defects in subtypes of immunoglobulins not quantified by the assay, and patients with normal or high values may nonetheless exhibit increased susceptibility to disease. Immunoglobulin values may be profoundly influenced by viral or bacterial infections and the presence of some drugs. T- and B-Cell Concentrations Immunotyping of T- and B-cell subsets by ethidium bromide and cytofluorometry techniques is used by many laboratories for screening studies of chemical-related injury. Concentrations of B cells, either in absolute terms or as a percentage of peripheral blood lymphocytes, can be expressed, and the distribution of B cells expressing different immunoglobulin types (IgM, IgG, IgA) can be measured. Some studies have sought to evaluate a potential immunosuppressive effect through measurement of the ratio of TH to TS lymphocytes in peripheral blood, using the CD4+ marker to indicate TH cells and the CD8+ marker for TS cells. As discussed above, these markers are not specific for TH and TS cells, however, and interpretation of a decreased CD4+ to CD8+ ratio as a loss of T help relative to T suppression is an oversimplification. A significant reduction in CD4+ cells is associated with several immunodeficient states (e.g., in patients with AIDS, undergoing radiotherapy, or chemotherapy), implying that diminished CD4+ is indicative of impaired antibody production. This assumption is not infallible, however, because there are also circumstances in which CD4+ cells may be reduced without loss of antibody production. Significant changes in absolute or relative concentrations of lymphocyte subsets may be suggestive of immunotoxic effects from chemical exposure, but are not, by themselves, reliable indicators of compromised function. Cutaneous Anergy Anergy is a generalized clinical condition of non-responsiveness to ubiquitous skin test antigens that is frequently observed in patients who are immunosuppressed. Cutaneous anergy may suggest functional impairment or abnormalities of the cellular immune system. The most cost-effective method for evaluation of cutaneous anergy is the use of a battery of attenuated, premeasured and well-standardized ubiquitous antigens that are available from commercial sources. The assessment of a person who is thought to be immunologically suppressed due to exposure to an environmental chemical can be attained within 48 h through the use of these antigens. The intradermal skin test antigens frequently used to measure cellular delayed hypersensitivity are: tetanus toxoid, diphtheria toxoid, Streptococcus (group C), old tuberculin (PPD), Candida albicans, Trichophyton mentagrophytes, and Proteus mirabilis. Measurement of specific IgG antibodies to diphtheria and tetanus toxoids in serum at 2 weeks following booster immunization is also useful in assessing the ability to form antibodies to protein antigens. In Vitro Tests Functional capabilities of lymphocytes can be evaluated by taking a blood sample and performing a variety of tests in vitro. In general, these tests involve isolating lymphocytes from a blood sample, placing them in culture, and exposing them to a stimulatory agent. The ability of the cells to proliferate in response to the stimulus and, in the case of B cells, to synthesize immunoglobulins, can be measured. For example, treatment of peripheral blood lymphocytes with pokeweed mitogen (PWM)



normally produces cellular proliferation and increased immunoglobulin synthesis. This response requires both TH and B cells, and provides an indication of the capability of these two cells to interact properly and of B cells to produce immunoglobulins. Lipopolysaccharide (LPS) is a mitogen effective selectively on B cells, while phytohemaglutinin (PHA) and concanavalin A (con A) are selective T-cell mitogens. Other stimulants to lymphocyte activation can be used, such as tetanus toxoid, diptheria toxoid, Candida, and PPD, if the subject has been previously exposed to these. The rapid cell division characteristic of a normal response to these mitogens is typically assessed by measuring incorporation of 3H-thymidine into DNA of the cells. Other endpoints of stimulation, such as increased expression of IL2 receptors on T cells, can also be evaluated. The results of these tests are particularly prone to variability, and the tests should be repeated on several occasions in order to demonstrate an abnormal response. In the mixed-lymphocyte reaction (MLR) test, lymphocytes from the test subject and another individual are mixed. Normally, contact with the allogenic lymphocytes will cause the test subject’s lymphocytes to become activated and proliferate. To conduct this assay, the target lymphocytes are rendered incapable of replication, often by irradiation or by treatment with mitomycin C. Test subject lymphocytes are then added, and the rate of their replication is evaluated by measuring incorporation of 3H-thymidine. The cytotoxic lymphocyte (CTL) assay takes the lymphocyte interactions one step further to evaluate the ability of cytotoxic T cells (TC) to destroy target cells. After incubation of the test subject and target lymphocytes, the subject TC are isolated, washed, and reincubated with target lymphocytes preloaded with 51Cr. As the target cells are destroyed, 51Cr is released into the medium and can be measured, providing an index of cytotoxic capabilities of the TC lymphocytes. Fluorescent Antinuclear Antibody Assay (FANA) The indirect immunofluorescence antinuclear antibody assay (FANA) may be the initial screening test used to show autoimmunity. However, several FANA patterns are recognized in various connective-tissue diseases and some low-titer staining patterns have also been reported in sera from persons exposed to environmental agents. The following staining patterns may be observed: 1. The diffuse (homogenous) staining pattern, which is usually associated with antibody directed to DNA-histone or histone subfractions. This staining pattern is frequently found in sera from patients receiving chronic treatment with procainamide, hydralazine, isoniazid, anticonvulsant drugs, and some environmental chemical agents. 2. A peripheral (rim) pattern, which is attributed to antibody reacting with native DNA and soluble DNA-histone complexes. This staining pattern is frequently seen in sera from patients with systemic lupus erythematosus (>95 percent). 3. Speckled FANA staining, which is usually attributed to antibodies reacting with saline-soluble antigens. These antibodies are directed to nonhistone antigens and include Sm, ribonucleoprotein, SS-A/Ro, SS-B/La, PM-1, and SCl-70. While these staining patterns frequently occur in patients with mixed connective tissue diseases, including Sjögren’s syndrome, polymyositis and progressive systemic sclerosis, they have also been found in sera from persons exposed to immunotoxic agents. 4. The nucleolar staining pattern, which has been restricted to antibodies reactive with nucleolar RNA. This pattern is associated with a particular form of systemic sclerosis (progressive systemic sclerosis).

10.5 TESTS FOR DETECTING IMMUNOTOXICITY IN ANIMAL MODELS For most chemicals, an assessment of their potential to produce immunotoxicity in humans is based on testing in animals. Many of the tests used in animal studies are the same as, or at least analogous to, those available for clinical assessment described above. However, studies in animals offer the



opportunity to evaluate directly toxic endpoints difficult or impossible to assess clinically, such as the development of immunopathology or loss of resistance to infectious disease. Currently, a tiered approach is recommended for standardized testing for immunotoxicity in animals. Tier I consists of a battery of tests intended to evaluate both humoral and cell-mediated immune system integrity. An assessment of immune system pathology is also included in tier I (see Table 10.2). If the results of tier I tests are negative, the chemical is considered not to possess significant immunotoxic potential at the dosages tested. If effects are observed in tier I tests, additional tests are conducted in tier II to better characterize the immunotoxic properties of the chemical. Tier II does not consist of a rigid battery of tests, but rather the opportunity to select more specific tests to follow up on observations made in tier I. Examples of tests that might be used in tier II are included in Table 10.2. Many of the endpoints examined in tier I are basic. Total and differential white cell counts are obtained from blood, body and specific organ weights are recorded, and tissues of particular relevance for immune function (viz., spleen, thymus, and lymph nodes) are examined histologically for evidence of injury. Humoral immunity is assessed with a plaque-forming cell (PFC) assay. In this assay, the test animal is injected with sheep red blood cells (SRBCs) as the source of antigen. Four days later the spleen is removed, and cells isolated from the spleen are cultured with intact SRBCs. B cells producing IgM directed to SRBC antigens result in lysis of the red cells, producing clear areas in the culture called plaques. The number of plaques (per spleen or per million spleen cells) provides an indication of the ability of splenic cells to synthesize and secrete antigen-specific antibodies. This, in turn, offers information regarding the ability of the immune system to mount a primary (IgM-mediated) response. Cell-mediated immunity is evaluated by measuring the responsiveness of peripheral blood T and B lymphocytes to mitogens (such as concanavalin A), and through the MLR assay. Nonspecific immunity is evaluated in tier 1 by measuring NK cell function. These tests are essentially identical to the in vitro methods described above for clinical assessment of potential immunotoxicity in humans. More detailed follow-up tests are available for tier II. For example, if disturbance in the numbers of immunocytes is suggested by tier I tests, the abundance of individual T- and B-cell types in the spleen or blood can be measured using reagents that detect specific cell surface antigens. In the assessment of humoral immunity, an abnormal primary response (IgM-mediated) to SRBCs detected in the PFC assay in tier I might lead to an evaluation of the secondary response (IgG-mediated) to SRBCs. Evidence of altered cell-mediated immunity could lead to expanded tests of T-lymphocyte cytotoxicity in tier II, commonly using tumor cells as targets. Tier II could also include an assessment of delayed-type hypersensitivity response. Evaluation of non-specific immunity may be extended in tier II to include enumeration of macrophages and tests of their function. For functional tests, macrophages are typically taken from the peritoneal or alveolar space of test animals, cultured, and examined for phagocytic activity, secretion of cytokines, and/or production of reactive oxygen or

TABLE 10.2 Tier I and Tier II Tests for Immunotoxicity Tier I

Tier II

Hematology, including CDC and differential counts Body and organ weights, including spleen, thymus, kidney, and liver Histology of lymphoid organs, including spleen, thymus, and lymph nodes Humoral immunity, assessed through IgM plaque-forming cell (PFC) response Cell-mediated immunity, assessed through T- and B-lymphocyte responses to mitogens, and the mixed-lymphocyte response (MLR) Nonspecific immunity, assessed through measurement of natural-killer (NK) cell activity Quantitation of individual T- and B-cell populations in blood and spleen Humoral immunity, assessed through IgG plaque-forming cell (PFC) response Cell-mediated immunity, assessed through cytotoxic T-cell (CTC) activity, as well as the delayed hypersensitivity (type IV) response Host resistance, assessed through challenge with pathogens or tumors



TABLE 10.3 Examples of Agents Used for Immune Challenge in Host Resistance Tests Type of Agent Virus



Tumor cells

Name Cytomegalovirus Herpes simplex virus type 2 Influenza virus Corynebacterium parvum Listeria monocytogenes Pseudomonas aeruginosa Streptococcus pneumoniae Plasmodium species Trichinella spiralis B16-F10 melanoma PYB6 fibrosarcoma

Typical Exposure Route Intraperitoneal or intratracheal administration Intraperitoneal, intravenous, or intravaginal administration Intranasal administration Injected intravenously Injected intravenously Injected intravenously Injected intravenously Intravenous or intraperitoneal injection of infected blood Intragastric administration Cells are injected intravenously Cells are injected subcutaneously

nitrogen species. The ability of macrophages in culture to phagocytize foreign materials is typically examined using light microscopy, with either biological (e.g., SRBCs or bacteria) or nonbiological materials (e.g., fluorescent beads) as targets. On activation, macrophages normally release specific cytokines (e.g., TNF-α and IL2), as well as reactive oxygen and nitrogen. Cytokine production by activated macrophages in culture can be measured by ELISA (enzyme-linked immunosorbent assay) using antibodies directed to specific cytokines, or by ELISPOT, which is capable of identifying the numbers of cells producing specific cytokines. Several techniques are available for quantitating reactive oxygen and nitrogen species. When immunosuppression (or, less commonly, immunostimulation) is suspected, one of the most direct means to test overall immune competence is through a host resistance model (also sometimes called a host susceptibility model). With this model, the ability of the animal to withstand an immune challenge is assessed with and without exposure to the drug or chemical. Immune challenge can take the form of an infectious microorganism or a syngeneic tumor. A variety of types of infectious microorganisms are used for these tests, including viruses, bacteria, yeast, fungi, and parasites. Syngeneic tumor lines are derived from the same strain and species as the test animal, requiring their recognition as tumor cells and not simply a source of foreign protein. Examples of microorganisms and tumor cell lines used for host resistance models are provided in Table 10.3. Many of these agents are human pathogens, and this type of test arguably provides the best direct evidence of the ability of a drug or chemical to produce clinically relevant immune suppression or stimulation.

10.6 SPECIFIC CHEMICALS THAT ADVERSELY AFFECT THE IMMUNE SYSTEM The number of drugs and chemicals associated with immunotoxicity in humans is extensive. As discussed in Section 10.3, immunotoxicity typically occurs as a hypersensitivity reaction, immunosuppression, or autoimmunity. Several agents commonly encountered in occupational settings are capable of producing contact, cell-mediated hypersensitivity, with common symptoms of rash, itching, scaling, and the appearance of redness and vesicles on the skin. Examples of these agents are shown in Table 10.4. The respiratory tract is also a common site of allergic symptoms from drug or chemical exposure. Inhalation of respiratory allergens can cause an immediate-type reaction (an early-phase reaction, occurring and waning rapidly) or a delayed-type reaction (sometimes called a late-phase reaction), which may appear 6–8 h later and require 12 to 24 hours to resolve. Both reactions are


IMMUNOTOXICITY: TOXIC EFFECTS ON THE IMMUNE SYSTEM TABLE 10.4 Examples of Agents that Produce Dermal Contact Sensitivity Drugs Benzocaine Thimerosal Neomycin Resins Acrylic resins Epoxy resins Formaldehyde resins Phenolic resins Other industrial chemicals Ethylenediamine Paraphenylenediamine and other dyes Antioxidants Chlorinated hydrocarbons Dinitrochlorobenzene Mercaptans

Metals Beryllium Cadmium Chromates Gold Mercury Nickel Silver Zirconium

IgE-mediated. Table 10.5 lists examples of common agents associated with respiratory allergy. Occupational asthma represents a special kind of inhalation disorder that is distinct from typical respiratory allergy. In general, a longer sensitization period is required, and symptoms may resemble an early-phase reaction, a late-phase reaction, or both. IgE may be responsible for some, but not all, of the manifestations of occupational asthma. In fact, the role of the immune system in occupational asthma may be different for asthma initiated or provoked by high-molecular-weight compounds, low-molecular-weight compounds, and irritatants. The potential for immunosuppression from occupational and environmental exposure to chemicals has been suggested by numerous in vitro studies and experiments in laboratory animals. Direct evidence for clinical immunosuppression following workplace or environmental exposures is extremely limited. However, there are many well-documented examples of the development or exacerbation of autoimmunity from chemical exposure. Most of these examples (shown in Table 10.6) are drugs, and for agents such as procainamide, up to 80 percent of patients treated chronically will develop increased levels of autoimmune antibodies. Many of these drugs produce signs and symptoms resembling systemic lupus erythematosus, while others produce autoimmune disease of the kidney, liver, thyroid, and other organs; scleroderma; or autoimmune hemolytic anemia. Evidence suggests that several environmental contaminants may also have the ability to either produce or worsen autoimmune disease, although the association with autoimmune disease is often less well substantiated.

TABLE 10.5 Examples of Agents that Product Respiratory Allergy Molds Aspergillus Cladosporum Hormodendrum Penicillium Rhizopus Pollens (various)

Dusts and Small Particulates Coffee Enzymes Flour Mites Sawdust Pet dander Cockroach proteins

10.6 SPECIFIC CHEMICALS THAT ADVERSELY AFFECT THE IMMUNE SYSTEM TABLE 10.6 Examples of Agents Associated with Autoimmune Disease Drugs Acebutalol Allopurinol Alprenolol Amiodarone Ampicillin Bleomycin Captopril Carbamazepine Cephalosporin Chlorpromazine Chlorthalidone Dapsone Diphenylhydantoin Ethosuximide Fenoprofen Iodine Isoniazid Lithium Lovastatin Mefenamic acid Methyldopa Minocycline Nitrofurantoin Penicillamine Phenylbutazone Propylthiouracil Quinidine Sulfonamides Amino acids L-Tryptophan L-Canavanine Environmental/industrial chemicals Aromatic amines Cadmium Chlordane Chorpyrifos Chromium Formaldehyde Gold Hydrazine Mercury Paraquat Pentachlorophenol Perchlorethylene Silicon (silica) Thallium Trichloroethylene Vinyl chloride




Some classes of chemicals or agents have, in particular, been associated with immunotoxic effects in humans. These are discussed briefly below. Metals Metals have been associated with various types of hypersensitivity reactions. Beryllium, nickel, chromium, cadmium, silver, and zirconium have all been found to produce contact dermatitis. Nearly 10 percent of women and 2 percent of men have sensitivity to nickel, and may develop rashes upon contact with nickel in jewelry, coins, and clothing fasteners. Sensitive individuals may also respond to chromium in tanned leather products. Metals are also associated with pulmonary hypersensitivity reactions and occupational asthma. One of the most serious of these diseases is berylliosis, a delayed hypersensitivity (type IV) reaction thought to result from beryllium acting as a hapten. Acutely, hypersensitivity to beryllium is manifested as pneumonitis and pulmonary edema. Chronically, workers exposed to beryllium develop a severe, debilitating granulomatous lung disease. Studies in experimental animals have shown that metals such as lead, mercury, nickel, and cadmium are associated with activation of CD4+ T cells or cause suppression of antibody responses and cell-mediated immunity, resulting in increased susceptibility to infection. There is some clinical and epidemiologic evidence that lead may decrease resistance to infectious disease, and the use of arsenic for medicinal purposes suggests that it, too, may have immunosuppressive effects. Arsenic was used in the early twentieth century to treat some inflammatory diseases, and currently appears to have some efficacy in treating leukemia. Also, patients treated with arsenicals were reported to have a relatively high incidence of the viral disease herpes zoster, suggesting some impairment of the immune system. A number of studies have reported increased or unusual autoantibodies in association with exposure to some metals in the workplace, suggesting potential autoimmune toxicity. For example, there is evidence of immune complex glomerulonephritis in nephrotoxicity from cadmium and mercury. Iodine and lithium have been linked to autoimmune thyroid disease, and chromium and gold have been associated with systemic lupus erythematosus-like disease. Polychlorinated Dibenzo(p)dioxins Studies in rodents have shown that perinatal exposure to 2,3,7,8-tetrachlorodibenzodioxin (TCDD) appears to affect the developing thymus, leading to a persistent suppression of cellular immunity. The depression of T-cell function from perinatal exposure appears to be greater and more persistent than when exposure occurs in adults. The potential for TCDD immunotoxicity in humans is less clear. Individuals exposed to very high TCDD doses during an industrial explosion in Seveso, Italy in 1976 have not shown demonstrable loss of immune function. Studies of individuals exposed to TCDD chronically in Times Beach, Missouri have revealed a few differences from a control population in some parameters, but overall the observations do not suggest significantly altered immunocompetence. These studies have focused on humans exposed as adults to TCDD, and it is possible that perinatal exposure to TCDD may have more profound effects, as has been observed in laboratory animals. Increased antinuclear antibodies and immune complexes have been reported in blood of dioxinexposed workers, but increases in clinical manifestations of autoimmunity have not been observed. Dusts and Particulates A number of occupations involve inhalation exposure to high-molecular-weight organic molecules or particles containing these molecules. Examples include flour and wood dust; enzymes (e.g., from B. subtilis and A. niger in the detergent industry); dusts from agricultural wastes; fungi and bacteria in moldy hay, feeds, and wood products; and dander, feces, pupae, and other residue from insect and rodent pests. These high-molecular-weight substances are capable of producing an IgE-mediated, type I allergic reaction. This reaction can manifest itself as eye and upper respiratory tract congestion, occupational asthma, and hypersensitivity pneumonitis. Acute inhalation of dusts from bacterial or



animal origin have also been shown to produce a short-term flulike illness called organic dust toxic syndrome. This is not a type I allergic reaction because no prior sensitization is required, nor are antigen-specific antibodies present during the illness. Inhalation of silica dusts both activates and damages alveolar macrophages. Activation of these macrophages can lead to pulmonary inflammation. Reported effects on lymphocyte responsiveness are somewhat conflicting, but suggest that immune function may be impaired. Pesticides Dermal and pulmonary symptoms among workers handling pesticides are not uncommon, but most of these cases appear to be due to irritant rather than hypersensitivity reactions. Studies of workers exposed to pesticides have sometimes found changes in various specific immune parameters, but there is currently little evidence that host resistance is compromised in these individuals. Isolated reports suggest an association of pesticides (i.e., paraquat) with the development of renal autoimmune disease. Also, recent studies in animals suggest that some chlorinated pesticides may accelerate the development of autoimmunity, although no studies are yet available to assess whether this occurs in humans as well. Solvents Benzene is capable of producing bone marrow hypoplasia and pancytopenia. Along with other formed elements of the blood, peripheral blood lymphocyte counts are diminished, leading to impaired immune function. Immunotoxic effects of benzene may extend beyond individuals experiencing bone marrow toxicity from benzene, as humans exposed chronically to benzene have been observed to have diminished serum immunoglobulins and immune complement. Immune abnormalities, such as alterations in serum immunoglobulin concentrations, immunocyte counts, or immunocyte ratios have been observed in workers exposed to solvents, either individually or as mixtures. The significance of these findings is unclear, however, as no deficits in host resistance or other clinical immune effects have been demonstrated. Exposure to vinyl chloride has been linked to the development of scleroderma, and there is epidemiologic evidence of an association between chronic exposure to trichloroethylene in groundwater and lupus syndromes. Miscellaneous Agents In 1981, thousands of individuals in Spain were poisoned with cooking oil adulterated with rapeseed oil containing aniline. The symptoms that developed were called toxic oil syndrome, and included pneumonitis, rash, gastrointestinal distress, and marked eosinophilia. These patients developed autoantibodies and a connective tissue disorder characterized by myalgia, neuropathy, myopathy, and cutaneous manifestations. Hundreds of poisoned patients died, attributed primarily to impairment of respiratory musculature. Acid anhydrides are used to produce a number of commercial products, including paints and epoxy coatings. On inhalation exposure, acid anhydrides can become haptens, binding to carrier proteins in the respiratory tract to elicit an immune response. After sensitization, subsequent exposure leads to asthma-like symptoms or to a reaction resembling hypersensitivity pneumonitis. Chronic exposure may lead to severe restrictive lung disease.

10.7 MULTIPLE-CHEMICAL SENSITIVITY Multiple-chemical sensitivity is a term applied to a subjective illness in individuals attributed to contact with a broad array of chemicals in the environment. Other terms for this condition include environmental illness, total allergy syndrome, chemical-induced immune dysregulation, chemical hypersen-



sitivity syndrome, and, more recently, idiopathic environmental intolerances. There is no defined symptomology for this condition; in fact, physical diagnostic and laboratory findings are typically normal. Complaints are almost always subjective, including fatigue, headache, nausea, irritability, and loss of concentration and memory. It appears almost exclusively in adults, primarily in women. Offending substances are commonly identified by odor, although symptoms can also be ascribed to substances in food, to drugs, and to electromagnetic fields. While sensitivity is thought to arise from a single initial exposure, perhaps to a single agent, it is visualized as progressing to eventually involve an expanded array of substances; hence the term multiple-chemical sensitivity. Diagnosis is made principally on the basis of history—the patient indicates intolerance to a variety of substances in the environment. The symptoms are triggered by exposure to these substances at levels generally well tolerated by the vast majority of the population. Improvement in symptoms is attributed to avoidance of these substances. Tests are sometimes performed on these patients, including a provocation–neutralization test and a panel of immunologic tests. In the provocation–neutralization test, reaction to various substances is tested by administering small doses sublingually, intracutaneously, or subcutaneously. The test agents are not necessarily those thought to be causing the patient’s illness, and the battery of agents tested can vary from practitioner to practitioner. After administration of the test substance, the patient records any symptoms that occur over the next 10 minutes. There is no standard for what constitutes a symptom in this testing. If no symptoms are recorded, the dose is increased until a positive response is obtained. Increased or diminished doses are then given until the symptom(s) abate. This becomes the “ neutralization” dose that may be recommended to the patient for subsequent treatment of the condition. Immunologic tests usually consist of quantitation of serum immunoglobulins, complement components, lymphocyte counts, autoantibodies, and immune complexes. The status of multiple-chemical sensitivity as a legitimate disease entity has been controversial. Mainstream medical organizations do not recognize it as a defined disease for several reasons: (1) there are no objective physical signs or symptoms, or clinical laboratory observations that characterize the disease; (2) there are no clearly defined diagnostic criteria—the provocation-neutralization test described above has no physiologic basis, and the diagnostic value of the immunologic tests typically performed has not been validated; (3) although several mechanisms have been proposed to explain symptoms in these patients, there are little data to support any of these, and many explanations are contrary to current understanding of immunology and toxicology; (4) objective evidence of any chemical agent as a specific cause of multiple chemical sensitivity is virtually nonexistent; and (5) there are no treatments of proven efficacy. Many theories have been proposed to explain multiple-chemical sensitivity, most involving the immune system. Some have proposed that environmental chemicals may act as allergens or haptens, and that an IgG (rather than IgE) response to these agents leads to an immune complex disease. However, the symptoms of multiple-chemical sensitivity do not resemble serum sickness, and IgG antibodies to the postulated array of triggering substances have not been demonstrated in these patients. An autoimmune mechanism has also been proposed, but patients with multiple-chemical sensitivity typically do not have demonstrably elevated autoantibody titers, nor do they have the usual clinical manifestations of any of the autoimmune diseases. A more general concept of immune dysregulation has also been advanced, commonly including the notion that T suppression has been impaired by environmental chemical exposure. Compelling evidence that T suppression (or any other specific immune abnormality) is a consistent feature of multiple chemical sensitivity is lacking, however, as is an explanation as to how several different chemical substances, at low exposure levels, could produce this effect. A high prevalence of psychiatric disorders has been observed in patients claiming to have multiple chemical sensitivity. This suggests that somatization (i.e., symptoms of psychogenic origin) may be involved, although proponents of multiple chemical sensitivity argue that this is a manifestation, rather than a cause, of the disease. Credibility for multiple-chemical sensitivity within the medical and scientific community has also been impaired somewhat by the significant percentage of individuals with this condition using it as a basis for workers’ compensation claims or other litigation.



The mainstay of treatment of multiple-chemical sensitivity involves avoidance of what are regarded as the inciting chemicals. In some cases, this can be taken to extremes, involving near isolation in specially controlled environments. Vitamins and mineral supplements are often recommended, as well as intravenous gammaglobulin, ostensibly to fortify the immune system. “ Neutralization” doses of extracts identified positively in provocation–neutralization tests are sometimes recommended to relieve or prevent symptoms. Reports of efficacy of these treatments are either anecdotal or from poorly controlled studies. Objective evidence that any of these treatments leads to improvement in the patient’s condition is generally considered to be absent.

10.8 SUMMARY A fully functioning immune system is vital for defense against pathogenic microorganisms and to prevent the emergence of cancerous cells. It is a complex system, requiring the cooperation of many types of cells. The immune system is capable of both specific and nonspecific responses to insults. Specific responses are elicited by macromolecules recognized by the body as being foreign, termed antigens. The presence of an antigen can trigger a humoral response (i.e., the production of antibodies that bind rather specifically to that molecule) or a cell-mediated response in which cells carrying the antigen on their surface are attacked by specialized immune cells (e.g., natural-killer cells or cytotoxic lymphocytes). Drugs and chemicals can produce adverse health effects by influencing the immune system in one of three ways: 1. Causing a Hypersensitivity Reaction. There are four basic types of hypersensitivity reactions (types I–IV), each with a different mechanism. Depending on the type of reaction, symptoms may be immediate or delayed, mild or severe, and involve different organs and tissues. Allergic reactions can cause considerable discomfort in the workplace, and some types (e.g., a severe type I reaction, or anaphylaxis) can be life-threatening. 2. Suppressing the Immune System. Normal function of the immune system requires participation by many components, and disruption of any of these could conceivably result in impaired capability. If impairment is sufficient, the individual is at increased risk of infection and cancer. This has been clearly demonstrated by patients on immunosuppressive therapy (e.g., transplant patients) and in animal studies involving a variety of chemicals. Although there are few clear examples of immunosuppression from occupational or environmental exposure in humans, there is no reason to expect that this effect cannot occur under these circumstances as well. 3. Causing or Exacerbating Autoimmune Disease. By producing a dysregulation of the immune system, drugs and chemicals are capable of causing the immune system to attack normal body constituents. This has been clearly demonstrated for several drugs, and a number of reports suggest that it may also occur from occupational and environmental exposures. The potential for a chemical to produce immunotoxicity can be assessed through a variety of in vivo and in vitro tests. Most of these tests focus on effects on a very specific aspect of the immune system. The immune system possesses considerable functional redundancy and extra capacity, and alterations (or “ abnormalities” ) in one or a few parameters may not necessarily result in diminished overall functional of the immune system. Consequently, the results of these tests must be interpreted carefully.

REFERENCES AND SUGGESTED READING Burleson, G. R., J. H. Dean, and A. E. Munson, Methods in Immunotoxicology, Wiley-Liss, New York (1995).



Burrell, R., D. K. Flaherty, and L. J. Sauers, Toxicology of the Immune System, A Human Approach, Van Nostrand-Reinhold, New York (1992). Farine, J-C., Animal models in autoimmune disease in immunotoxicity assessment. Toxicology 119: 29–35 (1997). Kammuller, M. E., N. Bloksma, and W. Seinen, eds., Autoimmunity and Toxicology: Immune Disregulation Induced by Drugs and Chemicals, Elsevier Science Publishing, New York (1989). National Research Council, Biologic Markers in Immunotoxicology, National Academy Press, Washington, DC (1992). Sell, S., Basic Immunology: Immune Mechanisms in Health and Disease, Elsevier Science Publishing, New York (1987). Smialowicz, R. J., and M. P. Holsapple, Experimental Immunotoxicology, CRC Press, Boca Raton, FL (1996). Vial T., B. Nicolas, and J. Descotes, Clinical immunotoxicology of pesticides, Journal of Toxicology and Environmental Health 48: 215–229 (1996). Vos, J. G., and H. Van Lovern, Experimental studies on immunosuppression: How do they predict for man? Toxicology 129: 13–26 (1998).

PART II Specific Areas of Concern

Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.

11 Reproductive Toxicology REPRODUCTIVE TOXICOLOGY


The possibility of disruptions in normal reproductive function or proper development is one of the potential health effects of chemical exposure that causes the most concern. Increases in the number of women in the workplace and the subsequent pregnancies that occur along with occupational exposure to various chemicals make the overall population especially aware of the potential for this type of toxicity. Also, our environment is frequently perceived to harbor more potential reproductive hazards due to growing awareness of the distribution and persistence of some human-made (or human-related) chemicals. How do these perceptions of increased reproductive risk correspond to currently available scientific information and, how can future scientific investigation be focused toward important issues? These are questions that can be addressed by considering how reproductive processes interact with chemical (and non-chemical) workplace or environmental exposure. Our understanding of reproductive toxicity has expanded rapidly over the last several decades, providing much more information on which to base critical evaluations of the potential for reproductive risk. This chapter will review the established toxic responses of the human male and female reproductive systems and of human fetal development. The focus will be on potential human health effects from occupational and/or environmental exposure and examples will be drawn from this area. However, there is much additional mechanistic information from experimental systems and many more chemicals for which there are experimental indications of potential reproductive toxicity. In evaluating the importance of such chemicals or mechanisms with regard to human exposure, two basic tenets of toxicology must be considered: 1) what are the characteristics of a likely exposure, and 2) what could be a relevant dose. Understanding how reproductive processes respond to chemical challenges is the key to addressing these issues. Topics to be covered include

• Male reproduction and the susceptibility of rapidly dividing germ cells • Female reproduction and the regulation of endocrine status as a potential target for toxic responses

• Fetal development—the major opportunities for toxic responses during development and the established causes of developmental defects

• Current research concerns—hot, timely topics Many of the toxicological principles pertaining to reproduction can be clearly illustrated with examples relating to the male. Also, many of the most fully characterized examples of toxic responses on human reproduction from occupational exposures are male-related. This state of knowledge may well reflect a historical bias toward interest in male-related effects due to the former predominance of men in the industrial setting. However, male reproductive physiology and function does provide certain susceptibilities to toxic agents that are useful for illustrating how toxicological responses relate to reproductive biology in general. Understanding the underlying mechanisms will help make sense out of the more complex, and less understood issues in female and developmental toxicology. As interest Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




continues to shift toward non-industrial occupational exposures and general environmental exposures, many new discoveries about the mechanisms of toxic injury to the female reproductive system and developing offspring can be expected. Most of the best described examples of reproductive toxicity rely on experimental results obtained with laboratory animals for their explanations. While inferring the actual human potential for an adverse occurrence from animal testing results must always be done critically and cautiously, there are certain factors relating to reproductive endpoints that add unique uncertainties. The relevance of the experimental dose level to potential human exposures is always an important factor for interpreting animal studies. For reproductive endpoints, not only is the dose level an issue, the duration and interval used for dosing is also critical since sequential, delicately timed progressions of physiological events are a hallmark of reproductive processes. Experimental testing for any given chemical must encompass the timeframe likely to be relevant for the mechanisms of toxicity involved. In evaluating developmental toxicity, concomitant maternal toxicity can be a problematic complication. In trying to demonstrate effects, dose levels are often pushed high enough to result in general wasting, nutritional problems, or other stresses on the animal. Thus, it can be difficult to distinguish between direct toxicity to the developing offspring and females that are simply too compromised to maintain a normal pregnancy. With such complications firmly in mind, animal testing remains a critical and valuable tool for characterizing reproductive toxicity. There is no way around the need for test systems that can be readily manipulated to tease out potential mechanisms of toxicity. The challenge is interpreting the implications of particular animal testing results and determining how they relate to potential concerns for humans.

11.1 MALE REPRODUCTIVE TOXICOLOGY In the most basic sense, the functions of the male reproductive system are to produce and deliver the male germ cells, spermatozoa, in such condition that union with a female germ cell and subsequent development can occur. Toxic responses must interfere with either germ cell production or delivery. The reproductive organs are obvious targets for such responses, but damage in the nervous and endocrine systems can also be important due to their role in controlling reproductive function. This section will describe some of the toxic chemicals that affect germ cell production and their mechanisms of action. Additionally, information about some of the toxic responses that can cause problems with sperm delivery will be presented.

Susceptibility of Spermatogenesis The process of germ cell production in the male, spermatogenesis, provides clear examples of how cells may have enhanced susceptibility to certain classes of chemicals at particular times. In spermatogenesis, germ cells are produced from a pool of progenitor cells, stem cells, through a series of mitotic and meiotic divisions that eventually produce a large number of spermatozoa from each original stem cell and provide replacement stem cells. Spermatogenesis occurs in specialized, thick walled tubules within the testis called seminiferous tubules. The germ cells are initially located at the outer edge of the tubule (Figure 11.1) and move progressively toward the center of the tubule. From here the spermatozoa move along the tubules and into the duct system that will carry them out of the body. In a human, it takes about 64 days to produce a mature spermatozoon through this process, which continues throughout adult life. The rate of spermatogenesis increases dramatically following puberty until a hundred million or more sperm are produced each day. Spermatogenesis can be equated to a mass production process where constant high rates of production and high quantity of output are the focus. In a biological system, this requires an extremely active, rapidly dividing cellular environment within the testis.



Figure 11.1 Drawing of seminiferous tubule showing the migration of germ cells to the center during development and stages of spermatogenesis.

The high rates of cellular division and metabolic activity associated with spermatogenesis are the basis for susceptibility to certain types of damage. During the duplication of genetic material and cell division, DNA is particularly vulnerable to damage. In addition, many specialized cellular proteins and enzymes are needed and a high level of cellular respiration is required. Therefore, chemicals that can cause DNA damage or interfere with cellular protein function or respiration are of particular concern in rapidly dividing tissues. Examples include reactive electrophilic chemicals such as alkylating agents and ionizing radiation. Many chemicals or their metabolites that are considered to be relatively toxic have the ability to undergo chemical reactions with DNA or important cellular proteins. Depending on the particular chemical, DNA damage may result from direct interaction with the strands or with other cellular macromolecules involved in stabilizing the DNA. Reactions with DNA can affect base pairing and strand linkage. Protein damage can include modifying enzymes and carrier molecules such that they cannot participate in biochemical reactions. Anti-neoplastic drugs used in chemotherapy, such as methotrexate, adriamycin, cyclophosphamide, vincristine, and vinblastine, are good examples of reactive compounds that can cause failures of germ cell production. Some examples of reactive chemicals with common occupational or environmental exposures and particular concerns with regard to rapidly dividing spermatogenic tissues include:

• Acrylamide & ethylene oxide—extensively industrial use • Polynuclear aromatic hydrocarbons (PAHs)—combustion products • Ethylene dibromide & dibromochloropropane—fumigants/pesticides



Understanding general mechanisms of action can help categorize chemicals as to the effects that are possible. It is important to remember that while a mechanism for cellular damage to spermatogenic cells exists for reactive chemicals in general, not every alkylating agent or reactive metabolite will actually act as a specific reproductive toxicant. Other factors control the susceptibility of spermatogenic tissues to particular chemicals. Among the critical factors are the dose level that is received, the extent of distribution to the target tissue that occurs, and how/where metabolism occurs for the particular chemical. Individual and species differences in such factors explain the differences in susceptibility and demonstrate the need to proceed carefully when predicting whether a chemical will be a human reproductive toxicant. The energy associated with ionizing radiation, including x-rays, can also result in chemical modifications to DNA that affect its potential to be copied correctly and to direct cellular functions. While germ cells at all stages of spermatogenesis can be affected, it appears that some of the early stages are most susceptible to DNA strand breaks from irradiation. Such breaks can result in chromosomal malformations in germ cells. In addition, the death of damaged somatic cells (non-germ cells) can result in the collection of cellular debris in the duct system that carries the germ cells. In addition to DNA damage that can lead to cell death, there is also the possibility that modifications of DNA can be repaired incorrectly, producing a mutation. There are cellular mechanisms available to remove and replace damaged segments of DNA, but there is a certain error rate, albeit low, associated with this type of repair. When such repair errors occur in a germ cell, there is the possibility that the resulting mutation could be passed to offspring and become heritable. While this is theoretically possible and can be demonstrated in some experimental systems, the generation of an inherited human mutation following chemical-induced DNA damage has not been documented. Direct and Indirect Modes of Toxicity Another important toxicological concept well illustrated in the male reproductive system is the distinction between direct acting toxicants and indirect acting toxicants. Direct acting toxicants may be reactive chemicals or chemicals with sufficient structural similarity to molecules used in cellular communication that they can interfere with signaling pathways. Indirect acting toxicants may eventually cause damage in the same ways, but must first be modified through metabolic reactions or bioactivated. Ironically, reactive metabolites result frequently from the chemical reactions also used to break down and eliminate foreign chemicals. The metabolism may take place in the cells or tissues that are eventually damaged, or may occur in other organs, such as the liver. In the latter case, the toxic metabolites must be transported to the target tissue. Some of the best known reproductive toxicants have a direct mode of action on male reproductive tissues. Lead and cadmium are two examples of metals in this category. Lead can damage genetic material, disrupting cell division and resulting in cell death. While many different cell types are susceptible to damage from lead, the importance of continual division during spermatogenesis makes this process particularly vulnerable. Cadmium has an interesting direct mode of action on the vasculature surrounding the testis and epididymis, the adjacent tubular organ in which sperm are stored and mature. Sperm must be kept at temperatures slightly below core body temperature. Extensive and specialized vascularization is provided to remove heat from the testis. This vasculature is extremely vulnerable to direct damage by cadmium. When such damage occurs, the remaining vessels are unable to carry away as much blood, and thus as much heat. The combination of reduced perfusion with oxygenated blood and higher temperatures can subsequently destroy the spermatogenic cells. Many of the alkylating chemotherapeutic drugs have direct modes of action resulting in male reproductive toxicity, including busulfan and cyclophosphamide. Ethylene oxide, used extensively as a chemical intermediate in industry and for gas sterilization of medical devices and even foods, also has direct actions on cellular biomolecules and experimental studies suggest later stage germ cells are particularly susceptible to its damage. DNA is stabilized by extensive interactions with structural



elements called protamines during later stages of spermatogenesis and protamines are particularly vulnerable to reactions with ethylene oxide. These medically related examples indicate the delicate balance between therapeutic or beneficial uses and potential reproductive toxicity. The powerful alkylating properties of chemotherapy drugs allow them to work against rapidly dividing cancer cells and the risk of ancillary effects on other cells, even frank reproductive toxicity, may represent an acceptable tradeoff for the cancer patient. Also, the safety and benefits of dry, gas sterilization with ethylene oxide are clearly significant. While we can document the mechanisms of action producing male reproductive toxicity experimentally, there is no evidence that the ethylene oxide exposures associated with sterilization has produced reproductive toxicity in men. Again, the toxicology indicates what could happen if the right conditions existed, not what happens under the typical situations in which the chemicals are encountered. Many of the compounds of interest in occupational or environmental toxicology require metabolic activation to produce reproductive effects. The testis has the enzymatic capabilities for oxidative metabolism, a pathway that frequently produces reactive intermediates. While this activity is low compared to the liver, it is sufficiently high to produce toxic amounts of metabolites for some compounds. Metabolism of common industrial chemicals including the solvents n-hexane and the glycol ethers appears to contribute to their reproductive toxicity. Some of the phthalates, a chemical class used extensively as plasticizers and distributed widely in the environment, are also capable of affecting male reproductive tissues after metabolism. All of these examples are discussed further with regard to the specific cells they affect and have at least purportedly affected humans. There are also many other examples of indirect acting male reproductive toxicants where there is at least experimental or mechanistic information on toxic potential including the intermediate acrylamide and vinyl chloride, another common industrial intermediate also found in the environment, often as a breakdown product of dry cleaning solvent.

Cell-type Specific Toxicity Another principle illustrated by examining male reproductive toxicology is the specificity of action on certain cell types due to the characteristics of the cells or their metabolic potential. For some reproductive toxicants, there is varying sensitivity among the somatic and germ cell types. While this may be explained in some cases by the high levels of cell division and activity among the germ cells, in some cases there appear to be more specific factors involved. Besides the germ cells, there are two major types of somatic cells in the testis required for spermatogenesis, Sertoli cells and Leydig cells. Both of these cell types may also be specific targets for some toxicants. In addition, the microvasculature of the testis can be a specific target and the functional consequences of impaired circulation in the testis have been described above. Developing Sperm Cells As germ cells proceed through spermatogenesis, several different terms are used to distinguish the varying degrees of maturity. At the earliest stages are the spermatogonia, followed by the spermatocytes, the spermatids, and finally spermatozoa (Figure 11.1). Besides describing the developmental stage of the germ cells, these distinctions also correspond to some degree of toxicological specificity as certain stages are targeted by certain compounds. This specificity generally relates to which stages are the most sensitive to a particular agent. In most cases, as the dose increases or exposure conditions change, more than one stage can be affected. One of the occupational episodes that stirred interest in effects on male reproductive function was reported sterility among workers handling the pesticide dibromochloropropane (DBCP). Subsequent investigations suggested a toxic effect that would certainly explain sterility. The most significant cell type damaged by DBCP is probably the spermatogonia. Since these progenitor germ cells are at the base of spermatogenic cellular expansion, their destruction precludes future cycles of spermatogenesis. Thus, the expected observation would be a depletion of all the later stages and an inability to recover



spermatogenic potential after the toxicant was removed. This fits the observed effects on occupationally exposed humans. Spermatocytes, particularly at the pachytene stages, are susceptible to damage from ethylene glycol monoethyl ether (EGME), one of the glycol ethers with considerable potential for human exposure. A metabolite of EGME, 2-methoxyacetic acid (MAA), may cause indirect damage to the spermatocyte by decreasing lactate production in the Sertoli cells. Lactate is a key metabolic substrate for developing spermatocytes. Spermatids may be a particular target for ethylene dibromide (EDB) toxicity. This is another compound, used as a fumigant, for which there is a least some information that occupational exposures may have adverse effects on male reproduction. Effects on the later, spermatid, stages of spermatogenesis would be consistent with the abnormalities and deficits observed in some occupationally exposed workers. Experimentally, however, EDB turns out to be an example where the stage specificity breaks down as the dose increases. Spermatozoa can be affected by various toxic mechanisms. Epichlorohydrin (widely used intermediate in plastics/rubbers) is an example of a compound that appears to affect sperm motility by interfering with metabolism. Motility is required for fertilization and the needed energy is produced through specialized metabolic pathways that operate within the sperm. Chlorpromazine, a drug used to treat psychosis, appears to cause metabolic effects on sperm secondary to permeabilizing their membranes. Agents such as mercury and lead may have a combined effect on both sperm cell membrane dynamics and on the epididymis. Sperm complete their maturation in the epididymis, and the secretory and absorptive functions of the epididymal epithelium are required for the maintenance of viable sperm. Another important mechanism of toxicity may also be pertinent to mature sperm. Many biochemical reactions result in the formation of highly reactive radical groups. Cellular phospholipids, important to the structure of cell membranes, are particularly sensitive targets for reactions with these radicals. The resulting damage, called lipid peroxidation, can impair the integrity of cell membranes. Free radicals can be generated during the oxidative metabolism of many different compounds. Some of those that may be male reproductive toxicants are adriamycin, ethylene dibromide, and the herbicides paraquat and diquat. Free radical-induced lipid peroxidation may be important to sperm for two reasons: 1) the detoxification pathways that typically keep free radicals in check are modified in reproductive tissues and appear to be especially limited in the sperm cells, and 2) sperm contain highly specialized membranes that can be easily compromised. Investigations of lipid peroxidation in sperm have only begun recently, and currently, the only clear occurrence of such damage in human sperm is found in frozen semen samples. Freezing seems to destroy one of the major anti-oxidant defense enzymes making the sperm especially susceptible. As further investigations of chemical-induced lipid peroxidation are carried out, some of the membrane disruption associated with spermatotoxicity may be better explained.

Sertoli Cells Sertoli cells are in direct contact with the germ cells and provide support for them, both structurally and functionally (Figure 11.1). By virtue of specialized junctions between Sertoli cells, which isolate the germ cells from any other somatic cells outside of the seminiferous tubule, the Sertoli cells create a barrier that provides a degree of insulation and protection from chemicals distributed through the circulatory system. Thus, when Sertoli cells are targeted by toxicants, not only is their support of germ cell production impaired, the blood/testis barrier may be disrupted, exposing the germ cells to more potential damage. Due to their close relationship with the germ cells, it is not surprising that toxicants which specifically affect Sertoli cells have a subsequent effect on germ cell production. Some of the characteristics of Sertoli cell damage are that all stages of developing germ cells are impacted and the damage is frequently irreversible. This is due to the limited replacement of Sertoli cells; they divide relatively little in mature males. Also, part of the function of Sertoli cells is to initiate the sequence of



germ cell development. Without Sertoli cells, remaining progenitor germ cells are unable to begin the spermatogenic cycle. The solvent n-hexane is a demonstrated reproductive toxicant. Its testicular toxicity is related to interference with microtubule formation in Sertoli cells by the metabolite 2,5-hexanedione. The susceptibility of Sertoli cells to a microtubule poison is understandable since Sertoli cells form a scaffolding supporting multiple layers of germ cells and this function relies on the assembly of microtubules. This process involves extensive remodeling of the Sertoli cell architecture as the germ cells are moved through the seminiferous tubule, and such remodeling is heavily dependent on microtubule formation. Some of the phthalate plasticizers also appear to affect Sertoli cells. The toxicity appears to occur in the Sertoli cells, involving a breakdown of the attachments between Sertoli cells and germ cells. Thus, all of the spermatogenic cells in the developmental sequence at the time of intoxication may be compromised. Bioactivation of tri-o-cresyl phosphate (TOCP) is required prior to its effects on Sertoli cells. Interestingly, the metabolism occurs in the Leydig cells but does not appear to interfere with their function. The reactive metabolite subsequently reaches the Sertoli cells, which are damaged, and spermatogenesis is subsequently impaired. Another Sertoli cell toxicant of interest in occupational toxicology is dinitrobenzene (DNB). This compound, and structurally related analogues, appears to disrupt Sertoli cell function, possibly through involvement in a metabolic reaction cycle that depletes the cells of important reducing equivalents, impairing their functional support for the spermatogenic cells. Subsequently, all of the stages of developing germ cells may be compromised. The Sertoli cell tight junctions can also be affected by toxic agents, disrupting the blood/testis barrier. Platinum-based anti-neoplastic drugs, such as cisplatin, appear to operate in this manner. The germ cells divide improperly subsequent to this toxicity; however, it is not clear whether this is due directly to exposure through the disrupted barrier or if the Sertoli cells are incapable of directing germ cell development properly. Leydig Cells Leydig cells are located outside, but surrounding the seminiferous tubules. Their major function is producing testosterone, a key to regulating spermatogenesis as well as male reproductive development and behavior. There are several toxicants that can be demonstrated to affect Leydig cells experimentally. It is not clear, however, whether any demonstrable human reproductive toxicity is primarily due to actions on Leydig cells. In part, this is because androgen regulation is so complex that it is difficult to determine which observations are primary toxic responses and which result secondarily from hormonal dysregulation. Also, several of the toxicants with specific actions on Leydig cells can also cause responses in other cell types, depending on the dose received. A well-described experimental Leydig cell toxicant which does not appear to directly affect other testicular cell types is ethane-1,2-dimethanesulfonate (EDS). This compound affects androgen production by the Leydig cells and appears to interfere with specific early steps in the synthesis of steroid hormones. While the mechanism leading to cell death is not clear, EDS does kill the Leydig cells. The eventual result of EDS toxicity is, somewhat predictably, impaired spermatogenesis. In addition, Sertoli cells, which are dependent to some degree on Leydig cell secretions, may also be damaged. Among the toxicants with significant human exposure that operate primarily on the testis, diethylhexyl phthalate and its metabolites are the only example with specific Leydig cell effects. In addition to their Sertoli cell effects, the phthalates also appear capable of interfering with steroid synthesis and Leydig cell function. It is not yet clear what the relative contribution of the Leydig cell damage is to the resulting spermatotoxicity. Hormonal Regulation and the Hypothalamic-Pituitary-Gonadal Axis Disruptions of male reproductive function can also occur secondary to toxic responses in the endocrine system. Androgen production in the testes is regulated primarily by luteinizing hormone (LH), a



gonadotropin released by the pituitary gland. Gonadotropin secretion is in turn regulated by gonadotropin releasing hormone (GnRH), secreted by the hypothalamic portion of the brain. This hierarchical arrangement, where the hypothalamus regulates the pituitary which in turn regulates the gonads, is known as the hypothalamic-pituitary-gonadal axis. The hypothalamic-pituitary-gonadal axis is illustrated for both males and females in Figure 11.2. From a toxicological perspective, this arrangement creates even more sites where toxic responses may have an impact on reproduction. With this in mind, it is not surprising that some compounds generally considered to affect the central nervous system, can impact Leydig cell function and male reproduction. The toxic effects of ethanol are wide ranging and complex. Experimental evidence for direct testicular toxicity is not clear; however, it is clear that alcoholics suffer decreased testosterone levels and subsequently, decreased gonadal function. In alcoholics, the ability to stimulate testosterone production appears to be impaired. Experimentally, it can be demonstrated that ethanol affects LH release. It is clear that alcohol interferes with male reproduction by affecting endocrine regulation, and ultimately, in part, Leydig cell production of testosterone, but the relative contributions of direct testicular action and toxic responses elsewhere in the regulatory axis are not known. A variety of other drugs and industrial compounds also affect male reproduction through endocrinerelated mechanisms. These include the anti-hypertensive drug propanolol, the opiates, and tetrahydrocannabinol (THC). The use of such drugs is pertinent to occupational toxicology, since their effects can confound observations on reproductive impairment related to direct occupational exposure. Carbon disulfide, the pesticide chlordecone, and the phthalates are examples of industrial chemicals that can disrupt the endocrine axis. Besides the potential impact on spermatogenesis, there is another aspect of toxicity to the hypothalamic-pituitary-gonadal axis that affects male reproductive function. Both libido, or behavioral drive, and physical aspects, such as penile erection and ejaculation, are controlled by the central nervous system. Libido is controlled primarily by androgens and any of the drugs or industrial compounds that can dysregulate the endocrine axis and affect androgen production can affect libido. Alcohol and THC

Figure 11.2 The hypothalamic-pituitary-gonadal axis for human males and females illustrating the sequence of control from the brain to the secretion of gonadotropins by the pituitary, to the production of steroid hormones in the gonads. The major hormonal products and their contribution to regulatory feedback loops are indicated. GnRH—Gonadotropin Releasing Hormone.



are common examples of compounds that decrease libido. Carbon disulfide and chlordecone-exposed workers also reported decreased libido. It is important to remember, however, that neurological and psychological factors play a tremendous role in libido and the relative role of chemical-induced mechanisms for decreasing libido is likely minor. Where there are demonstrable effects that relate to endocrine balance, it is not surprising to find reported effects on libido. However, where the claimed effect is simply a report of decreased libido, there is little likelihood of establishing a clear toxicological basis. Other factors that can preclude sperm delivery are impotence and inability to ejaculate. These aspects of reproductive function are centrally controlled, but the autonomic nervous system also plays a key role in coordinating the physical events. The most common examples of chemical-induced impotence are some of the adrenergic antihypertensives, methyl DOPA, clonidine and guanethidine, and the opiates and ethanol. Psychoactive drugs, such as chlorpromazine and diazepam, can also produce impotence. As with effects on libido, psychological factors are major contributors to impotence. Clinical findings with humans suggest that the preponderance of reported problems with impotence are psychological and not related to toxic responses.

Endocrine Feedback and Potential Dysregulation Some of the key signaling molecules essential for endocrine feedback loops are the steroid hormones and other protein-based hormones. Since both the absolute levels and the ratio between androgens and estrogens serve as signals to the hypothalamic-pituitary-gonadal axis, modifications of this balance can subsequently disrupt endocrine function and result in reproductive effects. A highly publicized case of occupational exposure leading to reproductive impairment involves the pesticide chlordecone. Occupationally exposed men appeared to have reductions in fertility. Subsequent analysis provided a possible mechanism for such effects, as chlordecone turned out to have estrogenic activity. The endocrine dysregulation produced by elevated estrogenic feedback could explain the neurotoxic and reproductive effects that have been attributed to chlordecone. The role of exogenous compounds with estrogenic, anti-estrogenic, and anti-androgenic activity is a current source of substantial controversy. The extremely potent toxicant dioxin has the potential to interfere with endocrine balance, and the contribution of this activity to its toxicity is hotly debated. A variety of pesticides, including DDT, the carbamates, and mirex, are reported to possess endocrine activity as are some of the polychlorinated biphenyls (PCBs). What is not yet clear is whether environmental levels of exposure to such compounds can actually produce endocrine-mediated toxic effects. Most synthetic steroid-like molecules can only displace the actual endogenous compounds in molecular interactions to a very limited degree. Also, typical exposure levels of the exogenous compounds are extremely low, and metabolic deactivation before they reach the target tissue further limits the potential activity. On the other hand, the endocrine system functions with very low effective concentrations of signal molecules at the target cells. The significance of endocrine-mediated toxicity in the male reproductive system is not yet clear, however, it is a topic of intense interest and potentially wide ranging ramifications for the future (see further discussion in Section 11.4).

Male Reproduction Summary There are many industrial and pharmaceutical compounds that are potential male reproductive toxicants (Table 11.1). Such chemicals may interfere with the development of germ cells directly, or indirectly, by disrupting the cellular and endocrine factors involved in supporting and regulating spermatogenesis. The regulatory roles of the neuroendocrine system are also important to the delivery of sperm. The potential for various compounds to work through a particular toxic mechanism, however, does not necessarily indicate that there is a relevant human reproductive risk associated with that mechanism, or even with the compound at all. Mechanistic possibilities and experimental results must be carefully evaluated in terms of those effects that are actually observed in exposed humans. There


REPRODUCTIVE TOXICOLOGY TABLE 11.1 Suspected Human Male Reproductive Toxicants Industrial/Environmental

Pharmaceutical Agents and Drugs

Cadmium Carbon disulfide Chlordecone Dibromochloropropane (DBCP) DDT Diethylhexyl phthalate Dinitrobenzene Epichlorohydrin Ethylene dibromide Ethylene oxide n-Hexane 2-Hexanedione Ionizing radiation Lead Mercury 2-Methoxyethanol Tri-o-cresylphosphate

Adriamycin Busulfan Chlorambucil Chlorpromazine Clonidine Cyclophosphamide Diazepam Ethanol Guanethidine Methotrexate Methyl DOPA Opiates Propanolol Tetrahydrocannabinol Vincristine Vinblastine

have been relatively few instances of occupational exposure leading to demonstrable decreases in male fertility. Cell type susceptibility to toxic injury can be roughly generalized with germ cells as the most sensitive, followed by Sertoli cells and then Leydig cells. The hierarchical regulation of spermatogenesis underlies this since development of germ cells is often affected by toxicants acting on the Sertoli and Leydig cells. In turn, Sertoli cells are often affected by both Sertoli cell and Leydig cell-specific toxicants. In addition, there need not be any specific site within the testis targeted by a reproductive toxicant. Reproductive function is susceptible to agents that interfere with the central nervous system and autonomic nervous function because of the importance of neuroendocrine regulation.

11.2 FEMALE REPRODUCTIVE TOXICOLOGY For the sake of this chapter, female reproductive toxicology will only include toxic responses of mature females not directly affecting the fertilized egg or subsequent development. All post-fertilization toxicity relating to the developing offspring will be considered developmental toxicology (see Section 11.3). With this limitation, female reproductive function can be described by the same characteristics outlined for the male—the key features being the production of female germ cells, eggs, and transport of the germ cells, in this case the sperm and eggs, to the site of fertilization. With reference to human occupational and environmental exposures, substantially less is known about female-specific toxicology compared to male or developmental toxicology. One of the reasons is that reproductive impairment of human females is difficult to both establish and analyze. Unless there is an observed prolonged inability to maintain a pregnancy there is often little reason to investigate whether toxic responses may have affected female reproductive function. While this may also be true of occupationally exposed men as well, fast and fairly sensitive means to test the reproductive capacity of men are available. Semen samples are easily obtained and evaluated, and, while such analysis cannot definitively determine fertility, abnormalities are obviously a sign of a potential problem. On the other hand, germ cell production by women is very difficult to monitor and potential indicators, such as failure to menstruate or irregular menstruation, occur frequently enough and for such a variety of



reasons that associating these occurrences with a particular exposure is difficult. In addition, pregnancy failure occurs frequently as well, and can be due to problems with either the female or the developing embryo. There are also some fundamental differences in female germ cell production that relate to the potential mechanisms of toxic injury. In contrast to the constant cellular division required to generate millions of new sperm each day, the ovary contains all of the eggs that will ever be ovulated, and then some, by the time of birth. All mitotic divisions and the initial stages of meiosis have been completed by the middle of fetal development and result in the generation of around 10 million primary oocytes arrested in the meiotic progression. Subsequently, a rapid process of degeneration (atresia) occurs, and there are around one million primary oocytes left at birth. Atresia continues somewhat more slowly throughout life, and the mature woman has around a half a million oocytes with the potential to develop into mature eggs. In the human, a handful of the primary oocytes begins the process of folliculogenesis during each menstrual cycle, but typically only one forms a completely mature follicle and is ovulated. Estimating the number of menstrual cycles in a typical reproductive lifetime, only around 500 germ cells ever complete development. There is no further division of the germ cells prior to leaving the ovary and while there are somatic cells that divide and develop to support the maturing egg, this occurs over the course of about two weeks. Clearly, female germ cell production does not rely on mass production and rapidly dividing cell populations. Correspondingly, the overall process is less susceptible to cytotoxic compounds that operate against dividing cells. Oocyte Toxicity This being noted, however, some of the powerful antineoplastic drugs are still capable of disrupting oocyte development, including cyclophosphamide, chlorambucil, busulfan, and vinblastine. Experimental results indicate that busulfan may be capable of destroying the arrested, primordial oocytes, preventing further maturation and ovulation. Other alkylating agents appear to work only on the follicles developing at the time of exposure. For the reproductive process as a whole, this is advantageous, since other primordial oocytes are still available for future cycles of folliculogenesis. While there is experimental evidence that ovarian-derived metabolites of some polycyclic aromatic hydrocarbons (PAHs–combustion byproducts) are also capable of destroying the primordial oocytes, there is limited information on ovotoxicity relating to occupational or environmental exposures. There are no good examples of human reproductive toxicants that appear to affect egg production under such circumstances. This is surely due in part to difficulties in analyzing ovarian toxicity that have slowed down the identification of toxicants with the ability to target the ovary. Experimental findings suggest that lead, mercury, and cadmium are capable of damaging oocytes, but in light of their generalized toxicity, this is hardly surprising and probably does not occur in the absence of their major toxic effects on other systems. Somatic Ovarian and Reproductive Tract Toxicity There are a few experimental examples of toxic compounds that can have a direct effect on the somatic ovarian cells or the reproductive tract. For the ovary, such toxicity relates to successful germ cell production since the ovarian cells differentiate into layers known as the granulosa and theca, both integral to the development of mature follicles. 4-Vinylcyclohexene, an industrial compound used in the production of epoxy resins, appears to be capable of producing generalized ovarian atrophy. Similar atrophy can be produced with the antibiotic nitrofurantoin, which also has some specific toxicity toward the cells lining the follicles. The granulosa layer in particular can be targeted by metabolites of some of the phthalates. It is not clear how such toxicity relates to other effects of these compounds. The female reproductive tract is responsible for transport of the eggs and sperm to the site of fertilization and subsequent transport of any fertilized embryo to the site of implantation in the uterus. Fertilization occurs in the uterine tube, also known as the Fallopian tube or oviduct (Figure 11.3). The


Figure 11.3 Cycle of follicular development and the progression through the human female reproductive tract. Note fertilization occurring in the upper portion of the oviduct and implantation occurring after the earliest cell divisions at the blastocyst stage. (Reproduced with permission from Dean J., American Journal of Industrial Medicine, 4 (1983) 3. pp. 31–39. Figure 1.)



early stages of embryonic development occur in the uterine tube, and then the embryo moves down to the uterus to implant. All of this transport requires a patent lumen in the oviduct, and the movement occurs due to a combination of ciliary beating and muscular action. In other words, the tract is more than a transport tube and appropriate biological function is required to support the early embryo in particular. Atrophy of the oviduct or uterus can clearly prevent the transport of the germ cells and embryo. Cadmium can produce such atrophy, and presumably other metals which cause overall tissue degeneration could as well. For cadmium, the response is a general metabolic inhibition of the cells in the reproductive tract, leading to cell death and declining organ weight. Not surprisingly, uterine, as well as ovarian, cyclicity is impaired. Lead is another example of a metal that may directly affect the cells lining the uterus, subsequently interfering with proper uterine cyclicity. The features responsible for moving the germ cells and early embryo appear to be a potential target for some components of cigarette smoke. Experimental exposures have resulted in both increases in muscular related oviductal and uterine motility and immobilization of the cilia lining the oviduct. The pertinence of these observations to human exposures and the observed effects of smoking are not clear. However, the potential outcomes, improper migration of the germ cells precluding fertilization, or the early embryo preventing proper implantation, are consistent with the overall decreased fertility and increases in irregular cyclicity reported for smokers.

Hormonal Regulation of Reproductive Function and Associated Toxicity Endocrine regulation of female reproduction is even extremely complex. The hypothalamic-pituitarygonadal axis is present in females, and GnRH release and the timing of changes in the relative levels of the two major gonadotropins, LH and follicle stimulating hormone (FSH) are linked to the ovarian follicular cycle (Figure 11.2). This is accomplished by endocrine feedback loops involving the steroid hormones estrogen and progesterone, as well as some protein hormones. In a simplified form, the female endocrine cycle can be considered to start with increasing levels of FSH production by the pituitary during the early stages of folliculogenesis (Figure 11.4). As the follicles develop, the granulosa cells surrounding the oocyte are a major source of estrogens. As the estrogen levels increase, FSH production is shut down and production of the other gonadotropin, LH increases. When the follicle is fully mature, there is a surge of LH release, directing ovulation and the subsequent formation of a progesterone secreting tissue, the corpus luteum, at the site where the follicle had been. Progesterone levels rise and support the establishment of a pregnancy. Progesterone also causes the levels of both gonadotropins to drop. If there is no pregnancy the corpus luteum degenerates and progesterone levels decline, releasing the inhibition of gonadotropin secretion. FSH can again rise, starting the cycle over. For the human, this is the point where menses occurs, lasting through the early stages of the next follicular cycle. Disruptions at the gonad, pituitary, or hypothalamus during the preovulatory stages can cause a failure of folliculogeneis, and there will be no ovulation for the affected cycle. Later disruptions can cause a failure of the corpus luteum maintenance, preventing the establishment of pregnancy if the egg had been fertilized, or causing a shortened cycle. Alternatively, interference with luteal degeneration can cause a cycle extension and this may be manifest as delayed menstruation. It is clear that there are plenty of opportunities for endocrine disrupting toxicants to interfere with both the follicular cycle and the ability to maintain a pregnancy. There are examples of toxicants that can interfere with female hormonal regulation. Lead toxicity, for instance, is associated with decreased progesterone production. This may in part explain its historical use as an abortofacient, since progesterone is the key hormone for establishing and maintaining pregnancy. The actual mechanism of lead-induced progesterone inhibition is not clear. However, the established effects of lead on the neuroendocrine system could reasonably be expected to interfere with the hypothalamic or pituitary secretion patterns required for the luteal phase of the cycle.



Figure 11.4 Sequence of hormonal peaks and hormone balance with concomitant follicle development during the human female reproductive cycle. Timeline indicates number of days beginning with the first day of menses; E—Estrogen; FSH—Follicle Stimulating Hormone; LH—Luteinizing Hormone; P—Progesterone. (Reproduced with permission from Mattison (Ed.), “ Reproductive Toxicology,” American Journal of Industrial Medicine, 4 (1983): 1–2. P. 20.)

Cigarette smoking, probably the nicotine, and alcohol are both capable of interfering with GnRH release. Alcoholics have been observed to be unable to produce the LH surge needed for ovulation, and tobacco use is associated with decreased estrogen levels. These observations are consistent with irregularities of the cycle that have been associated with alcohol and tobacco use. The decreased libido and lack of cyclicity associated with narcotics abuse may relate to the hypothalamic depression that these compounds cause. Clearly, drugs of abuse and smoking have wide ranging physiological effects, and their toxicity is not specifically tied to female reproductive function. However, the endocrine actions of such compounds clearly present a mechanism by which associated reproductive impairments could be explained. Clomiphene citrate is an example of a compound that is used therapeutically to interfere with the female endocrine balance. This anti-estrogen is used to hyperstimulate ovulation in infertile women. It appears to work by increasing gonadotropin release, allowing more follicular development to occur. The anti-estrogenic properties prevent the estrogen-mediated shut down of FSH production. There are, however, other ramifications of the anti-estrogenic activity including decreased luteal function and decreased ability of the uterus to establish a pregnancy. The mixture of desired and undesired outcomes is a good example of the complex outcomes associated with endocrine interference.



Occupational and environmental exposures that could dysregulate the female endocrine pattern are a major topic of current investigation. Dioxins, other polycyclic chlorinated compounds and organochlorine and organophosphate pesticides are all potential compounds of concern. As described in the Male Reproductive Toxicology section, the concern is that many of these compounds have some type of estrogenic or androgenic activity because they are able to replace the endogenous compounds in cellular interactions. The theoretical potential for such compounds to affect female reproductive function is clear. However, demonstrating such an effect is extraordinarily difficult, and there are still no convincing examples of endocrinologically active compounds causing reproductive impairment in women through typical occupational or environmental exposures. The endpoints that can generally be observed for women are menstrual interval, fertility as measured by time to pregnancy, and ability to carry a pregnancy to term. This last potential effect will be discussed in the Developmental Toxicology section. There are such extreme interpersonal differences in menstrual interval and regularity and so many established causes for missed or delayed menstruation that associating any variation with a particular chemical exposure is difficult. Many cultural and occupational factors are clearly relevant for affecting time to pregnancy, and difficulty achieving a pregnancy when desired may affect as much as 25 percent of couples in the United States at times. It is clear that this is not always due to female reproductive problems, but this “ naturally” occurring background obscures potential toxicologically mediated effects. The high degree of interest in endocrine disruption as a potential mechanism for female reproductive toxicity is driving extensive investigations of this hypothesis. In the future it should become clearer whether environmental estrogens and other endocrinologically active compounds can actually reach levels at which they can produce a significant endocrine disruption and subsequent reproductive impairment. Currently, we are left with a potential mechanism for reproductive effects and candidate compounds that could act through this mechanism, but no clear demonstration of any of the candidates posing an actual risk through such a mechanism for humans following occupational or environmental exposures. The potential for exposure to chemicals that could alter endocrine processes and the need to use pharmacological agents known to cause reproductive toxicity opens up controversial occupational and societal issues about restricting women’s chemical exposure. What types of data or experimental results should be sufficient to indicate the need to control occupational exposures? When considering whether women should be excluded from certain jobs during certain segments of their reproductive lives, suddenly, the need to get beyond the uncertainties of extrapolating doses and mechanisms of toxicity from animal testing becomes crystal clear. The associated issues are as widely disparate as the economic impacts of possibly needing to move employees in and out of certain jobs or requiring specialized exposure control equipment to the potential for claims of discrimination, should women be excluded from opportunities on the basis of concerns they do not believe are relevant for them. Alternatively, when deciding a certain therapy is needed, what constitutes an adequate representation to the patient of the risks to herself or a developing fetus? Clearly, we cannot always discard effective treatments. The recent return of thalidomide, discussed below as the cause of one of the most notorious cases of human developmental toxicity, is a shining example. Thalidomide turns out to be a particularly effective treatment for patients suffering complications of leprosy or some complications of AIDS. It may further be an effective sedative for cancer patients and those suffering autoimmune diseases. These uses expand the patient population where reproductive effects are a possible concern. Effective patient education and carefully planned distribution policies may be relatively straightforward for thalidomide, where the toxic timing and dosage is established and the outcomes are readily documented, but what is the appropriate balance between protection and restrictiveness for other drugs? Female Reproduction Summary Compared to the male, there are relatively few female-specific reproductive toxicants that are not related to developmental toxicity (Table 11.2). (Developmental toxicants will be covered in the following section.) In part, this is due to the difficulties in analyzing oocyte production and determining


REPRODUCTIVE TOXICOLOGY TABLE 11.2 Suspected Human Female Reproductive Toxicants Industrial/Environmental

Pharmaceutical Agents and Drugs

Arsenic Cadmium Diethylhexyl phthalate Dioxins Ionizing radiation Lead Mercury PCBs—coplanar forms

Androgens, estrogen, and progestins Busulfan Chlorambucil Cyclophosphamide Ethanol Opiates Vinblastine

when there is a reproductive impairment. Also, the relatively small proliferative cell population in the ovary and the intermittent nature of the proliferative stage makes the ovary less susceptible to disruptions of cell division. The compartmentalization of the active germ cell and its supporting follicle is also pertinent to ovarian toxicology since this means that only a few germ cells are vulnerable during a given cycle. This decreases the likelihood that a toxic injury will cause permanent interference with oogenesis. While there are occasional toxicants, such as busulfan, that can wipe out the arrested population of primordial oocytes and prevent future follicular development, in general, the arrested cells are fairly resistant to damage. Probably the most significant type of toxicants for female reproductive function are those that interfere with the dynamic endocrine balance required for folliculogenesis and ovulation. A wide array of toxicants have at least the potential to interfere with the female hormonal pattern, including heavy metals, drugs of abuse, and some chlorinated biphenyls. Chemicals that can structurally mimic the steroid hormones, or have a competing functional activity, can be found among a variety of therapeutic drugs, pesticides, and environmental contaminants. Though a mechanism for interfering with female reproductive function is suggested, the impact of endocrinologically active chemicals, especially through environmental exposures on human reproduction, is not yet clear.

11.3 DEVELOPMENTAL TOXICOLOGY Recognition that environmental factors could cause congenital defects grew following the 1941 report by Gregg that there was an association between exposure to Rubella virus (German measles) during pregnancy and the occurrence of blindness and deafness in the offspring. Further analysis following a Rubella epidemic and the thalidomide incident solidified people’s awareness of the potential for prenatal exposures to cause developmental defects. Part of the reason that recognition was so late in coming, even in the scientific community, is that the placenta was thought to serve as a barrier preventing any potentially harmful agents from reaching the fetus. Since the 1960s it has become clear that the placenta is actually quite porous to chemicals of the molecular size that encompasses all but the largest drugs and industrial compounds. Developmental toxicity testing has now become commonplace, and many agents that can affect development, chemical and biological, as well as physical phenomena, have been identified. As we have seen in male and female reproductive toxicology, experimental demonstrations of toxic potential far outnumber demonstrated cases of human developmental toxicants. This may be due in part to species differences and the high doses used in experimental protocols; however, with developmental defects, it is also not always clear whether the experimentally demonstrable differences in structure or behavior can be extrapolated to humans and considered abnormal. Overall, the class of chemicals with demonstrated human effects are less important in the population than both biologically infective agents, such as Rubella virus, syphilis and cytomegalovirus, and maternal metabolic disor-



ders, such as diabetes and phenylketonuria. In turn, genetically based developmental disorders still surpass all of the environmental phenomena that can affect development in terms of proportional importance within the population. This section will cover the toxicology relevant to all stages of development from fertilization onward. This will include the action of toxicants on the mother that affect the ability to establish and maintain pregnancy, as well as direct actions on the fetus. The developmental stages can be divided into 1) the preimplantation stage, where toxicity generally affects the entire organism and there is typically an all or none response (i.e., there is repair or the developing organism is aborted), and 2) the later embryonic and fetal stages, where specific structural defects can occur. The most sensitive period for teratogenesis, the production of congenital defects, is during organogenesis in the embryonic period. Mammalian development can be thought of as an expansive flow diagram (Figure 11.5). Following fertilization there is a very particular sequence of events that is followed, directed by the expression of certain genes at certain times. From a toxicological point of view, disruptions during this relatively linear phase generally derail the entire developmental sequence. Certain developmental steps serve as branch points and once particular branches are followed, the occurrence of events on that branch will not necessarily affect other branches. As development progresses, many smaller branches are reached, which may each relate to the development of certain tissues, cell types, or regions of a structure. Toxicologically, when specific sequences are affected, the response may be restricted to the features that develop out of that particular sequence. This illustrates how very specific defects can occur in response to toxicants or other environmental factors and exhibit clear time dependence. Spontaneous Abortion and Embryonic Loss Recent improvements in the ability to measure human chorionic gonadotropin, a very early indicator of the presence of an embryo, have allowed reasonable estimates of early pregnancy loss. Overall, more than 50 percent of fertilized eggs/embryos are lost through spontaneous abortion. Around 30 percent are lost after implantation but before the first menstrual period is missed. An additional 20–25 percent are lost after they have been clinically recognized as a pregnancy. There are also probably substantial pre-implantation losses, but these are much harder to accurately estimate. Presented another way, it appears that the chance of achieving a full-term pregnancy for any one menstrual cycle in which fertilization is likely to have occurred (based on non-contracepted intercourse with ovulation) is around 25 percent. The preponderance of embryonic loss occurs during what was described as the linear phase of development. Based on chromosomal analysis of spontaneously aborted embryos, approximately two-thirds of this loss can be explained by gross genetic abnormalities. Around 10 percent can be attributed to a known environmental cause, and a cause cannot be determined for the remainder. While some of the genetically associated loss could be related to environmentally mediated DNA damage, it is clear that most is due to major chromosomal aberrations associated with germ cell production and fusion. Once again, the potentially chemically induced responses are hidden within an extremely high background rate of embryonic loss. In general, the response to toxicological insult during the early embryonic stage is considered to be an all or none event, where damage up to minor cell death is completely repairable and major cellular disruption or death results in abortion of the pregnancy. This is based on the flexibility of the cells in the early embryo, which allows them to functionally replace a few lost cells. As development progresses, most cell lines become committed to a particular fate and such compensation is less likely. The limited responses available during early embryogenesis mean that typically the endpoint of concern for toxic exposures is spontaneous abortion. While this criterion has held over the years for most toxicants, there are recent experimental results which suggest that very early exposure to ethylene oxide and some other mutagens may cause responses that are manifest much later in development. This implies that the exposure may result in a sublethal injury that is not repaired, nor are all of the


Figure 11.5 Developmental tree indicating the time during human gestation at which the development of various major organ systems becomes a separate progression from the rest of the organs and tissues. The branching corresponds roughly to the periods of tissue-specific teratogenic sensitivity.



affected cells replaced. It is not clear how many toxicants could create developmental defects in this manner, as they would likely have to be capable of producing specific, minor changes in DNA, nor is it clear that the experimental conditions are relevant to humans. Spontaneous abortion remains the most useful indicator of early embryonic toxicity, and there are both experimental and occupational examples where chemical exposures appear to affect spontaneous abortion rates. One of the most investigated, and most controversial, occupational examples relates to anesthetic gases. Studies both large and small have reported both positive and negative results when looking for elevated rates of spontaneous abortion among health care professionals using gas anesthetics such as nitrous oxide and halothane. The potential effect does not appear to relate to paternal exposure since there is not an observable elevation in the spontaneous abortion rate among the wives of occupationally exposed men. Methodological flaws have called some of the positive results into question. At this point, the most defensible conclusion is that based on the evidence there is an elevated incidence of spontaneous abortion among women in such occupations; however, the association between anesthetic exposures and the spontaneous abortion rates cannot be reliably demonstrated. This suggests that other, unidentified factors present in the study populations could play an important role. Carbon disulfide, dimethylformamide, and some of the phthalates are other examples in which investigations of spontaneous abortion rates have detected differences that may be attributable to occupational exposures of women. The data for carbon disulfide are the most convincing in terms of documenting an association, but even this conclusion is weakened because the proportionate increases are small and not clearly out of the expected background range. Chloroprene, an industrial chemical used in polymer manufacture, is an example in which male exposure may have subsequent effects on spontaneous abortion. In this case, the wives of occupationally exposed men showed elevated spontaneous abortion rates. This could be classified as a male reproductive effect, if it results from an effect on the sperm that necessarily occurs prior to fertilization, and suggests that some types of sperm damage may still be compatible with the ability to fertilize an egg. While such a mechanism of toxicity could explain the observations, this is not recognized as a common mode of toxic action, and it is thus difficult to exclude some other explanation not directly linked to the chloroprene exposure of the men. There are many compounds that can be shown experimentally to cause spontaneous abortions. Some of the classes might be expected based on their common toxic effects, such as the antineoplastic drugs and heavy metals. Their cytotoxicity is well known, so embryonic interference is hardly surprising. In addition, solvents such as benzene and toluene, many chlorinated pesticides and herbicides, PAH’s, and aldehydes such as formaldehyde can all experimentally cause early pregnancy failure. In short, most cytotoxic chemicals have the ability to interfere with early development under experimental conditions. The relevance to human exposure conditions and potential dose levels, especially in the occupational setting, is not clear. A common occupational chemical exposure that illustrates the difficulties in establishing embryotoxicity occurring in humans is the use of ethylene oxide. Large quantities of this chemical are used in manufacturing, especially for the production of ethylene glycol antifreezes. Female workers clearly have potential industrial exposures. In terms of the number of exposed workers, even more significant is the use of ethylene oxide as a sterilant of medical devices. Ethylene oxide exposure during unloading of sterilizers and in the area where the sterilized packages are aerated can be significant enough to produce toxic responses in other organs systems. This was especially true prior to interest in the potential long-term effects that grew during the 1980s. Also, ethylene oxide clearly causes embryotoxicity and death and structural abnormalities during the fetal stages in animal tests. These factors suggest that ethylene oxide could be a developmental concern for occupational exposure levels. Epidemiological studies of ethylene oxide exposed workers are equivocal. Though occasional findings suggesting elevated spontaneous abortion rates among potentially exposed workers have been reported, this result has been inconsistently observed. Furthermore, the largest studies, best designed to account for exposure levels and potential biases, have been routinely negative. So, the database stacks up as follows: 1) the toxic potential from animal tests is clear, 2) the potential for human exposure



is clear, 3) suggestive associations have been reported in relatively few worker studies that have been criticized for inability to clearly establish exposure levels and for sensitivity to possible biases, and 4) no association is supported by the biggest and best controlled studies.

Toxic Responses of the Embryo and Fetus For many people, the possibility of congenital defects is the most alarming aspect of reproductive toxicology. Congenital defects are any morphological, biochemical, or functional abnormalities that result from an occurrence prior to birth. The defect may not be detected until later, as with some learning deficits, but the biological basis for the defect occurs during uterine development. Most congenital defects are not due to chemical exposures, but it is clear that some defects have been caused by drugs and environmental exposure. The rate of major congenital structural malformations runs at about two to three percent of live births. When other more subtle congenital defects are added in, the rate reaches about seven percent of live births. Approximately 15 percent of these defects are linked to an inheritable disorder at a known gene locus, such as Tay Sachs disease or hemophilia. Another 10 percent are linked to major chromosomal malformations, such as monosomies or trisomies. Overall, genetic factors can account for up to 35 percent of congenital defects. Identifiable external factors (physical, biological, and chemical) account for around 10 percent of congenital defects. Estimates of additional, uncharacterized chemical and drug-induced defects account for between one and five percent. There remains a large proportion of defects without a well understood cause. Teratogens are agents, chemical or otherwise, capable of creating congenital defects. They are generally considered to create specific defects during the period of organogenesis, which begins around five weeks after fertilization for humans, and continues for various organs through most of the second trimester of pregnancy. For many teratogens with specific structural targets, the fetus shows a period of sensitivity corresponding to the development of the target structure. This is significant to human concerns since it means that to cause a defect, exposure must generally occur in a particular window of time during pregnancy. Disruptions of Tissue Organization The prototypical example of a teratogen with a narrow window for toxic potential is thalidomide. This drug was widely used in the late 1950s and into the 1960s to treat morning sickness and as a sedative. A sudden rise in children with limb deformities was associated with mothers who had taken thalidomide. The critical window of sensitivity was identified based on the severity of the deformities and period during which the mother had used the drug. A relatively narrow window in weeks 6–7 of gestation was identified in which exposure to thalidomide produced deformities in nearly all infants. Exposures after this time were associated with minor and less prevalent defects. In addition to the striking limb deformities, thalidomide exposed infants also exhibited congenital heart and renal defects along with ear deformities. Despite extensive animal experimentation in the aftermath of the thalidomide incident, the mechanism of action has still not clearly been determined. The probable reactive metabolites have been identified, and mechanisms of actions, such as interference with vitamin or amino acid metabolism in the developing limb bud and direct disruption of DNA in this region, have been suggested. It seems unfortunately ironic that while so many chemicals have a clear potential mechanism of action, yet no clearly observable human effects, perhaps the best example of an actual human teratogen has been recalcitrant to the studies that might identify the mechanisms that actually operate to disrupt human development. With thalidomide removed from use by pregnant women, the most significant teratogenic drug is isotretinoin, or Accutane, a highly effective agent against cystic acne. This drug is especially important in relation to teratogenicity, precisely because it is so effective and there is not a suitable replacement. Thus, despite its ability to effectively produce major fetal deformities, its use continues. Despite aggressive warnings by physicians and many exclusionary policies that attempt to prevent patients at



risk for becoming pregnant from taking the drug, developmental deformities related to isotretinoin exposure continue to be reported. The defects associated with isotretinoin are wide ranging and include craniofacial deformities, including cleft palate, and cardiac and central nervous system abnormalities. This can be understood on the basis of the role of the retinoids in normal development. Isotretinoin is a synthetic retinoid, or chemical relative of Vitamin A. Gradients of certain retinoids in tissues appear to play a major role in the organization and orientation of tissue growth. The direction of cellular growth needed to extend a limb, for instance, is guided by retinoid signals. Disrupting this road map with exogenous retinoids could clearly be a basis for inappropriate development. Exogenous exposure to most retinoids can produce developmental defects, at least experimentally. Isotretinoin is a good example of a teratogen that works by interfering with the chemical signaling used to guide development. Fetal Hydantoin Syndrome Another class of teratogenic drugs is still used by pregnant women because the developmental risks are less than the risks of removing the drug. Diphenylhydantoin— phenytoin, valproic acid, and other anticonvulsants are used in epileptics to prevent seizures. They are also teratogenic. In these cases, however, the therapeutic regimens are not associated with substantial numbers of congenital defects, and the potential for injury to both the fetus and mother should a seizure occur is more of a concern. A specific set of characteristic developmental features associated with anticonvulsant treatment has been classified as Fetal Hydantoin Syndrome. There are craniofacial features, limb alterations, as well as growth and learning deficits. While the structural effects may be mild, the growth and learning retardation are commonly permanent. The syndrome is not particularly common, however, and many studies conclude that the risk is low, especially when the potential for epileptic seizures is considered. In experimental protocols, phenytoin produces more severe, specific craniofacial defects, including cleft palate when given in the window of time associated with palatogenesis. Later exposures produce the limb effects. Valproic acid is clearly teratogenic in experimental protocols as well. The primary effects appear to be on the central nervous system, however, skeletal and craniofacial defects can also be produced. There are reports of neural tube defects and spina bifida in humans exposed to valproic acid through maternal treatment, and the occurrence appears to be higher than expected in some studies. However, most of these studies have selected cases to examine, and it is not clear what the actual incidence rate of valproic acid-related human defects is. Again, the concurrent epilepsy confounds the situation. DES: A Teratogen Associated with Cancer Endpoints Another type of teratogen is exemplified by another human tragedy. Diethylstilbestrol (DES) is a synthetic steroid hormone that was used to help prevent miscarriage in women with difficulty maintaining a pregnancy. In this case, rather than the half decade it took for thalidomide’s effects to become clear, around a quarter of a century passed between the mid-1940s and 1970 before the teratogenicity of DES became clear. An extremely rare form of reproductive tract cancers, clear cell adenocarcinomas, was detected in a series of women whose mothers had taken DES during the first trimester of pregnancy. Though the cancer risk was first noted in some early teenage girls, it peaked around age 19–22. This explains the delay in discovering DES’s effects and illustrates an example where the congenital defect was not immediately obvious. DES is also an interesting example of a teratogen that produces cancer as its congenital defect. It is currently the only established human carcinogen that acts transplacentally. In addition to the cancer risk associated with DES, a variety of other reproductive disorders were noted as the exposed children grew up. Among the female children, these included an increased risk of ectopic pregnancy, spontaneous abortion, menstrual irregularities and infertility. For the male children, abnormalities of the genitals, decreased sperm production, cryptorchidism, and a possible increase in testicular cancer were observed. All of these reproductive tract defects help point out the likely actions of DES in the developing children. The development of the internal and external genitalia is coordinated by steroid hormone production, primarily by the fetal gonads. The hyperestrogen environment produced by DES is consistent with improper formation of the male internal and external



genitalia and may have disrupted the development of the steroid feedback loops in the hypothalamicpituitary-gonadal axis of the females. Additional Human Teratogens Other drugs that are established human teratogens include lithium, tetracyclines, aminopterin, and the coumarin anticoagulants. Antineoplastic agents such as busulfan and cyclophosphamide are teratogenic in addition to their multiple other reproductive toxicities. Androgenic hormones, used to help maintain pregnancies, are also teratogenic and interfere with reproductive development in the fetus, somewhat like the estrogenic DES. With all of these therapeutic agents, the dose resulting in teratogenicity, the issues of leaving the underlying illness untreated, and the actual likelihood of teratogenic effects must be considered in characterizing their toxicological potential. Drugs of Abuse and Maternal Nutrition Some of the drugs of abuse are teratogenic, but their effects typically relate more to generalized metabolic disruptions than to interference with specific features of development. Fetal alcohol syndrome (FAS) is the best example of this class of teratogens. The primary features of FAS include growth retardation, psychomotor dysfunction, and craniofacial anomalies. Growth retardation is the most sensitive and prevalent effect following alcohol consumption during pregnancy. This can be demonstrated at fairly low doses, around 1 drink per day, but the effects at low dose exposure are controversial. At higher doses, 5 or more drinks at a time and at least 2 each day, however, the risk of having an infant classified as small-for-gestational-age increases three-fold. Among children born to chronic alcoholics, one-third have been classified as FAS by some studies. Subsequent development is characterized by reduced height and an inability to catch up during postnatal development. Among children classified as FAS, the rate of mental retardation is 85 percent. Impulsiveness, attention disorders, and language deficits are commonly observed. The mechanism by which FAS is produced is not yet clear. Especially for the chronic alcoholics, it seems most likely to relate primarily to the overall nutritional state, metabolic and endocrine imbalances of the mother. Nutrient deficits and interference with placental metabolism and transport are clearly mechanisms that can affect fetal growth and neural function. Infants born to heroin addicts also exhibit growth retardation that is probably related to overall maternal nutritional and metabolic status. Cigarette smoking by the mother has also been associated with general developmental deficits. Growth retardation of the fetuses of smokers is clear. The array of chemicals in tobacco smoke, however, makes defining the key pathway difficult. It has been suggested that nicotine, carbon monoxide, and cyanide interfere with the transport of amino acids across the placenta. In addition, cadmium is capable of producing placental necrosis and could affect placental exchange. Also, the PAH’s induce metabolic enzymes in the placenta that may create toxic reactive metabolites. Which of these mechanisms actually operate in humans, and which are the major causes of growth retardation is not clear. Maternal deficiencies of specific nutrient factors have been reported to be associated with teratogenicity. Zinc, folic acid, and retinoic acid deficiencies have all been reported to have negative effects on development. It is clear that these nutrients are all required by the fetus, just as they are for any human, but the degree of deficiency that must be reached to cause developmental defects is not clear. Severe deficiencies of any of the vitamins, minerals, or amino acids could reasonably be expected to interfere with development. Declines in zinc, folic acid, and retinoic acid may not be tolerated very well because developmental growth processes are heavily dependent on them. Zinc and folic acid are utilized extensively in metabolism within the rapidly growing tissue. The role of retinoic acid as a major signal molecule during fetal growth has been described above. Methylmercury Poisoning During the 1950s, Minamata disease was described and related to mercury contamination of the Minamata Bay in Japan by industrial facilities. Fish from these coastal waters, a major food source for women in the area, were accumulating elevated methylmercury levels. Mercury levels realized by the women were not high enough to produce obvious signs of mercury



toxicity. However, methylmercury accumulation was sufficient to cause CNS abnormalities in developing fetuses, with cerebral palsy being the most common problem. It was later discovered that mercury levels increased in the placenta and fetal membranes of pregnant women that were exposed to metallic mercury during work. However, these exposures were apparently not sufficient to cause developmental toxicity since there was no increase in spontaneous abortion rates and no defects found in the offspring.

Developmental Toxicity Summary Developmental toxicity can be separated into two categories. Early in pregnancy, the predominant effect of chemical and other stresses is spontaneous abortion. Later, when the specific differentiation of the various organs and structures is taking place, the response to some toxicants is congenital defects of structure or function. Determining whether spontaneous abortions have been caused by a particular chemical exposure is extremely difficult, primarily because there is such a high background rate and so many non-toxicological causes. Detecting a difference in the spontaneous abortion rate within a population is difficult for similar reasons. Most congenital defects are due to inherited or developmental genetic factors rather than teratogenic chemicals. Though in many cases there is a desire to establish whether a defect is the result of external factors, clearly identifying relevant factors and isolating the definitive cause is frequently impossible. From a population perspective, finding a relevant, common factor to ascribe to a certain set of defects is difficult and establishing a causal role for such a factor is even more challenging. It is relevant to consider that the best examples of chemical-induced teratogenesis relate to therapeutic doses of chemicals given with the knowledge and documentation of a health professional. Experimental results have clearly established mechanisms by which both early and later development can be affected by exogenous chemicals. This information is useful in prioritizing investigations of potential human health effects and judging whether reported effects could reasonably be expected

TABLE 11.3 Suspected Human Developmental Toxicants and Teratogens Industrial/Environmental

Pharmaceutical Agents and Drugs

Anesthetic gases (e.g., Halothane) Benzene Cadmium Carbon disulfide Chloroprene Chlorobiphenyls Diethylhexyl phthalate Dioxins Ethylene oxide Lead Methylmercury Toluene

Aminopterin Busulfan Coumarin anticoagulants Cyclophosphamide Diethylstilbestrol (DES) Dimethylformamide Ethanol Isotretinoin and other retinoids Lithium Phenytoin Opiates Tetracyclines Thalidomide Valproic acid

Miscellaneous Nonchemical Agents Cytomegalovirus Diabetes Ionizing radiation Phenylketonuria

Rubella Syphilis Toxoplasmosis



to occur from certain candidate chemicals. However, development is a very finely regulated process that depends on small pools of cells to serve as the starting point for various structures, and the potential of many experimental protocols to interfere with this process is not surprising. The challenge is to decide which of these experimental sources of developmental effects should be of concern to humans. Table 11.3 provides a summarization of the suspected developmental toxicants.

11.4 CURRENT RESEARCH CONCERNS This section will describe some areas of reproductive toxicology that are currently drawing intensive interest from researchers. Some of these research areas are likely to follow a progression into applied toxicology and become issues affecting the future regulatory framework in the United States and, consequently, affect industrial and environmental decisions and concerns. The goal is to point out both exciting areas for investigation and suggest areas of toxicology that may be important in the future for both government and industry.

Endocrine Disruption Since passage of the 1996 Food Quality Protection Act, a law that required EPA to screen pesticides for the ability to produce estrogenic effects in humans, endocrine disruption has moved to the forefront in terms of toxicological research and regulatory controversy. The potential for some chemicals to manifest adverse effects through interactions with the endocrine system has been discussed for male, female, and developmental toxicology in earlier sections of this chapter. Although scientific and regulatory attention is recent, ecologists, agricultural scientists and farmers have known about the practical ramifications of hormonally active agents for many years. A large number of chemicals produced by one organism can affect the hormonal status of other organisms. On the ecological scale, this type of chemical signaling between species is critical to the functioning of certain communities. Farmers have long known that grazing sheep or cattle on rich, new growth clover reduces pregnancies. This affect is due to compounds produced by the clover that can mimic estrogen, called phytoestrogens. Many different classes of chemicals produced by plants or fungi can affect reproduction. One question drawing much toxicological and regulatory attention is whether exposure to synthetic chemicals at levels relevant in the environment or workplace can also have hormonal effects. A second question is whether synthetic chemicals with weak hormonal potency could adversely affect endocrine functioning given that the human diet already contains large amounts of naturally-derived hormonally active agents. A third critical question is whether it is practical to regulate chemicals based on presumed mechanisms of action—i.e., on the basis of a potential endocrine mechanism—rather than on production of adverse effects such as reproductive or developmental impairment. There is little doubt that certain wildlife exposures to high concentrations of synthetic chemicals have produced reproductive and developmental effects. However, scientists disagree as to which chemicals or environmental factors may be responsible, and whether the effects are caused by hormonal mechanisms or by other types of toxicity. There is even greater controversy regarding whether adverse effects are also occurring in humans and other wildlife at lower exposure levels (the so-called endocrine disruptor hypothesis). Proponents of the hypothesis claim support from three main tenets: the effects observed in wildlife inhabiting highly contaminated environments; the chemical similarities among endogenous hormones, naturally occurring hormonally active agents, and certain synthetic chemicals; and, the fact that the endocrine system is responsive to minute levels of hormones. Those who find the hypothesis unsupportable point out different tenets: that synthetic chemicals are much less potent than natural hormones; that the human diet already contains many naturally-occurring hormonally active agents that may actually enhance health; that there is a propensity for weakly hormonal chemical signals to cancel each other by antagonistic actions; and, that the reproductive health of humans in industrialized societies has tended to improve rather than decline over recent decades.



The basis for presuming that exposure to hormonally active agents can lead to significant risks is mechanistically sound and clearly operates in certain high-dose situations, such as the birth control pills and other therapeutic uses of synthetic hormones such as DES. However, there are well-understood biological reasons to expect that the characteristics of the endocrine response differ under dramatically different levels of stimulation (i.e., dose). Since doses of DES prescribed during the first three months of pregnancy are equivalent to more than 150 years worth of a woman’s natural estrogen production, this example is probably not relevant to potential risks from low levels of weakly estrogenic environmental contaminants. It is important to recognize that the potency of “ environmental estrogens” typically range from hundreds to millions of times less than estradiol itself. Although there are factors that might tend to reduce these potency differences, such as binding to serum proteins, these factors are insufficient to answer the serious questions raised as to whether synthetic chemicals could affect endocrine signaling in humans at realistic exposure levels. Other putative developmental effects of hormonally active agents in humans derive from environmental exposures, some of which occurred in accidental, high-dose poisoning incidents. Prenatal and postnatal exposure to PCBs and polychlorinated dibenzofurans (PCDFs) in high-dose accidental poisonings from contaminated rice oil in Yusho, Japan and Yucheng, Taiwan have resulted in various developmental defects. The syndrome of effects includes low birth weight, dark pigmentation of the skin and mucous membranes, gingival hyperplasia, exophthalmic edematous eyes, dentition at birth, abnormal calcification of the skull, rocker bottom heel, and low birth weight. Most of the affected infants were found to be shorter and had less total lean mass and soft-tissue mass. Follow-up studies on poisoned individuals suggest neurobehavioral effects and cognitive deficits. Gross developmental defects have not been observed in populations exposed at lower levels, but proponents of the endocrine disruptor hypothesis point out that typical body burdens of PCB’s and dioxins are relatively close to the levels measured in epidemiological study groups where adverse effects on IQ and neuromuscular development have been reported. Nonetheless, these body burdens of PCBs have not been shown to cause any specific adverse effects, and the overall epidemiological evidence is equivocal and does not support a causal association between typical body burdens of PCBs and adverse developmental outcomes. Another highly publicized putative consequence of endocrine disruption in humans is reduced sperm counts in men living in industrialized nations. Increasing background levels of a variety of persistent, estrogenic environmental chemicals have been identified as a potential cause. This theory has some mechanistic plausibility because sperm production is controlled by androgen levels, and some effects of androgens can be antagonized by estrogens. This theory also has some high-dose precedent from DES, a potent estrogenic compound that may have reduced sperm counts among males exposed to therapeutic levels in utero. However, several of the studies that report declining sperm counts have been criticized for methodological flaws, failing to account for alternative factors, and biases in data collection. The statistical tests used and the proper interpretation of the tests have also been called into question. Based on these difficulties and criticisms, many scientists question whether sperm counts have actually declined in men from industrialized nations. It is important to recognize that suggested reduction in sperm numbers is not the type of readily apparent pathological condition observed in DBCP manufacturers (discussed above under Male Reproductive Toxicology). The sperm count decline suggested by some authors (up to around 50 percent) would not be expected to correspond to a general fertility reduction because of the large excess of sperm that are produced by most men. Numerous methodological difficulties arise in evaluating sperm counts from different laboratories over a long time frame, and hence, the degree of change that is purportedly related to environmental exposures may be too subtle to be easily measured. Collection and preparation methods have varied over the years, and differing criteria have been used to categorize typical, or “ normal” sperm counts. Different studies have handled samples from possibly infertile patients differently, some studies including and others excluding them according to differing criteria. These differences in methodology have confounded attempts to combine the results into a larger database for integrated analysis. For



these reasons, many scientists are convinced that reported sperm count declines are an artifact of methodological and analytical flaws of the studies. Many synthetic chemicals have been suggested as potential human endocrine disruptors based upon widespread human exposure and their hormone-like activity in certain laboratory assays. Various lists of putative endocrine disruptors have been published or otherwise publicized in the media or on the internet. It is important to recognize that the quality of data supporting inclusion of chemicals on these lists varies considerably, and there is no generally accepted scientific source providing an authoritative listing at this time. Most lists include chemicals from diverse chemical classes, many of which have produced a positive result in at least one of a variety of bioassays and receptor-binding methods devised to determine the potential interaction of a chemical with the endocrine system. Despite positive results in laboratory assays, few chemicals—e.g., those drugs and chemicals already discussed in this section—have been shown to produce adverse developmental outcomes in exposed humans. Some prominent examples of chemicals listed as endocrine disruptors include organochlorine pesticides (e.g., toxaphane, methoxychlor, chlordecone, DDT and metabolites), alkylphenol ethoxylates (detergents or dispersing agents in household cleaners), PAHs (combustion products) dioxins (TCDD), co-planar PCBs, phthalate and phenolic plasticizers (e.g., benzyl butyl phthalate, di-n-butyl phthalate, bisphenol A). However, more definitive laboratory studies and risk assessments developed for a number of such chemicals (e.g., alkylphenol ethoxylates, phthalate and phenolic plasticizers) indicate little or no potential for adverse effects in humans at environmentally relevant exposure levels. Two particular issues have arisen in the controversy over endocrine disruption that deserve special mention. In 1996, just months before Congress passed the 1996 Food Quality Protection Act, Arnold and coworkers published a paper in Science that brought national attention to the subject of endocrine disruption. The report claimed that a combination of synthetic chlorinated pesticides were one-thousand times more potent than any of the chemicals individually in stimulating an estrogenic response. This so-called demonstration of estrogenic synergism was later shown to be in error, and the publication was retracted more than a year later. Despite its failure to demonstrate synergy, this study raised a debate within the scientific, regulatory, and regulated communities over the frequency with which synergistic interactions are likely to occur and their relevance to human and environmental health. Though the interest in synergy has subsided considerably since the retraction of the Arnold publication, a considerable amount of effort is still underway to determine whether such chemical interactions are important considerations for risk assessment. The second issue of debate involves the dose-response function for endocrine active agents. First, is there a threshold for endocrine-mediated adverse effects and second, do toxic effects of high doses of hormonally active agents mask more subtle adverse effects that can only be detected at low doses using specialized assay systems? These issues arise from two publications suggesting that very low doses of plasticizing agents could produce subtle effects on the developing male reproductive tract not seen at higher doses, possibly because subtle effects are masked by more overt toxicity at higher doses. Neither study has been replicated, despite attempts that employed more comprehensive study designs. Nonetheless, the issue has lead to an outcry from consumer and environmental activist groups to cease the use of certain plastics in baby bottles and childrens’ toys. Former Surgeon General of the United States Dr. C. Everett Koop has responded, calling this reaction irresponsible. In summary, a number of critical questions have been raised with respect to the identification of hormonally active agents in general, and laboratory studies that purport to demonstrate potential hormonal activity in particular.

• Are positive results in short-term in vivo and in vitro laboratory assays predictive of adverse health effects in humans?

• Can measurements of hormonal potency in laboratory assays be extrapolated to human populations at environmentally relevant exposure levels?



• Given that the human diet contains high amounts of naturally-derived hormonally active agents, is it feasible that synthetic chemicals with weak hormonal potency could adversely affect human endocrine functioning?

• Do the dose-response curves of hormonally active agents lack a threshold for adverse effects? • Do toxic effects of high doses of hormonally active agents mask more subtle adverse effects that can only be detected at low doses using specialized assay systems?

• Are hormonally active agents more prone to exhibiting interactive effects (synergism or antagonism) than chemicals that operate through other mechanisms?

• Is it practical to regulate chemicals based on presumed mechanisms of action—i.e., on the basis of a potential endocrine mechanism—rather than on production of adverse effects such as reproductive or developmental impairment? The way that the scientific and regulatory communities answer these questions could have a profound impact on the risk assessment of hormonally active agents in the workplace and in the environment.

Lead Poisoning and the Lowering of the Threshold Currently, a hot area of research is the sensitivity of the developing nervous system to low-dose lead exposure. Lead toxicity is apparent in a variety of organ systems. As mentioned above, lead effects on both male and female reproduction have been investigated and the use of lead salts for inducing abortion reaches back to antiquity. The neurological system is recognized as one of the key targets for toxic responses to lead. Some reports have recently suggested that the levels of environmental lead exposures received by large populations, especially in urban areas, could be sufficient to produce adverse cognitive effects. This has lead to substantial investigation of both lead toxicity mechanisms in animals and the occurrence of cognitive deficits in children. Though reports of low-dose lead effects have struck parental and societal chords, the body of research on intelligence and cognitive outcomes does not support a consistent association with today’s common levels of environmental lead exposure. Rather than the traditional applied dose, lead exposure is typically considered on the basis of a measured blood level. There is little dispute about the potential for lead toxicity in children when chronic blood levels reach the 30–50 µg/dl range or higher. A standard regulatory criterion of concern is 10 µg/dl. However, there are suggestions that cognitive effects may accrue even at this threshold, or perhaps even up to 10-fold lower. Unfortunately, the endpoints of intelligence and verbal ability that have been suggested as the most sensitive indicators are exceedingly difficult to measure in a repeatable, reliable, and objective manner. A further complication is the considerable plasticity in learning processes and the ability of children to “ make up ground” as they develop. Scientific arguments rage over the verbal abilities of two-year olds and the meaning of IQ differences of less than one or two points on the typical scale. Research has suggested that verbal development is a brain function particularly vulnerable to lead. However, despite claims of statistical significance in some studies, the uncertainty associated with evaluating these endpoints, which is not captured statistically, clearly makes definitive conclusions impossible. The testing methods for assessing cognitive development and verbal ability in infants and toddlers are not generally regarded as sensitive enough to reliably distinguish between inter-individual variability and exposure-associated effects at the required levels. However, information from animal studies has begun to shed light on mechanisms by which lead could affect brain development. There does appear to be a heightened sensitivity of fetal and neonatal brain cells to lead effects compared to adults. This may relate to the much more active process of forming connections among neural cells and expansion of vascular, blood carrying elements during fetal and neonatal stages. It is not clear what degree of change in this process must occur to represent an adverse reaction to lead, however, since there is considerable variation and plasticity in the process anyway.



Two important areas for additional investigation are: 1) developing better tools for investigating human cognitive function and abilities and 2) characterizing the relationship between the doseresponse characteristics of experimental animals and that of humans for lead. These questions are important in terms of both public health and economics. In general, the scientific and regulatory communities have regarded the clear and dramatic drop in children’s blood lead levels since the 1970s as a real public health improvement realized through the control of lead from gasoline and paints. If neurocognitive development turns out to be as sensitive as some suggest to the effects of lead, much tougher questions about whether and how to address exposures, down to the range associated with naturally occurring lead, will be up for consideration. Without obvious and readily replaceable major exposure sources, like gasoline or paint, the costs associated with additional incremental reductions in lead exposure for the population as a whole may be dramatic.

11.5 SUMMARY This chapter has outlined the toxic responses of the male and female reproductive systems and the developing fetus. Some of the mechanisms of toxicity, generally described using experimental toxicants, have been presented to illustrate the types of responses and effects that should be considered. In most cases, however, the experimental toxicants have limited direct application to human health effects. Especially for occupational exposures, the gap between toxic potential and demonstrated effects is large. Examples of actual human reproductive and developmental toxicants have been pointed out so that those chemicals, which are currently known to represent a risk to humans, can be identified. Some of the key points in the chapter included:

• The differential sensitivity of various tissues and cell types in the male and female repro• • • •

ductive organs to certain types of toxicants. The functional and toxicological implications of the different patterns of cellular division and germ cell maturation used by males and females. The multiple interactions between the reproductive and endocrine systems and the balance of endocrine regulation that may be vulnerable during certain toxic responses. The relationship of the sequential course of developmental processes to toxic responses. The major difference in toxic responses between the embryonic and fetal periods of development.

REFERENCES AND SUGGESTED READING Alvarez, J. G., and B. T. Storey, “ Evidence for increased lipid peroxidative damage and loss of superoxide dismutase as a mode of sublethal damage to human sperm during cryopreservation.” Jo. of Androl. 13: 232–241 (1992). Arnold, S. F., D. M. Klotz, B. M. Collins, P. M. Vonier, L. J. Jr., Guillette, and J. A. McLachlan, Synergistic activation of estrogen receptor with combinations of environmental chemicals [see comments] [retracted by McLachlan JA. In: Science 1997 Jul 25; 277(5325):462–463], Science, 1996; 272: 1489–1492. Ashby, J., J. Odum, H. Tinwell, and P. A. Lefevre, “ Assessing the risks of adverse endocrine-mediated effects: where to from here?” Regulatory Toxicology and Pharmacology 26: 80–93 (1997). Ashby, J., H. Tinwell, P. A. Lefevre, J. Odum, D. Paton, S. W. Millward, S. Tittensor, and A. N. Brooks, “ Normal sexual development of rats exposed to butyl benzyl phthalate from conception to weaning.” Regulatory Toxicology and Pharmacology 26(1 Pt 1):102–118 (1997). Auger, J., J. M. Kunstmann, F. Czyglik, and P. Jouannet, “ Decline in semen quality among fertile men in Paris during the past 50 years.” New England Journal of Medicine 332: 281–285 (1995). Bromwich, P., J. Cohen, I. Stewart, and A. Walker, “ Decline in sperm counts: and artefact or changed reference range of ‘normal’?” British Medical Journal 309: 19–22 (1994).



Cagen, S. Z., J. M. Jr., Waechter, S. S. Dimond, W. J. Breslin, J. H. Butala, F. W. Jekat, R. L. Joiner, R. N. Shiotsuka, G. E. Veenstra, and L. R. Harris, “ Normal reproductive organ development in CF-1 mice following prenatal exposure to bisphenol A.” Toxicological Sciences 50(1): 36–44 (1999). Carney, E. W., A. M. Hoberman, D. R. Farmer, R. W. Jr., Kapp, A. I. Nikiforov, M. Bernstein, M. E. Hurtt, W. J. Breslin, S. Z. Cagen, and G. P. Daston. “ Estrogen modulation: tiered testing for human hazard evaluation.” American Industrial Health Council, Reproductive and Developmental Effects Subcommittee. Reproductive Toxicology 11(6): 879–892 (1997). Colborn, T., F. S. Vom Saal, and A. M. Soto, “ Developmental effects of endocrine-disrupting chemicals in wildlife and humans.” Environmental Health Perspectives 101: 378–384 (1993). Crisp, T. M., E. M. Clegg, R. L. Cooper, W. P. Wood, D. G. Anderson, K. P. Baetcke, J. L. Hoffman, M. S. Morrow, D. J. Rodier, J. E. Schaeffer, L. W. Touart, M. G. Zeeman, and Y. M. Patel, “ Environmental endocrine disruption: An effects assessment and analysis.” Environmental Health Perspectives 106 (Supplement 1): 11–56 (1998). Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC), “ Endocrine Disruptor Screening and Testing Advisory Commitee (EDSTAC) Final Report.” Washington, D.C. USEPA, editor, (1998). Faber, K. A., and C. L., Jr., Hughes, ” Clinical Aspects of Reproductive Toxicology” in Witorsch, R. J., ed., Reproductive Toxicology. 2nd edition. New York: Raven Press, Ltd, (1995). Gorospe, W. C., and M. Reinhard, “ Toxic Effects on the Ovary of the Nonpregnant Female.” in Witorsch, R. J., ed., Reproductive Toxicology. 2nd edition. New York: Raven Press, Ltd, (1995). Koop, C. E., “ The Latest Phoney Chemical Scare.” The Wall Street Journal, June 22, 1999. Manson, J. M., and L. D. Wise, “ Teratogens.” in Amdur, M. O., Doull, J., and Klaassen C. D., eds., Casarett and Doull’s Toxicology: The Basic Science of Poisons. 4th edition. New York: Pergamon Press (1991). Matt, D. W., and J. F. Borzelleca, “ Toxic Effects on the Female Reproductive System During Pregnancy, Parturition, and Lactation.” in Witorsch, R. J., ed., Reproductive Toxicology. 2nd edition. New York: Raven Press, Ltd, (1995). Mattison, D. R., D. R. Plowchalk, M. J. Meadows, A. Z. Al-Juburi, J. Gandy, and A. Malek, “ Reproductive Toxicity: Male and Female Reproductive Systems as Targets for Chemical Injury.” Medical Clinics of North America 74: 391–411 (1990). McLachlan, J. A., Retraction: Synergistic activation of estrogen receptor with combinations of environmental chemicals, Science 277: 462–463 (1997). NagDas, S. K. “ Effect of chlorpromazine on bovine sperm respiration.” Archives of Andrology 28: 195–200 (1992). Nair, R. S., F. W. Jekat, D. H. Waalkens-Berendsen, R. Eiben, R. A. Barter, and M. A. Martens, “ Lack of Developmental/Reproductive Effects with Low Concentrations of Butyl Benzyl Phthalate in Drinking Water in Rats.” The Toxicologist, 48(1-S): 218 (1999). National Research Council, Committee on Hormonally Active Agents in the Environment, Board on Environmental Studies and Toxicology, Commission on Life Sciences, 1999. “ Hormonally Active Agents in the Environment.” National Academy Press, Washington. Nimrod, A. C. and W. H. Benson, “ Environmental estrogenic effects of alkylphenol ethoxylates.” Critical Reviews in Toxicology 26: 335–364 (1996). Olsen, G. W., K. M. Bodner, J. M. Ramlow, C. E. Ross, and L. I. Lipshultz, “ Have sperm counts been reduced 50 percent in 50 years? A statistical model revisited.” Fertility and Sterility 63: 887–893 (1995). Peltola, V., E. Mantyla, I. Huhtaniemi, and M. Ahotupa, “ Lipid peroxidation and antioxidant enzyme activities in the rat testis after cigarette smoke inhalation or administration of polychlorinated biphenyls or polychlorinated naphthalenes.” Jo. of Androl. 15: 353–361 (1994). Safe, S. H., “ Do environmental estrogens play a role in development of breast cancer in women and male reproductive problems?” Human and Ecological Risk Assessment 1: 17–23 (1995). Schardein, J. L. Chemically Induced Birth Defects. New York: Marcel Dekker, Inc (1985). Schilling, K., C. Gembardt, and J. Hellwig, “ Reproduction toxicity of di-2-ethylhexyl phthalate (DEHP)” The Toxicologist, 48; (1-S): 692 (1985, 1999). Sharpe, R. M., J. S. Fisher, M. M. Millar, S. Jobling, and J. P. Sumpter, “ Gestational and lactational exposure of rats to xenoestrogens results in reduced testicular size and sperm production.” Environmental Health Perspectives 103(12): 1136–1143 (1995). Shepard, T. H., Catalog of Teratogenic Agents. 6th edition. Baltimore: Johns Hopkins University Press (1989).



Sundaram, K., and R. J. Witorsch, “ Toxic Effects on the Testes.” in Witorsch, R. J., ed., Reproductive Toxicology. 2nd edition. New York: Raven Press, Ltd, (1995). Thomas, J. A. “ Toxic Responses of the Reproductive System.” In Amdur, M. O., Doull, J., and Klaassen, C. D., eds., Casarett and Doull’s Toxicology: The Basic Science of Poisons. 4th edition. New York: Pergamon Press (1991). vom Saal, F. S., B. G. Timms, M. M. Montano, P. Palanza, K. A. Thayer, S. C. Nagel, M. D. Dhar, V. K. Ganjam, S. Parmigiani, and W. V. Welshons, “ Prostate enlargement in mice due to fetal exposure to low doses of estradiol or diethylstilbestrol and opposite effects at high doses.” Proceedings of the National Academy of Sciences 94(5): 2056–2061 (1997).

12 Mutagenesis and Genetic Toxicology MUTAGENESIS AND GENETIC TOXICOLOGY


Genetic toxicology combines the study of physically or chemically induced changes in the hereditary material (deoxyribonucleic acid or DNA) with the prediction and the prevention of potential adverse effects. Modification of the human genetic material by chemical agents or physical agents (e.g., radiation) represents one of the most serious potential consequences of exposure to toxicants in the environment or the workplace. Nevertheless, despite increasing research interest in this area, the number of agents or processes that are known to cause such changes is quite limited. This chapter presents information regarding the following areas:

• Types and characteristics of genetic alteration • Common research methods for the assessment of genetic change • Practical significance of test results from animal and human studies in the identification of potential mutagens

• Theoretical relationships between mutagenesis and carcinogenesis 12.1 INDUCTION AND POTENTIAL CONSEQUENCES OF GENETIC CHANGE

Historical Perspective The term mutation is defined as a transmissible change in the genetic material of an organism. This actual heritable change in the genetic constitution of a cell or an individual is referred to as a genotypic change because the genetic material has been altered. While all mutational changes result in alteration of the genetic material in the parent cells, not all are immediately expressed in cell progeny as functional, or phenotypic, changes. Thus, it is possible to have genetic change that is not associated with a transmissible change. These distinctions are discussed in greater detail in subsequent sections. Potential environmental and occupational mutagens may be classified as physical, biological, or chemical agents. Ames and many subsequent researchers have identified representative chemical mutagens in at least 10 classes of compounds, including the following: cyclic aromatics, ethers, halogenated aliphatics, nitrosamines, selected pesticides, phthalate esters, selected phenols, selected polychlorinated biphenyls, and selected polycyclic aromatics (PAHs). Despite nearly 50 years of research concerning chemical-induced genetic change, ionizing radiation still represents the best described example of a dose-dependent mutagen and was first demonstrated in the 1920s. Chemical mutagenesis was first demonstrated in the 1940s, and many of the characteristics of radiation-induced mutation are believed to be common to chemically induced mutation. This is particularly true for molecules known as free radicals, which are formed in radiation events and some chemical toxic events. Radicals contain unpaired electrons, are strongly electrophilic, and extraordinarily reactive, features that are well correlated with both mutagenic and carcinogenic potency. Such reactive molecules Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




probably are responsible for at least some of the alterations of nucleic acid sequences that are observed in genotoxic processes. Over 3500 functional disorders or disease states have been linked to heritable changes in humans, and the ambient incidence of genetic disease may be as great as 10 percent in newborns. In the case of some cancers, a change in the genotype of a cell results in a change in phenotype that is grossly defined by rapid cellular division and a reversion of the cell to a less specialized type (dedifferentiation). The subsequent generations eventually may form a growing tumor mass within the affected tissue. This simplified sequence has been termed the somatic cell mutation theory of cancer. While not all chemically-induced cancers can be explained by this hypothesis, general applicability of the somatic cell mutation theory is supported by the following points:

• Most demonstrated chemical mutagens are demonstrably carcinogenic in animal studies • Carcinogen-DNA complexes (adducts) often can be isolated from carcinogen-treated cells • Heritable defects in DNA-repair capability, such as in the sunlight-induced disease • • •

xeroderma pigmentosum, predispose affected individuals to cancer Tumor cells can be “ initiated” by carcinogens but may remain in a dormant state for many cell generations, an observation consistent with permanent DNA structural changes Cancer cells generally display chromosomal abnormalities Cancers display altered gene expression (i.e., a phenotypic change)

The issue of correlation between genotoxicity or mutagenicity assays and cancer is discussed in greater detail in subsequent sections of this chapter. Although genetic changes in somatic cells are of concern because consequences such as cancer may be debilitating or lethal, mutational changes in germ cells (sperm or ovum) may have even more serious consequences because of the potential for effects on subsequent human generations. If a lethal and dominant mutation occurs in a germinal cell, the result is a nonviable offspring, and the change is not transmissible. On the other hand, a dominant but viable mutation can be transmitted to the next generation, and it need only be present in single form (heterozygous) to be expressed in the phenotype of the individual. If the phenotypic change confers evolutionary disadvantage to the individual (e.g., renders it less fit), it is unlikely to become established in the gene pool. In contrast, individuals that are heterozygous for recessive genes represent unaffected carriers that are essentially impossible to detect in a population. Thus, recessive mutations are of the greatest potential concern. These mutations may cause effects ranging from minor to lethal whenever two heterozygous carriers produce an offspring that is homozygous for the recessive trait (i.e., the genes are present in both copies). Figure 12.1 describes the potential consequences of mutagenic events. Occupational Mutagens, Spontaneous Mutations, and Naturally Occurring Mutagens In considering the potential adverse effects of chemicals, it is important to recognize that both physical and chemical mutagens occur naturally in the environment. Radiation is an ubiquitous feature of our lives, sunlight representing the most obvious example. Incomplete combustion produces mutagens such as benzo[a]pyrene, and some mutagens occur naturally in the diet, or may be formed during normal cooking or food processing (e.g., nitrosamines). In addition, drinking water and swimming-pool water have been shown to contain potential mutagens that are formed during chlorination procedures. Thus, the genetic events that influence the human evolutionary process appropriately may be viewed as a combination of normal background incidence of spontaneous mutations that may be occurring during cellular division, coupled with the exposure to naturally occurring chemical or physical mutagens. Mutagenic chemicals in the workplace, or those that are introduced into the environment via industrial operations, represent a potential contribution to the genetic burden, though the practical significance of this contribution is not known with precision. It is estimated that over 70,000 synthetic



Figure 12.1 Possible consequences of mutagenic event in somatic and germinal cells.

organic compounds currently are in use, a number which increases annually. Only a very small fraction of these have been confirmed as human carcinogens (see Chapter 13), and no compound has been shown unequivocally to be mutagenic in humans. However, animal and bacterial tests have demonstrated a mutagenic potential for some occupational and environmental compounds at high exposure levels, and it is reasonable to consider human exposure to these compounds, particularly in occupational situations where contact may be frequent and/or intense. This is not to suggest that very small exposures common to environmental circumstances are likely to be associated with adverse effects.


Transcription and Translation DNA (deoxyribonucleic acid) is the structural and biochemical unit on which heredity and genetics are based for all species. It is the only cellular macromolecule that is self-replicating, alterable, and transmissible. Subunits of the DNA molecule are grouped into genes that contain the information, which is necessary to produce a cellular product. An example of such a cellular product is a polypeptide or protein, which may have a structural, enzymatic, or regulatory function in the organism. Figure 12.2 illustrates how the sequence of messages on the DNA molecule is transcribed into the RNA (ribonucleic acid) molecule and ultimately is translated into the polypeptide or protein. The sequence of base pairs in the DNA molecule specifies the appropriate complementary (“ mirror image” ) sequence that governs the formation of the messenger RNA (mRNA). Transfer RNAs (tRNA), each of which is specific for a single amino acid, are matched to the appropriate segment of the mRNA. When the amino acids are released from the tRNAs and are linked in a continuous string, the polypeptide (or protein) chain is formed. Recognition of the mRNA regions by the tRNA-amino acid complex is accomplished by a system of triplet, or three-base, codons (in the mRNA) and complementary anticodons (in the tRNA). The critical features of this coding system are that it is simultaneously unambiguous and degenerate. In



Figure 12.2 Schematic representation of transcription and translation.

other words, no triplet codon may call for more than a single specific tRNA-amino acid complex (unambiguous), but several triplets may call for the same tRNA-amino acid (degenerate). This results from the fact that four nucleotides, which form DNA (DNA is composed of adenine, cytosine, guanine, and thymine), and the nucleotides forming RNA (RNA is made up of A, C, G, and uracil) may be combined in triplet form in 64 different ways (4 × 4 × 4 or 43). The 20 amino acids and three terminal codes account for less than half of the available codons, leaving well over 30 codons of the possible 64. The biological significance of this degeneracy is that such a characteristic minimizes the influence of minor mutations (e.g., single basepair deletions or additions) because codons differing only in minor aspects may still code for the same amino acids. The significance of having an unambiguous code is clear; the formation of proteins must be perfectly reproducible and exact. Table 12.1 depicts the amino acids that are coded for by the various triplet codons of DNA, as well as the initiation and termination signal triplets. The process of mutagenesis results from an alteration in the DNA sequence. If the alteration is not too radical, the rearrangement may be transmitted faithfully through the mRNA to protein synthesis,


Leucine Leucine Leucine Leucine Valine Valine Valine Valine Phenylalanine Phenylalanine Leucine Leucine




Serine Serine Serine Serine

Alanine Alanine Alanine Alanine

Proline Proline Proline Proline

Threonine Threonine Threonine Threonine


Tyrosine Tyrosine STOP STOP

Aspartate Aspartate Glutamate Glutamate

Histidine Histidine Glutamate Glutamate

Asparagine Asparagine Lysine Lysine


Cysteine Cysteine STOP Tryptophan

Glycine Glycine Glycine Glycine

Arginine Arginine Arginine Arginine

Serine Serine Arginine Arginine

(Video) Toxicology (Part-01)Principle of Toxicology = General Terminology | Management of Poisonings






Third position in triplet

*The sequence AUG, in addition to coding for methionine, is part of the initiator sequence that starts the translation process by which mRNA is formed from the DNA template.

Isoleucine Isoleucine Isoleucine *Methionine



First position in triplet

Second position in triplet

TABLE 12-1. Correspondence of the Genetic Code with the Appropriate Amino Acids (Note Unambiguity and Degeneracy)



which results in a gene product that is partially or completely unable to perform its normal function. Such changes may be correlated with carcinogenesis, fetal death, fetal malformation, or biochemical dysfunction, depending on the cell type that has been affected. However, cause and effect relationships for such correlations typically are lacking. Initiation and termination of DNA transcription are controlled by a separate set of regulatory genes. Most regulatory genes respond to chemical cues, so that only those genes that are needed at a given time are expressed or available. The remaining genes are in an inactive state. The processes of gene activation and inactivation are believed to be critical to cellular differentiation, and interruption of these processes may result in the expression of abnormal conditions such as tumors. This represents an example of a case in which a non-genetic event may result in tumorigenesis. Oncogenes represent an example of a situation where activation of a genetic phenomenon may act to initiate carcinogenicity. In contrast, loss of “ tumor suppressor” genes may, by omission, result in initiation of the carcinogenic process. Chromosome Structure and Function The DNA of mammalian species, including humans, is packaged in combination with specialized proteins (predominantly histones) into units termed chromosomes, which are found in the nucleus of the cell. The proteins are thought to “ cover” certain segments of the DNA and may act as inhibitors of expression for some regions. Each normal human cell (except germ cells) contains 46 chromosomes (23 pairs). Chromosomes may be present singly (haploid), as in germ cells (sperm or ovum), or in pairs (diploid), as in somatic cells or in fertilized ova. In haploid cells, all functional genes present in the cell can be expressed. In diploid cells, one allele may be dominant over the other, and in this case, only the dominant gene of each functional pair is expressed. The unexpressed allele is termed recessive, and recessive genes are expressed only when both copies of the recessive type are present. Some cell types in mammals are found in forms other than diploid. Functionally normal liver cells, for example, are occasionally found to be tetraploid (two chromosome pairs instead of one pair). Some features and terminology that are important to cytogenetics, or the study of chromosomes, include:

• Karyotype—the array of chromosomes, typically taken at the point in the cell cycle known • • • • •

as metaphase, which is unique to a species and forms the basis for cellular taxonomy; may be used to detect physical or chemical damage Centromere—the primary constriction, which represents the site of attachment of the spindle fiber during cell division; useful in identifying specific chromosomes, as its location is relatively consistent Nucleolar organizing region—the secondary constriction, which represents the site of synthesis of RNA, subsequently used in ribosomes for protein synthesis Satellite—the segment terminal to the nucleolar organizing region; useful in specific chromosome identification Heterochromatin—tight-coiling region; relatively inactive Euchromatin—loose-coiling region; primary transcription site

Mitosis, Meiosis, and Fertilization The process by which a somatic cell divides into two diploid daughter cells is called mitosis. The first stage of mitosis is called prophase, during which the spindle is formed and the chromatin material (DNA and protein) of the nucleus becomes shortened into well-defined chromosomes. During metaphase, the centriole pairs are pulled tightly by the attached microtubules to the very center of the cell, lining up in the equatorial plane of the mitotic spindle. With still further growth of the spindle, the chromatids in each pair of chromosomes are broken apart, a stage called anaphase. All 46 pairs



(in humans) of chromatids are thus separated, forming pairs of daughter chromosomes that are pulled toward one mitotic spindle or the other. In telophase, the mitotic spindle grows longer, completing the separation of daughter chromosomes. A new nuclear membrane is formed, and shortly thereafter the cell constricts at the midpoint between the two nuclei, forming two new cells. Meiosis is the term for the process by which immature germ cells produce gametes (sperm or ova) that are haploid. During meiosis, DNA is replicated, producing 46 chromosomes with sister chromatids. The 46 chromosomes arrange into 23 pairs at the center of the nucleus, and in the first division the pairs separate. In a second division, the sister chromatids separate, with one chromosome of each pair being incorporated into four gametes. At the time of fertilization, or zygote formation, the fusion of gametes once again forms a cell with a full complement of 46 chromosomes.

Genetic Alteration Tests for genotoxicity in higher organisms may be placed into one of three basic categories: gene mutation tests, chromosomal aberration tests, and DNA damage tests. These tests are conducted individually or in combination to identify various types of mutagenic events (Figure 12.3) or other genotoxic effects. For the purpose of this discussion, the principles of each test category will be reviewed and specific tests will be discussed by broad phylogenetic classifications. Over 200 individual test methods have been developed to assess the extent and magnitude of genetic alteration; however, less than 20 have been validated or are in common use. Numerous mutagenic agents have the demonstrated capacity to cause genetic change in one or more of these test systems, but no well-documented cases of human mutation are available. This latter conclusion may change as a result of improvements in the ability to detect human genetic change. Nevertheless, as discussed in this section, use of a reasonable battery of tests is capable of identifying almost all of the known human carcinogens, consistent with the hypothesis that somatic cell mutations are, at least in part, responsible for a large proportion of human cancers. A transmissible change in the linear sequence of DNA can result from any one of three basic events:

• Infidelity in DNA replication • Point mutation • Chromosomal aberration

Figure 12.3 Types of mutagenic changes (Adapted from Brusick, 1980).



One possibility, infidelity or inexact copying of a DNA strand during normal cellular replication, may result from inaccurate initiation of replication, failure of the transcription enzymes to accurately “ read” the DNA, or interruption of the transcription process by agents that interpose (intercalate) themselves within the DNA molecule or between the DNA and an enzyme. Point mutations, as the second possibility, may be subdivided into basepair changes and frameshift mutations (Figure 12.4). The former result from transition or transversion of DNA base pairs so that

Figure 12.4 Schematic representation of point mutations (frameshift and basepair changes).



the number of bases is unchanged but the sequence is altered. Because the genetic code is “ degenerate,” this may or may not result in an altered product after transcription and translation. A frameshift mutation, however, results from insertion or deletion of one or more bases from the linear sequence of the DNA. This causes the transcription process to be displaced by the corresponding number of bases and virtually assures an altered genetic product. Proflavine, which has been used as a bacteriostatic agent, is an example of a compound that intercalates within the DNA molecule. It is a flat, planar molecule and inserts itself neatly between the bases. When it intercalates, it forces the DNA strand out of its normal configuration, so that when the replication enzymes or transcription enzymes try to read the bases, the bases are not spatially arranged the normal way, and the enzymes cannot read the base sequence properly. The enzymes may skip over one or several bases, or may put an additional base into the DNA or RNA strand at random. Proflavine does not chemically bind with the bases in DNA. In contrast, many of the environmentally prevalent polynuclear aromatic hydrocarbons (PAHs) may intercalate into the DNA, leading to frameshift, and also may chemically react directly with it, an event that can lead to basepair substitution. An example of this is benzo(a)pyrene (BaP), which is found at low concentrations throughout the environment as a product of combustion of fossil fuels, in grilled steaks, tobacco smoke, and many other places. BaP by itself is seldom considered to be mutagenic. However, after metabolism, many highly reactive epoxide intermediate metabolites are formed, one of which (BPDE I) is highly mutagenic. BPDE I combines with guanine to form what is called a DNA adduct. These adducts have been found in extremely small quantities by highly specialized and sensitive techniques such as enzyme-linked immunosorbent assay (ELISA) and fluorescence. A scheme of activation and adduct formation for BaP is given in Figure 12.5. Basepair changes, described earlier, are of two kinds: transitions or transversions. In transitions, one base is replaced by the base of the same chemical class. That is, a purine is replaced by the other purine (e.g., adenine is replaced by guanine); in the case of pyrimidine bases, cytosine would be replaced with thymine or vice versa. An example of a chemical that causes transitions is nitrous acid (see Figure 12.6). Nitrous acid is formed from organic precursors such as nitrosamines, nitrite, and nitrate salts. It reacts with amino (NH2) groups in nucleotides and converts them to keto (C?O) groups. In transversions, a base pair is replaced in the DNA strand by a base of the other type: a purine is replaced by a pyrimidine or vice versa. Another group of chemicals that can cause mutations are alkylating agents. Some well-known alkylating agents are the mustard gases, originally developed for chemical warfare. Chemicals in this group add short carbon–hydrogen chains at specific locations on bases. The experimental agent ethyl methanesulfonate (EMS) can alkylate guanine to form 7-ethylguanine (see Figure 12.7), which can cause the bond between the base and deoxyribose in the backbone of the DNA strand to become unstable and break. This leads to a gap in the DNA strand which, if unrepaired at the time of DNA replication, is filled with any of the four available bases. Not all point mutations are caused by radiation or chemicals; some may occur because of the nature of the bases themselves. The bases have their preferred arrangement of hydrogen atoms, but on rare occasions undergo rearrangements of the hydrogen atoms, called tautomeric shifts. The nitrogen atoms attached to the purine and pyrimidine rings are usually in the amino (NH2) form and only rarely assume the imino (NH) form. Similarly, the oxygen atoms attached to the carbon atoms of guanine and thymine are normally arranged in the keto (C?O) form, but rarely rearrange to the enol (COH) form. The changes in configuration lead to different hydrogen bonding patterns, and, if a base is in the alternate form during replication, a wrong base can be put into the new growing strand leading to a mutation. A group of chemicals, base analogs, that resemble the normal bases of DNA may lead to mutations by being incorporated into DNA inadvertently during repair or replication. These chemicals go through tautomeric shifts more often and result in inappropriate base pairing during replication so that changes in the base sequence occur. An example of a base analogue is 5-bromouracil, which can replace thymine. Gene mutation tests measure those alterations of genetic material limited to the gene unit, that are transmissible to progeny unless repaired. Brusick (1980) refers to gene mutations as “ microlesions” because the actual genetic lesion is not microscopically visible. Microlesions are classified as either

248 Figure 12.5 Example of DNA adduct formation with benzo-a-pyrene.



Figure 12.6 Cytosine modified to uracil by nitrous acid.

basepair substitution mutations or frameshift mutations. These two categories of gene mutations are induced by different mechanisms and, often, by distinctly different classes of chemical mutagens. Yet both types of gene mutation are virtually always monitored by measuring some phenotypic change in the test organism. Microlesions occur at a much lower frequency (10–5 to 10–6) in comparison to chromosome aberrations or “ macrolesions,” which may be as frequent as 10–2 to 10–3. As described earlier, the basepair changes induced by point mutations (Figure 12.4) will also alter RNA codon sequences, which, in turn, change the amino acid sequence of the peptide chain being

Figure 12.7 Alkylating agent (EMS) effects on DNA.



formed, which may result in an alteration of some measurable cellular function. The phenotypic changes that can be monitored by this type of test include auxotrophic changes (i.e., acquired dependence on a formerly endogenously synthesized substance), altered proteins, color differences, and lethality. It is extremely difficult to detect those alterations in mammalian DNA caused by insertions or deletions of one or a few bases, except in rare instances where the specific protein product is known and its formation can be monitored. It is somewhat easier in bacterial or prokaryotic systems, and this has led to the use of bacterial or in vitro screening assays to detect potential mutagens. These issues are discussed in greater detail in Brusick (1980, 1994). Chromosomal aberrations, the third type of genetic change, may be present as chromatid gaps or breaks, symmetrical exchange (exchange of corresponding segments between arms of a chromosome), or asymmetric interchange between chromosomes. Point mutations can result in altered products of gene expression, but chromosomal aberration or alteration in chromosome numbers passed on through germ cells can have disastrous consequences, including embryonic death, teratogenesis, retarded development, behavioral disorders, and infertility. Some naturally occurring abnormalities of human chromosomal structure or number are shown in Table 12.2. The frequency of these events may be increased by mutagenic agents. Because these genetic lesions may be visualized by microscopy, they are referred to as macrolesions. One type of macrolesion is caused by an incomplete separation of replicated chromosomes during cell division. This is characterized by the abnormal chromosome numbers that result in the daughter cells and may be recognized as a change in the number of haploid chromosome sets (ploidy changes) or in the gain or loss of single chromosomes (aneuploidy). A second type of macrolesion caused by damage to chromosome structure (clastogenic effects) is categorized by the abnormal chromosome morphology that results. Two theories are currently available to explain the mechanism of chromosome aberration. One is the classic “ breakage-first” hypothesis. This theory assumes that the initial lesion is a break in the chromosomal backbone that is indicative of a broken DNA strand. Several possibilities exist following such an event: (1) the ends may repair normally and rejoin to form a normal chromosome; (2) the ends

TABLE 12.2 Examples of Human Genetic Disorders Chromosome Abnormalities Cri-du-chat syndrome (partial deletion of chromosome 5) Down’s syndrome (triplication of chromosome 21) Klinefelter’s syndrome (XXY sex chromosome constitution; 47 chromosomes) Turner’s syndrome (X0 sex chromosome constitution; 45 chromosomes) Dominant Mutations Chondrodystrophy Hepatic porphyria Huntington’s chorea Retinoblastoma Recessive Mutations Albinism Cystic fibrosis Diabetes mellitus Fanconi’s syndrome Hemophilia Xeroderma pigmentosum Complex Inherited Traits Anencephaly Club foot Spina bifida Other congenital defects



may not be repaired, resulting in a permanent break; or (3) they may be misrepaired or join with another chromosome to cause a translocation of genetic material. A second theory is the “ chromatid exchange” hypothesis. If the exchange occurs with a chromatid from another chromosome, an “ exchange figure” results. This theory assumes that the initial lesion is not a break and that the lesion can either be repaired directly or may interact with another lesion by a process called exchange initiation. Most chromosomal abnormalities result in cell lethality and, if induced in germ cells, generally produce dominant lethal effects that cannot be transmitted to the next generation. The traditional method for determining chromosomal aberrations is the direct visual analysis of chromosomes in cells frozen at the metaphase of their division cycle. Thus, metaphase-spread analysis evaluates both structural and numerical chromosome anomalies directly. Chemicals inducing changes in chromosomal number or structure also may be identified by the micronucleus test, an assay that assesses genotoxicity by observing micronucleated cells. It is a relatively simple assay because the number of cells with micronuclei are easily identified microscopically. At anaphase, in dividing cells that possess chromatid breaks or exchanges, chromatid and chromosome fragments may lag behind when the chromosome elements move toward the spindle poles. After telophase, the undamaged chromosomes give rise to regular daughter nuclei. The lagging elements are also included in the daughter cells, but a considerable proportion are included in secondary nuclei, which are typically much smaller than the principal nucleus and are therefore called micronuclei. Increased numbers of micronuclei represent increased chromosome breakage. Similar events can occur if interference with the spindle apparatus occurs, but the appearance of micronuclei produced in this manner is different, and they are usually larger than typical micronuclei. Historically, lymphocytes and epithelial cells have been the most commonly used cell populations. Many point mutations are detected by the cell and are deleted by various repair mechanisms. Some, however, remain undetected and are passed to daughter cells. The significance of the mutations varies with the type of cell, and the location within the DNA. If the cell is of somatic lineage, altered gene products can result from gene expression. If the cell is a gonadal cell (or germ cell), the change can be passed on to offspring and may cause problems in future generations. Much of the DNA in organisms is never expressed. If the mutation occurs in that portion of the DNA that is not expressed, no problem occurs. However, if the mutation occurs in the active portion of the DNA, the altered gene products can be expressed. An example of a problematic point mutation is in the gene that causes sickle cell anemia. A change of one basepair (a transversion from thymine to adenine) results in the amino acid glutamate being replaced by another amino acid, valine, in one of the molecules that makes up hemoglobin, the oxygen-carrying molecule in red blood cells. When the blood becomes deoxygenated, such as under heavy exercise conditions, the valine allows the red blood cells to assume a sickle shape instead of the normal circular shape. This leads to clumping of blood cells in capillaries, which in turn may limit blood flow to the tissues. This behavior of the blood cells exacerbates other effects of sickle cell anemia, which result in oxygen deprivation because the hemoglobin content of the blood in persons with sickle cell anemia is about half that of other persons.

12.3 NONMAMMALIAN MUTAGENICITY TESTS Because results from bacterial or prokaryotic assays often establish priorities for other testing approaches, it is of interest to briefly describe the assays currently used to screen for mutagenic capacity, particularly those done in industrial settings. Rapid cell division and the relative ease with which large quantities of data can be generated (approximately 108 bacteria per test plate) have made bacterial tests the most widely utilized routine means of testing for mutagenicity. These systems are the quickest and most inexpensive procedures. However, bacteria are evolutionarily far removed from the human model. They lack true nuclei as well as the enzymatic pathways by which most promutagens are activated to form mutagenic compounds. Bacterial DNA has a different protein coat than seen in eukaryotes. Nevertheless, bacterial systems have great utility as a preliminary screen for potential mutagens.



In addition to bacteria, fungi have been used in genotoxicity assays. The Saccharomyces and Schizosaccharomyces yeasts, as well as the molds Neurospora and Aspergillus, have been utilized in forward mutation tests, which are similar in design to the salmonella histidine revertant assays that will be described in the next section.

Typical Bacterial Test Systems The most widely utilized bacterial test system for monitoring gene mutations and the most widely utilized short-term mutagenicity test of any type is the Salmonella typhimurium microsome test developed by Dr. Bruce Ames and co-workers and commonly called the Ames assay. The phenotypic marker utilized for the detection of gene mutations in all the Ames Salmonella strains is the ability of the bacteria to synthesize histidine, an amino acid essential for bacterial division. The tester strains of bacteria have mutations rendering them unable to synthesize histidine; thus, they must depend on histidine included in the culture medium in order to be able to multiply. Bacteria are taken directly from a prepared culture and incorporated with a trace of histidine into soft agar overlay on a dish containing minimal growth factors. The bacteria undergo several divisions, which are necessary for the expression of mutagenicity and, after the available histidine has been used up, a fine bacterial lawn is formed. Bacteria that have back-mutated in their histidine operon sites (and thus have reverted to the ability to synthesize histidine) will keep on dividing to form discrete colonies, while the nonmutated bacteria will die. A chemical that is a positive mutagen will demonstrate a statistically significant dose-related increase in “ revertants” (colonies formed) when compared to the spontaneous revertants in control plates. Five Ames S. typhimurium tester strains are recommended for routine mutagenicity testing: TA1535, TA1537, TA1538, TA98, and TA100. The TA1535 tester strain detects basepair substitution mutations. The TA1538 tester strain detects frameshift mutagens that cause basepair deletions. The TA1537 tester strain detects frameshift mutagens that cause basepair additions. The TA100 (basepair substitution) and TA98 (frameshift) strains are sensitive to effects caused by certain compounds, such as nitrofurans, which were not detectable with the previous three strains. The lack of oxidative metabolism to transform promutagens (those mutagens requiring bioactivation to the active form) is overcome in these bacterial assays by two means. First, a suspension of rat liver homogenate containing appropriate enzymes may be added to the bacterial incubation. The liver preparation is centrifuged at 9000g for 20 min at 4°C, and the resultant supernatant (S9) is added to the culture medium. In a slightly more complex procedure, called the host-mediated assay, the bacterial tester strains are injected into the body cavity of a test animal such as the mouse. This host is treated with the suspected mutagen and, after a selected period, the bacteria are harvested and assayed for mutation (revertants) as described earlier. Other bacterial species used in mutagenicity screens include Escherichia coli and Bacillus subtilis. Assays that measure DNA repair in bacterial systems have also been developed. These tests are based on the premise that a strain deficient in DNA repair enzymes will be more susceptible to mutagenic activity than will a similar strain that possesses repair enzymes that can correct the mutagenic damage. A “ spot” test consists of placing the chemical to be tested in a well or on a paper disk on top of the agar in a petri dish. The test chemical will diffuse from the central source, causing a declining concentration gradient as the distance from the source increases. A strain deficient in repair enzymes will exhibit a greater diameter of bacterial kill than the repair-sufficient strain tested with a mutagen. In a “ suspension” test, a given number of bacteria are preincubated with and without the test compound. The bacteria are then plated and the colonies counted. The repair-deficient strain will demonstrate a greater percentage kill than will the sister DNA-repair-sufficient strain. A liver S9 activation system can also be incorporated in bacterial DNA repair tests. The most widely used bacterial DNA repair test utilizes the polA+ and polA– strains of E. coli. The polA– strain is deficient in DNA polymerase I, whereas the polA+ strain is sufficient in this enzyme.



Drosophila Test Systems The fruitfly (Drosophila melanogaster) has received wide use in the sex-linked recessive lethal test. The endpoint phenotypic change monitored in this test is the lethality of males in the F2 generation. Brusick has gone to the extent of labeling Drosophila an “ honorary mammalian model” by virtue of its widespread application and correlation with positive mutagens in mammalian testing. Drosophila melanogaster has also been utilized to monitor two types of chromosomal aberration endpoints through phenotypic markers: loss and nondisjunction of X or Y sex chromosomes and heritable translocations. The monitoring of translocations has the advantage of a very low background rate, facilitating comparisons between controls and treated groups. Dominant lethal assays are also performed with insects and can theoretically be applied in any organism where early embryonic death can be monitored. The male is treated with the test agent, then mated with one or more females. If early fetal deaths occur, these are demonstrative of a dominant lethal mutation in the germ cells of the treated male.

Plant Assays A number of assay types are available in plant systems as well, including specific locus tests in corn (Zea maize) and multilocus assays in Arabidopsis. Cytogenetic tests have been developed for Tradescantia (micronucleus test), as have chromosomal aberration assays in the root tips of onions (Allium sepa) and beans (Vicia faba). Finally, DNA adducts analysis is applicable to somatic and germinal plant cell systems. It is anticipated that one or more plant species may prove to be useful indicators of the potential for genetic damage that may be related to emissions of environmental pollutants.

12.4 MAMMALIAN MUTAGENICITY TESTS Testing chemicals for mutagenicity in vivo in mammalian systems is the most relevant method for learning about mutagenicity in humans. Mammals such as the rat or mouse offer insights into human physiology, metabolism, and reproduction that cannot be duplicated in other tests. Furthermore, the route of administration of a chemical to a test animal can be selected to parallel normal human environmental or occupational conditions of oral, dermal, or inhalation exposures. Human epidemiologic findings may also be compared with the results of tests done in animals. While the monitoring of human exposures and their effects does not constitute planned, controlled mutagenicity testing, human epidemiology offers the opportunity to monitor and test for correlations suggested by other mutagenicity tests. Thus, these studies are the only opportunity for direct human modeling of a chemical’s mutagenic potential. It is worth noting that despite extensive investigation, to date no chemical substances have been positively identified as human mutagens. The advantages and limitations of a wide variety of genetic test systems are presented in Brusick (1994). One perceived disadvantage of in vivo mammalian test systems is the time they require and their cost. A larger commitment of physical resources and personnel is required than is required with in vitro testing. Human epidemiology studies are further complicated by the fact that not all of the environmental variables can be controlled. Frequently, the duration and extent of exposure to single or multiple compounds can only be estimated. Nevertheless, progress is being made to lessen the cost and decrease the time required for in vivo mammalian testing. Also, new data handling, statistical techniques, and increased cooperation from industry have increased the reliability of human epidemiology studies. More regular sampling of workplace exposures has helped to improve the quality and accessibility of human data. Mutagenic potential can vary greatly across a class of analytes, as shown for the metals (Costa, 1996). Mutagenicity data for metals can be quite difficult to interpret due to the breadth of mechanisms at work, as illustrated by differences between Cr, Ni, As, and Cd.



Germ Cell Assays A basic test used to detect specific gene mutations induced in germ cells of mammals is the mouse-specific locus assay. This test involves treatment of wild-type mice, either male or female, with a test compound before mating them to a strain homozygous for a number of recessive genes that are expressed visibly in phenotype. If no mutations occur, then all offspring will be of the wild type. If a mutation has occurred at one of the test loci in the treated mice, then the recessive phenotype will be visibly detectable in the offspring. The mouse-specific locus test is of special significance in human modeling because it is the only standardized assay that directly measures heritable germ cell gene mutations in the mammal. A major drawback of the mouse-specific locus test is that extensive physical plant facilities are required to execute this assay, and it has been estimated that one scientist and three technicians could execute 10 single-dose mouse-specific locus tests in one year, provided there are facilities for 5000 cages. New and promising test procedures have been described for detecting germ cell mutations by using alterations in selected enzyme activity as the phenotypic endpoint. A large group of somatic cell enzymes can be monitored for changes in activity and kinetics in the F1 generation. These changes indicate changes in the parental genome. A similar biochemical approach has been proposed for identifying germ cell mutations in humans through the monitoring of placental cord blood samples. The activity of several erythrocyte enzymes, such as glutathione reductase, can be monitored because the enzyme proteins are the products of a single locus and because heterozygosity of a mutant allele for the chosen enzymes will result in abnormal levels of enzyme activity. Likewise, it has been proposed that gene mutations be directly monitored in mammalian germ cells by searching for phenotypic variants with biochemical markers such as lactic acid dehydrogenase-X (LDH-X), an isozyme of lactic acid dehydrogenase found only in testes and sperm. The test is based on the fact that a monospecific antibody for rabbit LDH-X reacts with rat but not mouse LDH-X in sperm. The rat sperm fluoresce as a result of the reaction but the mouse sperm do not fluoresce unless a phenotypic variant is present. If adapted to humans, this test has potential use as a noninvasive screening test of germ cell mutations in males. It has been proposed that the induction of behavioral effects in the offspring of male rats exposed to a mutagenic agent may represent a genotoxic endpoint. For example, studies have demonstrated that the mutagen cyclophosphamide can induce genotoxic behavioral effects in the progeny of male rats and that these effects correspond to observed genetic damage caused in the spermatozoa following meiosis. A similar effect has been attributed to vinyl chloride in at least one instance of occupational exposures. Mammalian germ cells can be monitored for chromosomal aberrations, and normally the testes are used as the cell source. Mammalian male germ cells are protected by a biological barrier comparable in function to the barrier which retards the penetration of chemicals to the brain. The blood–testes barrier is a complex system composed of membranes surrounding the seminiferous tubules and the several layers of spermatogenic cells organized within the tubules. This barrier restricts the permeability of high-molecular-weight compounds to the developing male germ cell. An advantage of in vivo mammalian germ cell mutagenicity testing is that the protective contribution of this barrier is automatically taken into account. Conventional procedures for harvesting mammalian male germ cell tissue for metaphase-spread analysis involve mincing or teasing the seminiferous tubules to liberate meiotic germ cells in suspension. This homogenate is centrifuged, the centrifuged pellet is discarded, and the suspended cells are collected and analyzed. However, it was found that the tissue fragments discarded during this conventional procedure contained more spermatogonial cells and meiotic metaphases than did the suspension. Thus, the method has been refined by using tissue fragments and adding collagenase to dissociate them. After collagenase treatment, the tissue fragments are gently homogenized and centrifuged, and the pellet containing meiotic cells is resuspended and prepared for microscopic analysis.



Dominant Lethal Assays Dominant lethal assays can be performed in any organism where early embryonic death can be monitored. As described earlier, mammals are commonly used in dominant lethal assays, although it is possible to do so with insects as well. The male animals (typically mice or rats) are treated with the suspected mutagen before being mated with one or more females. Each week these females are removed and a new group of females is introduced to the treated male. This process is repeated for a period of 6–10 weeks. The females are sacrificed before parturition, and early fetal deaths are counted in the uterine horns. This test has become standardized, and a large number of compounds have been screened in mouse studies. As with most in vivo mammalian assays, costs and commitment of resources can be extensive. However, the applicability of the data is typically quite good. The dominant lethal test in rodents is of significance for human modeling because it gives an indication of heritable chromosomal damage in a mammal. Even though the endpoint of early fetal death may seem of minor significance when considering only its effects on the human gene pool, it does provide a signal that viable heritable chromosomal damage and gene mutations may also be produced.

Heritable Translocation Assay Results of dominant-lethal assays frequently correlate well with another test used for determining clastogenic effects in mammalian germ cells: the heritable translocation assay (HTA). Translocation represents complete transfer of material between two chromosomes. In HTA procedures, male mice are mated to untreated females after treatment with the test compound, and the pregnant females are allowed to deliver. Male offspring are subsequently mated to groups of females. If translocations are produced through genotoxic action, then the affected first-generation male progeny will be partially or completely sterile; this can be noticed in the litter size produced from those females to which they were mated.

Micronucleus Tests Application of the micronucleus test to mammalian germ cells recently has been reported. This test procedure is analogous to the bone marrow micronucleus test (somatic cells), but it involves the sampling of early spermatids from the seminiferous tubules of male rats. The number of micronuclei are quantified by using fluorescent stain and counting micronucleated spermatids. To date, the technique has not been widely used in occupational evaluations.

Spermhead Morphology Assay Some relatively new tests have been developed that evaluate the ability of a test chemical to induce abnormal sperm morphology when compared to controls. It has been proposed that an increase in abnormal sperm morphology is evidence of genotoxicity because there seems to be an association between abnormal sperm morphology and chromosome aberrations. However, recent investigations have reported that induction of morphologically aberrant sperm can be caused by nongenotoxic actions, such as dietary restriction. In addition, some known mutagens, including 1,2-dibromo-3-chloropropane (a pesticide with mutagenic, carcinogenic, and gonadotoxic properties), were reportedly unable to induce production of spermhead abnormalities in mice, when tested. It should be noted that sperm abnormalities are fairly common in humans and may occur at rates of 40–45 percent. Thus, more verification is needed before strong conclusions can be drawn about the mammalian spermhead morphology assay.



Tests for Primary DNA Damage A historical test thought to monitor primary DNA damage in mammalian germ cells in vivo involves the monitoring of sister chromatid exchange. The observation of sister chromatid exchanges through differential staining involves exposing the cells to bromodeoxyuridine for two rounds of replication, so that the chromosomes consist of one chromatid substituted on both arms with 5-bromodeoxyuridine and the other substituted only on a single arm. Differential staining between sister chromatids is due to the differences in bromodeoxyuridine incorporation in the sister chromatids. Unscheduled DNA repair has been induced by chemical mutagens in mammalian male germ cells from the spermatogonial to midspermatid stages of development. The test is based on the fact that cells not undergoing replication (scheduled DNA synthesis) should not exhibit significant DNA synthesis. Thus, incorporation of radiolabeled tracer molecules into the DNA of these cells should be minimal. However, if a chemical mutagen damages the DNA, the DNA repair system may be activated, causing unscheduled DNA synthesis (UDS). If such is the case, radiolabeled tracers will be incorporated into the DNA; these can be monitored by autoradiography or by direct measurement of radioactivity in the repaired DNA. Male germ cells lose DNA repair capability when they have advanced to the late spermatid and mature spermatozoa stages; unscheduled DNA synthesis cannot then be induced by chemical mutagens. The genotoxic agents methyl methanesulfonate, ethyl methanesulfonate, cyclophosphamide, and Mitomen have been shown to induce unscheduled DNA repair in vivo in male mouse germ cells. Similar procedures are available to evaluate UDS in some types of somatic cells as well.

Transgenic Mouse Assays In the late 1980s and early 1990s the development of a new genotoxicity assay system was reported by Gossen et al. (1989), Kohler et al. (1991), and colleagues. Briefly, the test system involves mutagen dosing to a specific mouse strain (C57BL6) that has been infected with a viral “ shuttle vector,” isolation of the mouse DNA, recovery of the phage segment (lacI or LacZ), and infection of an E. coli strain with the recovered phage. The phage will form plaques on a lawn of E. coli. The plaques are colorless if no mutation has occurred or blue if a mutation has occurred. The assay may be performed to gather information on mutations in somatic cells or in germ cells. The lacZ assay also is known as the “ Muta-Mouse” assay, while the lacI assay also is known as the “ Big Blue” assay. Advantages of the assay include its in vivo treatment regime, the fact that it can be conducted in a few days from the isolation of DNA through plaque formation to mutation scoring. However, it may be difficult to use extremely high dosages (e.g., approaching lethal doses), since the mice must survive for 1–2 weeks in order to fix the mutation in the affected tissues. The performance of the transgenic mouse assays that have been conducted on 26 substances was evaluated by Morrison and Ashby (1994), including a review of the results of the tests that had been performed in the lacZ case (14 reports) and in the lacI case (16 reports). These authors concluded that the variability of data reporting formats and the rapid developments and modifications in the assay protocols make it difficult to perform direct comparisons among tests or between this assay type and the results of other historically available methods. Nevertheless, the initial results are generally promising, and there are no examples of internal disagreement between responses for the same chemical in the same tissue.

In Vitro Testing Test systems have been developed that use mammalian cells in culture (in vitro) to detect chemical mutagens. Disadvantages in comparison with in vivo mammalian tests are that in vitro tests lack organ–system interaction, require a route for administration of the agent that cannot be varied, and lack the normal distributional and metabolic factors present in the whole animal. The obvious advantages are that costs are decreased and that experiments are more easily replicated, which facilitates verification of results. Cases where human cells have been cultured successfully (e.g., lymphocytes) provide the only viable in vitro experiments on the human organism. Several endpoints



can be used during testing of potential mutagens in vitro. One of the most common involves the monitoring of mutations in specific well-characterized gene loci, such as those coding for hypoxanthineguanine phosphoribosyl transferase (HGPRT), thymidine kinase (TK), or ouabain resistance (OVAr). Mutagenic modification in the segments of the DNA coding for these proteins (enzymes) results in an increased sensitivity of the cell, which can often be evaluated by the cell’s heightened susceptibility to other agents (e.g., bromodeoxyuridine or 8-azaguanine). As described in the section on in vivo mammalian testing, evaluations of sister chromatid exchange, DNA repair activity, and chromosomal aberration through interpretation of metaphase spreads may be applied to in vitro testing of mutagens. An additional procedure that has been correlated with chemical mutagenicity is examination for cell culture transformation; following treatment with mutagens, some cells in culture lose their normal, characteristic arrangement of monolayered attachment and begin to pile up in a disorganized fashion. Two major drawbacks in looking for this feature are that considerable expertise is necessary to interpret the results accurately and that the criteria for evaluation are more subjective than for other mutagenicity assays. A comparison of the sensitivity and specificity of selected short-term tests by two recognized systems (National Toxicology Program (NTP) and Gene-Tox) is shown in Table 12.3.


Areas of Concern: Gene Pool and Oncogenesis The potential significance of occupationally acquired mutations can be divided into two areas. The first is concern for the protection of the human gene pool. This factor may represent the most significant reason for genetic testing, but it is often underemphasized by nongeneticists involved with safety evaluation because of the inability to demonstrate induced mutation in humans to date. The second area is that of oncogenesis. The intimate relationship between the tumorigenic and genotoxic properties of many chemicals (Figure 12.8) makes genetic testing a potentially powerful screening technique for establishing priorities for future testing of chemicals of unknown cancer-causing potential. This factor has been one of the primary driving forces behind the rapid expansion of genetic toxicology as a discipline. Once again, however, the paucity of proved human carcinogens compared with the number of demonstrated animal carcinogens suggests weaknesses in the process of extrapolating from animal studies to human exposure in the workplace. At the heart of the present legal and regulatory approach toward environmental and occupational exposure to mutagens is the possibility that they may cause human genetic damage. Two important assumptions underlie this central concept:

• Environmental or occupational mutagens may cause aneuploidy, chromosome breaks, point mutations, or other genetic damage in humans

• Environmental or occupational mutagens that can be controlled by regulatory efforts represent a significant component of total human exposure Much of the interest in potential environmental and occupational mutagens is related to the prevalent opinion that many cancers are initiated by a mutagenic event. This premise is supported by the strong correlation between some specific occupational chemical exposures and cancer incidence in humans. One good example is the relationship between liver cancer (angiosarcoma) and exposure to vinyl chloride in some manufacturing operations. Another example is the respiratory tract cancers that may be caused by exposure to bis(chloromethyl)ether.


175/223 45/54 40/41 84/104 77/106 40/54 8/9 100/101 21/21 19/22

(+)/Total 78 87 98 81 73 74 89 99 100 86

%(+) 20/44 31/44 — — 4/18 24/44 9/15 31/44 10/15 6/30



45 70 — — 22 55 60 70 67 20


Source: Kier (1988). a Sensitivity is the proportion of positive results for carcinogens. b Specificity is the proportion of negative results for noncarcinogens. c Gene-Tox data include combined results for sufficient and limited-evidence carcinogens and noncarcinogens. d NTP specificity assumes that equivocal evidence compounds are noncarcinogens.

Ames/Salmonella Mouse/lymphoma CHO/HGPRT V79 Drosophila SLRL In vitro cytogenetics In vivo cytogenetics In vitro SCE In vivo SCE UDS in hepatocytes




TABLE 12-3 NTP and Gene-Tox Evaluation of Short-Term Test Sensitivities and Specificities

29/47 0/5 1/1 3/3 9/16 2/6 0/0 0/10 0/0 0/0



62 0 100 100 60 33 — 0 — —


25/29 13/29 — — 9/9 20/29 11/12 13/29 5/12 13/14


Specificityb NTP

86 45 — — 100 69 92 45 42 93




Figure 12.8 Proposed relationships between mutagenicity and carcinogenicity. (Adapted from Fishbein, 1979).

A Multidisciplinary Approach Although each of the mutagenicity tests described in this chapter has individual strengths, likewise each is weak in some facets of detection capacity. Clearly, therefore, the accurate and efficient testing of chemicals and the protection from potential occupational mutagens require a multidisciplinary approach that integrates toxicology, clinical chemistry, microbiology, pathology, epidemiology, industrial hygiene, and occupational medicine. Testing intact animals has the advantage of increasing the reliability of any extrapolations that must be made from the data; however, cost considerations often limit the application of in vivo mammalian assays except when it is expected that they will verify lower tier assays. There are three areas where the results of mutagenicity testing of a given substance may be applied:



• Extrapolation of test results in order to make a quantitative evaluation of the hazard of exposures for humans

• Prioritization of human hazards caused by specific compounds • Institution of remedial procedures that should be undertaken to minimize the human hazard One of the most difficult areas of analysis is the correct application of mutagenicity tests to arrive at a quantifiable human hazard from exposure to a given substance. There is frequently good correlation between the mutagenicity and carcinogenicity of a substance in animal tests (Table 12.4). However, this may be misleading because models for carcinogenicity determination are often characterized by chronic procedures utilizing very high doses in nonprimate species. These may bear little resemblance to aspects of exposure in the human model, such as magnitude and route of exposure, metabolic patterns, and environment (which are qualitative factors), and exposure dose (which is quantitative). As noted previously, the time and expense that are involved with lifetime carcinogenicity assays have strongly influenced the use of test batteries as predictive measures of carcinogenic potential. Among many others, Ashby and Tennant (1994), Anderson et al. (1994), Benigni and Giuliani (1987), and Blake et al. (1990), have addressed the question of applicability of multiple test systems to the classification of a substance as genotoxic or not, and carcinogenic or not. It is important in these efforts to distinguish among “ sensitivity” (ability to identify a known carcinogen), “ specificity” (ability to identify a noncarcinogen), and “ accuracy” (correct results of either type). Parodi et al. (1990) reported on qualitative correlations associated with studies of up to 300 substances conducted during the period 1976 through 1988. Initial measures of sensitivity, specificity, and accuracy were approximately 90 percent, if the decision is based solely on Salmonella assays. As more substances have been tested, this estimate has ranged from 45 to 91 percent. Best results typically are reported for sensitivity, where accuracy generally is on the order of 65 to 75 percent. Consideration of the quantitative correlation between short-term genotoxicity tests and carcinogenic potency has yielded extremely variable estimates, ranging from approximately 30 to over 90 percent. The overwhelming conclusion was that a battery of test systems that addresses differing endpoints is required if the goal is to develop a confident conclusion regarding predictivity. As is the case with many areas of toxicology, one may choose between in vivo and in vitro test systems, each with their attendant advantages and disadvantages. The testing of chemicals in experimental animals has all the advantages of any intact in vivo system; that is, it has all of the biochemical and physiological requirements to make anthropomorphization more reliable. However, in vivo mutagenicity testing may require an investment of many thousands of dollars and a long period of time. These disadvantages often force the tester to use a less expensive, well-established short-term

TABLE 12.4 Comparative Mutagenicity of Various Compounds

Compound Epichlorohydrin Ethyleneimine Trimethyl phosphate Tris Ethylene dibromide Vinyl chloride Chloroprene Urethane N = no; Y = yes; O = not tested.

Established human carcinogen




Mammalian cells

Human cells









bioassay, such as the Ames Salmonella bacterial assay, to determine the mutagenicity of a chemical, and then extrapolate these results into the in vivo mutagenicity model.

Occupational Monitoring and Biomarkers for Genetic Damage Cytogenetic analysis (chromosome evaluation) of human lymphocytes has been a standard industrial technique for monitoring human genetic damage. However, several limitations are inherent in the conventional use of human lymphocytes as indicators of exposure to genotoxic chemicals or radiation:

• Individual and population variability in normal levels of chromosomal aberrations may mask

• •

small changes in the frequency of mutations. To overcome this obstacle, specific defects that occur with low frequency in normal individuals may also be tested for, but typically several thousand cells must be scored per individual to achieve sufficient sample size. Evaluation of chromosomal aberrations is subject to substantial variation between laboratories. Therefore, replicate readings should be obtained; this substantially increases the effort required when thousands of samples are involved. Since chromosomal aberrations are considered indicators of relatively gross damage, the techniques may miss many more subtle effects of mutagens.

Evaluation of sister chromatid exchange (SCE) may be potentially valuable in answering some of these difficulties. For example, SCEs are elevated in patients undergoing chemotherapy, which is not unexpected, as many of the chemotherapeutic agents are powerful mutagens. These elevations tend to be dose-related, which supports the usefulness of the technique as a potential screening device. It must be emphasized that SCE may not be a damaging lesion in itself, but may prove a useful marker for other detrimental effects on the DNA induced by the agents in question. This caveat is underscored by the observation that SCE is poorly correlated with radiation exposure and exposure to other agents that break DNA. Agents that alkylate the DNA (bind tightly to the molecule) show a better correlation with mutagenic potential and may be a sensitive indicator for the monitoring of chromosomal aberrations, which are otherwise more difficult and time-consuming to detect. The Ames-type mutagenicity testing of urine from exposed individuals (e.g., tobacco, chemotherapy patients) has yielded promising results as a simple, rapid, and inexpensive screening technique, although the timing and impact of cumulative versus acute exposures are not yet fully understood. The evaluation of other biomarkers for genetic damage are under development or investigation, particularly with regard to germinal cell populations. These methods include techniques to detect formation of DNA and protein adducts, and changes in sperm morphology or fertility indices. These issues recently were reviewed by the National Research Council (1989).

Areas for Future Activity Many mutagenicity assays have been proposed, each with a unique attribute and measurable biochemical or visible endpoint. These tests are being incorporated into routine safety assessment programs in all regulatory agencies. Furthermore, these tests have been proposed as part of a regulatory decision-making policy by the Occupational Safety and Health Administration (OSHA) for the classification of chemical carcinogens in the workplace, and by the U.S. Environmental Protection Agency (USEPA) for the regulation of pesticides and for regulating the disposal of toxic wastes. A tremendous amount of information is available through the Environmental Mutagen Information Center (EMIC), housed at the Oak Ridge National Laboratory. The short-term mutagenicity tests actually serve two purposes. They not only assist in the assessment of a chemical’s potential for cancer induction but also assess the potential for inducing germ cell mutations in humans. Some of the organizations involved in the development of guidelines for germ cell mutagenicity tests are the International Commission for Protection against Environmental Mutagens



and Carcinogens (ICPEMC), the World Health Organization, and the Commission for European Communities. In the past, most estimates of genotoxic risks were more qualitative than quantitative, and the emphasis has rested on somatic effects (e.g., those leading to cancers) rather than on germinal cells (sperm and ovum). On the basis of evidence in animals demonstrating germinal cell effects, it is imperative to develop human screening methods capable of detecting such effects. Therein lies one of the premier challenges to genetic toxicology and occupational medicine. The uncertainties of accurate extrapolation of mutagenicity test data to a human hazard model have supported the philosophy that if uncertainty is to occur in extrapolation it should favor the side of safety. This concept is particularly important in the consideration of whether or not threshold characteristics may exist. In the case of carcinogens, discussed further in the next chapter, good evidence supports the view that genotoxic (DNA-damaging) carcinogens may be distinct from epigenetic carcinogens (those that induce or potentiate cancer by means other than direct DNA interaction). For the purposes of this discussion, mutagens are assumed to exert nonthreshold effects. That is, even as one approaches zero dose, there is still a calculable risk of DNA effects. The concern for the potential mutagenic hazards in the workplace from exposure to chemicals should include routine tests of nonpregnant females and males, as well as the more traditional monitoring of pregnant and lactating women. For example, vinyl chloride, mentioned earlier in relation to its suggested role in angiosarcoma of the liver, has been correlated with an increased incidence of nervous system malformations in infants fathered by exposed workers. It has also been demonstrated to cause elevations in chromosomal aberration in the occupationally exposed. 1,2-dibromo-3-chloropropane (DBCP), a pesticide linked to sterility in exposed male workers, causes increases in indices of mutagenic capacity in humans and animals. Monitoring of male populations may prove particularly important in that the spermatogenic cycle is continuous in adults and therefore poses continuous opportunities for genetic damage to be expressed as damaged chromosomes. Since the female carries the full lifetime complement of ova at birth, susceptibility to propagation of genetic alteration during cell division is reduced except in those periods of division following conception. By the same token, the cessation of exposure in the male should allow for recovery from a mutagenic event in premeiotic spermatocytes, providing that spermatogonia are not affected. If chromosome damage occurs in sperm or ovum, then fetal death frequently occurs. Greater than 50 percent of spontaneous abortions in humans show chromosomal defects. Once mutagenic potential is established for a compound, the risks posed by exposure under expected conditions must be assessed. As discussed, complications may be encountered in situations where mutagenic effects are due to “ multihit” phenomena and therefore reflect threshold-type responses. A more complete discussion on risk assessment is presented in Chapter 18.

12.6 SUMMARY Modification of genetic material by mutagenic agents poses a serious environmental and occupational threat. Chemical or physical mutagens may induce cancer or lead to germ cell alteration.

• The mutagens that lead to cancer alter the DNA of somatic cells so as to cause modifications in gene expression, which results in tumorigenesis.

• Germ cell (sperm, ovum) mutagens may exert their effects through decreased fertility, birth defects, spontaneous abortion, or through changes that may not become evident for several subsequent generations (such hidden mutagenic effects remain essentially undetectable except when expressed as a gross malformation). Many screening tests have been developed to investigate the mutagenic potential of chemical agents.

• These assays use bacteria, insects, mammals, and various cells in culture.



• Although in vitro tests are less expensive and less complex, in vivo mammalian tests give results that can be extrapolated to human circumstances more realistically, but in vivo studies are expensive and labor-intensive. Persons whose occupations expose them to potential mutagens may undergo chemically induced changes at a greater rate than the general population does. Validation of this hypothesis is the subject of extensive ongoing research.

• Epidemiology seeks to identify groups with increased susceptibility to chemical mutagens, or increased incidence of exposure, in order to limit exposures.

• No single method currently stands out as the most comprehensive and thorough screen for identifying mutagenic agents; often, a multidisciplinary approach employing several tests is best suited to the accurate identification of industrial mutagens. Once mutagenic potential has been demonstrated for a compound, typically an analysis must be made of the risks posed to exposed individuals. Such a determination is essential in the qualitative evaluation of the occupational hazard of mutagens.

REFERENCES CITED AND SUGGESTED READING Anderson, D., M. Sorsa, and M. D. Waters, “ The parallelogram approach in studies of genotoxic effects,” Mutat. Res. 313: 101 (1994). Ashby, J., and R. W. Tennant, “ Prediction of rodent carcinogenicity for 44 chemicals: results,” Mutagenesis 9: 7 (1994). Ashby, J., and H. Tinwell, “ Use of transgenic mouse lacI/Z mutation assays in genetic toxicology,” Mutagenesis 9: 179 (1994). Auerbach, C., J. M. Robson, and J. G. Carr, “ The chemical production of mutations.” Science 105: 243 (1947). Barlow, S. M., and F. M. Sullivan, Reproductive Hazards of Industrial Chemicals, Academic Press, New York, 1982. Benigni, R., “ Rodent tumor profiles Salmonella mutagenicity and risk assessment,” Mutat. Res. 244: 79 (1990). Benigni, R., and A. Giuliani, “ Which rules for assembling short-term test batteries to predict carcinogenicity,” Molec. Toxicol. 1: 143 (1987). Berg, K., ed., Genetic Damage in Man Caused by Environmental Agents, Academic Press, New York, 1979. Blake, B. W., K. Enslein, V. K. Gombar, and H. H. Borgstedt, “ Salmonella muatgenicity and rodent carcinogenicity: Quantitative structure-activity relationships,” Mutat. Res. 241: 261 (1990). Brusick, D. J., Principles of Genetic Toxicology, Plenum Press, New York, 1980. Brusick, D. J., ed., Methods for Genetic Risk Assessment, Lewis Publishers, New York, 1994. Calabrese, E. J., Pollutants and High Risk Groups, Wiley-Interscience, New York, 1978. Cohen, B. H., A. M. Lilienfeld, and P. C. Huang, eds., Genetic Issues in Public Health and Medicine, Charles C. Thomas, Springfield, IL, 1978. Costa, M., “ Introduction to metal toxicity and carcinogenicity of metals,” in L. W. Chang, ed., Toxicology of metals, CRC Press, Boca Raton, 1996. Fishbein, L., Potential Industrial Carcinogens and Mutagens, Elsevier Scientific, Amsterdam, 1979. Gossen, J. A., W. J. F de Leeuw, C. H. T. Tan, E. C. Zwarhoff, F. Berends, P. H. M. Lohman, D. L. Knook, and J. Vijg, “ Efficient rescue of integrated shuttle vectors from transgenic mice: A model for studying mutations in vivo,” Proc. Natl. Acad. Sci. (USA) 86: 7971 (1989). Hoffmann, G. R., “ Genetic toxicology,” In M. O. Amdur, J. Doull, and C. D. Klaassen, eds., Casarett and Doull’s Toxicology: The Basic Science of Poisons, 4th ed., Macmillan, New York, 1991. Hollaender, A., ed., Chemical Mutagens: Principles and Methods for Their Detection, Vols. 1–8, Plenum Press, New York, 1971–1984.



ICPEMC (International Commission for Protection against Environmental Mutagens and Carcinogens), “ Estimation of genetic risks and increased incidence of genetic disease due to environmental mutagens,” Mutat. Res. 115: 255 (1983b). ICPEMC (International Commission for Protection against Environmental Mutagens and Carcinogens), “ Regulatory approaches to the control of environmental mutagens and carcinogens,” Mutat. Res. 114: 179 (1983b). Kier, L. D., “ Comments and perspective on the EPA workshop on the relationship between short-term test information and carcinogenicity,” Environ. Molec. Mutag. 11: 147–157 (1988). Kirsch-Volders, M., Mutagenicity: Carcinogenicitry and Teratogenicity of Industrial Pollutants, Plenum Press, New York, 1984. Kohler, S. W., G. S. Provost, A. Fieck, P. L. Kretz, W. O. Bullock, D. L. Putman, J. A. Sorge, and J. M. Short, “ Analysis of spontaneous and induced mutations in transgenic mice using a lambda ZAP/lacl shuttle vector,” Environ. Molec. Mutag. 18: 316 (1991). Livingston, G. K., “ Environmental mutagenesis,” in W. N. Rom, ed., Environmental and Occupational Medicine, Little, Brown and Co., Boston, 1992. Mendelsohn, M. L., and R. J. Albertini, Mutation and the Environment, Wiley-Liss, New York, 1990. Miller, E. C., J. A. Miller, I. Hirono, P. Sugimura, and S. Takayama, Naturally Occurring Carcinogens, Mutagens, and Modulators of Carcinogenesis, University Park Press, Baltimore, 1979. Morrison, V., and J. Ashby, “ A preliminary evaluation of the performance of the Muta-Mouse (lacZ) and Big Blue (lacI) transgenic mouse mutation assays,” Mutagenesis 9: 367 (1994). Muller, H. J., “ Artificial transmutation of the gene,” Science 64: 84 (1927). NRC (National Research Council), Biomarkers in Reproductive and Neurodevelopmental Toxicology, National Academy Press, Washington, DC, 1989. Osterloh, J. D., and A. B. Tarcher, “ Environmental and biological monitoring,” in Principles and Practice of Environmental Medicine, A. B. Tarcher, ed., Plenum Medical, New York, 1992. Parodi, S., M. Taningher, P. Romano, S. Grilli, and L. Santi, “ Mutagenic and carcinogenic potency indices and their correlation,” Teratogen. Carcinogen. Mutagen. 10: 177 (1990). Rom, W. N., ed., Environmental and Occupational Medicine, Little, Brown, Boston, 1983. Shaw, C. R., ed., Prevention of Occupational Cancer, CRC Press, Boca Raton, FL, 1981. Sorsa, M., and H. Vainio, eds., Mutagens in Our Environment, Alan R. Liss, New York, 1982. Suutari, A., and T. Sjoblom, “ The spermatid micronucleus test with the dissection technique detects the germ cell mutagenicity of acrylamide in rat meiotic cells,” Mutat. Res. 309(2): 255 (1994). Tarcher, A. B., ed., Principles and Practice of Environmental Medicine, Plenum Medical, New York, 1992. Teaf, C. M., “ Mutagenesis,” in Industrial Toxicology, P. L. Williams and J. L. Burson, eds., Van Nostrand-Reinhold, New York, 1985. Thilly, W. G., and R. M. Call, “ Genetic toxicology,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 3rd ed., C. D. Klaassen, M. O. Amdur, and J. Doull, eds., Macmillan, New York, 1986. Tweats, D. J., “ Mutagenicity,” in General and Applied Toxicology, B. Ballantyne, T. Marrs, and P. Turner, eds., Stockton Press, New York, 1995. Waters, M. D., H. F. Stack, J. R. Rabinowitz, and N. E. Garrett, “ Genetic activity and pattern recognition in test battery selection,” Mutat. Res. 205: 119 (1988). Weisberger, J. H., and G. M. Williams, “ Chemical carcinogenesis,” in Casarett and Doull’s Toxicology: The Basic Science of Poisons, 2nd ed., J. Doull, C. D. Klaassen, and M. O. Amdur, eds., Macmillan, New York, 1980.

13 Chemical Carcinogenesis CHEMICAL CARCINOGENESIS


There are few people living today who have not been affected in some way by cancer, through either personal experience or that of a family member. Current statistics indicate that one out of two men and one out of three women in the United States will develop cancer over the course of their lifetime. Approximately 1.2 million people will be diagnosed with cancer this year alone, and this number excludes common and easily treatable basal and squamous cell skin cancers. While long-term survival rates are improving, cancer is still the second leading cause of death in the United States behind heart disease. In 1999, one of every four deaths or approximately 560,000 will be from cancer. In addition to the price cancer exacts in human lives lost, economic costs are estimated to be a staggering 107 billion dollars per year. This figure includes direct medical expense as well as the cost of lost productivity due to increased morbidity and early death. Clearly, there are many reasons for modern society to be concerned about cancer. The disease we call cancer is actually a family of diseases having the common characteristic of uncontrolled cell growth. In normal tissue, there are a myriad of regulatory signals that instruct cells when to replicate, when to enter a resting state, and even when to die. In a cancer cell these regulatory mechanisms become disabled and the cell is allowed to grow and replicate unchecked. Cancer is largely a disease of aging. The overwhelming majority of cancers are first diagnosed when patients are well over the age of 50. Carcinogenesis, or the sequence of events leading to cancer, is a multistep process involving both intrinsic and extrinsic factors. We know this because certain individuals inherit a genetic predisposition to certain types of cancer. The majority of cancers, however, are not associated with any particular inheritance pattern. Still, many of the same steps have been implicated. These incremental steps typically occur over the span of decades. At the most fundamental level, cancer is caused by abnormal gene expression. This abnormal gene expression occurs through a number of mechanisms, including direct damage to the DNA and inappropriate transcription and translation of cellular genes. Carcinogenesis has been shown to be induced or at least accelerated by exposure to certain types of chemicals. These chemicals are known as carcinogens. In the pages that follow, we will discuss the carcinogenic process and how chemicals can contribute to that process. This chapter will discuss:

• • • • • • •

Tumor classification and nomenclature Properties of carcinogenic chemicals An overview of the molecular basis of carcinogenesis Methods for testing chemicals for carcinogenic activity Chemicals identified as human carcinogens Risks associated with occupational carcinogens Factors that modulate carcinogenic risk

Principles of Toxicology: Environmental and Industrial Applications, Second Edition, Edited by Phillip L. Williams, Robert C. James, and Stephen M. Roberts. ISBN 0-471-29321-0 © 2000 John Wiley & Sons, Inc.




13.1 THE TERMINOLOGY OF CANCER Like most scientific disciplines, carcinogenesis has its own language. This section will familiarize you with some of the terms that you will encounter in your study of cancer. Since cancer is a disease that is characterized by uncontrolled or disregulated cell growth, most of the following terms relate to cell growth and differentiation. Anaplasia. Literally means “ without form.” Characterized by a marked change from a highly differentiated cell type to one that is less differentiated or more embryonic in nature. Thus, anaplastic tissue is less organized and functional than is the normal tissue. Anaplasia probably occupies the borderline between dysplasia and neoplasia. Benign. A term applied to neoplasms that are localized and encapsulated. Growth generally occurs via expansion and compression of adjacent tissue. Growth is generally slow, and there may be regression. If the growth is progressive, it is usually orderly and uniform. Dysplasia. A reversible change in cells, which may include an altered size, shape, and/or organizational relationship. This change usually affects epithelium and often results from chronic irritation. Hyperplasia. Increased organ or tissue size due to an increase in cell number. Hyperplasia may be physiological (e.g., tissue development and wound healing) or pathological (e.g., nodular liver regeneration in chronic alcoholics). Several hallmarks distinguish neoplasia from hyperplasia: Neoplasia Growth in excess of needs Purposeless Persistent Irreversible Autonomous

Hyperplasia Not excessive to needs Purposeful Ceases when stimulus ceases Reversible Regulated

Malignant. A term applied to neoplasms that are locally invasive. Growth may be rapid and is disorderly and progressive. Malignant neoplasms often have areas of necrosis. May spread by extension or metastasis. Metaplasia. A reversible change in which one differentiated cell type is replaced by another cell type. Metastasis. Presence of a disease process (usually cancer) at a site distant from the site of origin (the primary tumor). Metastasis (v. metastasize, adj. metastatic) is the primary hallmark of malignancy. Neoplasia. Literally means “ new growth.” Often used synonymously with cancer. The pathologist R. A. Willis offered the following definition, “ A neoplasm is an abnormal mass of tissue, the growth of which exceeds and is uncoordinated with that of normal tissues and which persists in the same excessive manner after cessation of the stimuli which evoked the change.” Tumor. A mass or swelling; one of the cardinal signs of inflammation. By common usage, a tumor is specifically a neoplasm.

13.2 CLASSIFICATION OF TUMORS Neoplasms can be formed in any tissue, and a variety of benign and malignant tumors can occur throughout the body. These tumors are classified using a binomial system based on (1) the tissue or cell type of origin and (2) their actual or predicted behavior (i.e., benign or malignant).



Histogenesis: Tissue Origin Nearly all tumors arise from either epithelial (ectodermal or endodermal) or mesenchymal (mesodermal) tissues. Epithelial tissues include lining epithelium (e.g., skin epidermis and the epithelium of the gastrointestinal tract and urinary system) and glands (e.g., pancreas, liver, mammary, prostate, sweat, and sebaceous glands). Mesenchymal or connective tissues include cartilage, bone, muscle, lymphoid, and hematopoietic cells.

Behavior of Neoplasm The distinction between benign and malignant tumors is extremely important because the malignancy of a tumor is typically what defines human cancer as a disease state. As noted previously, the ability to metastasize is the definitive characteristic of a malignant neoplasm. Not all malignant tumors metastasize (e.g., CNS tumors, intraocular tumors), but no benign tumors do. Metastasis is the major cause of morbidity and mortality associated with cancer. Treatment is much less likely to be successful once metastasis has occurred. In contrast to malignant tumors, benign neoplasms do not grow beyond their boundaries. However, benign tumors can inflict damage by localized obstruction, compression, interference with metabolism, and even secretion of unneeded hormones. Once a benign tumor is removed, so is its harmful influence. Table 13.1 provides further distinction between benign and malignant neoplasms.

Combined Classification Benign neoplasms, whether arising from epithelial or mesenchymal tissues, are named by adding the suffix oma to the tissue type. A benign epithelial neoplasm is an adenoma. This can be made more specific by the addition of modifiers that refer to the tissue of origin (e.g., mammary adenoma). There are many exceptions that are too rooted in common usage to be eliminated. For example, melanoma usually refers to a malignancy of melanocytes; the preferred term is malignant melanoma. Lymphoma usually refers to a malignancy of lymphoid tissue; the preferred term is lymphosarcoma. Malignant neoplasms arising from epithelial tissues are named by adding carcinoma to the tissue or cell of origin (e.g., thyroid carcinoma). The prefix -adeno is added if the tumor is gland forming (e.g., thyroid adenocarcinoma). Malignant neoplasms arising from mesenchymal tissues are named by adding sarcoma to the tissue or cell of origin (e.g., fibrosarcoma). Table 13.2 provides several examples of tumors classified using this scheme.

TABLE 13.1 Distinctions between Benign and Malignant Neoplasms Benign Well differentiated; resembles a cell of origin Grows by expansion Well circumscribed, often encapsulated by a peripheral rim of fibrous tissue Grows at a normal rate Few mitotic figures Growth may be limited Does not metastasize, seldom dangerous Adequate blood supply

Malignant Poorly differentiated or anaplastic Grows by expansion and infiltration Poorly circumscribed and invades stroma and vessels Growth rate usually increased Frequent mitotic figures Progressive growth Metastasis usual; can be fatal Often outgrows blood supply, becomes necrotic



TABLE 13.2 Some Examples of Tumor Classification and Nomenclature Tissue (Cell) Origin Epithelial Biliary tract Liver cells Lung Mammary gland Squamous epithelium Mesenchymal Blood vessels Bone Fibroblasts Hematopoietic cells Fat Striated muscle Smooth muscle



Cholangioma Hepatocellular adenoma Pulmonary adenoma Mammary adenoma Papilloma

Cholangiocarcinoma Hepatocellular carcinoma Pulmonary carcinoma Mammary adenocarcinoma Squamous cell carcinoma

Hemangioma Osteoma Fibroma No benign form recognized Lipoma Rhabdomyoma Leiomyoma

Hemangiosarcoma Osteosarcoma Fibrosarcoma Leukemia Liposarcoma Rhabdomyosarcoma Leiomyosarcoma

13.3 CARCINOGENESIS BY CHEMICALS The following definitions should help clarify subsequent discussions about chemical-induced carcinogenesis.

• A carcinogen, as defined in this chapter, is a chemical capable of inducing tumors in animals or humans.

• Carcinogenesis is the origin or production of cancer; operationally speaking, this includes any tumor, either benign or malignant.

• A direct-acting or primary carcinogen is a chemical that is reactive enough to elicit carcinogenic effects in the parent, unmetabolized form. Often these chemicals produce tumors at the site of exposure (e.g., alkylating agents, radiation).

• A procarcinogen is a chemical that requires metabolism or bioactivation to another chemical form before it can elicit carcinogenic effects (e.g., polycyclic aromatic hydrocarbons).

• A cocarcinogen is a chemical that increases the carcinogenic activity of another carcinogen when coadministered with it. While not carcinogenic itself, the agent may act to increase absorption, increase bioactivation, or inhibit detoxification of the carcinogen administered with it.

Early Epidemiologic Evidence Some of the earliest evidence that exposure to chemicals could play a role in the development of cancer came from the observations of the British physician Sir Percivall Pott. In 1775, Pott reported on a relationship between scrotal cancer and occupation in men who in their youth, had been employed as chimney sweepers. He suggested that the soot to which these men were exposed on the job played a causal role in the development of their cancer. Over the next century, mounting evidence implicated other chemicals and industrial processes in human cancer. In 1884, Bell and Volkman independently reported on the increased prevalence of skin cancer in workers who were exposed to oils that were distilled from coal and shale. Sporadic reports of other occupational exposures and cancer began appearing in the literature. In 1895, Rehn reported on bladder cancer in aniline dye workers in Germany.



In the early 1900s, radiation was associated with lung tumors in uranium miners and in skin tumors and leukemia in technicians working with the recently discovered X rays. If it were true that exposure to chemicals could cause cancer in humans, researchers at the beginning of the twentieth century were hopeful that they could be identified and their mechanisms of action studied. Unfortunately, and a bit ironically, early attempts at reproducing cancer in laboratory animals were fruitless. Experimental validation of Pott’s original hypothesis finally came in 1915, when the Japanese pathologists Yamagiwa and Ichikawa reported that rabbits developed malignant skin tumors following repeated topical applications of coal tar. The line of research that Pott began was brought full circle when, in the 1930s, a group of investigators led by Cook and Kennaway implicated polycyclic aromatic hydrocarbons as putative carcinogens in coal tar and other industrial oils. They identified benzo[a]pyrene as the first carcinogenic hydrocarbon of known structure isolated from coal tar. Soon afterward, the structures of other chemical carcinogens were identified and their carcinogenic effects replicated in experimental animal models. The Somatic Mutation Theory In the early 1900s very little was known about the mechanism of cancer induction by chemicals. Theodor Boveri is often credited with the proposition that cancer involved a permanent alteration of the genetic material in somatic cells. In what became known as the somatic mutation theory, he attributed cancer to an “ abnormal chromatin complex, no matter how it arises. Every process which brings about this chromatin condition would lead to a malignant tumor.” It is important to point out, however, that this theory had as its basis only gross morphological observations of cancer cells. It was only after the pioneering work of Watson and Crick in 1950s, that it became evident that interference with DNA basepairing could be a mechanism by which chemicals induced mutation. A great deal of evidence has now accumulated in support of the somatic mutation theory (i.e., a genetic mechanism of cancer). Since the early 1970s many carcinogens have been shown to produce permanent, heritable changes in DNA. It has further been shown that these changes are involved in the carcinogenic process. Smart (1994) has described several lines of evidence that support a genetic mechanism for cancer.

• Cancer is a heritable change at the cellular level. • Tumors are generally clonal in nature. • Many carcinogens or their activation products can form covalent bonds with DNA and produce mutations.

• The inheritance of certain recessive mutations in genes associated with genomic integrity predisposes affected individuals to cancer.

• Most cancers display chromosomal abnormalities. • The phenotypic characteristics of a tumor cell can be transferred to a nontumor cell by DNA transfection. Initiation and Promotion Since the eighteenth century, investigators have realized that the carcinogenic process involved a period of latency following exposure to chemical carcinogens prior to the appearance of any clinical symptoms. Early attempts at inducing cancer with chemicals in experimental animals met with little success. It was not until the work of Yamigawa and Ichikawa that it was realized that the generation of cancer in experimental animals required long-term repeated exposures to chemicals. With this discovery, researchers quickly developed animal models with which to test the carcinogenic potency of chemicals and mixtures. The most common was the mouse skin model, in which repeated applications of a potential carcinogen were applied to the shaved back of a mouse and the number of tumors generated and the time required for their development was recorded. In the early 1940s,



researchers working with rodent skin models discovered that croton oil could stimulate the rapid development of tumors but only if it was applied after treatment with polycyclic aromatic hydrocarbons. Rous and coworkers were the first to use the terms initiation and promotion to describe two stages of carcinogenesis observed in the experimental induction of skin tumors in rodents. The term progression was added later to describe the sequence of events leading to the development of malignant tumors. These three stages are remarkably similar to those described in the development of human skin cancer, induced by soot and paraffin oils. It is now known that initiation, the first step in this process, involves an irreversible mutation in the DNA of a somatic cell. Chemical initiators are usually electrophiles or metabolically activated to electrophiles (see Chapter 3 for a discussion of metabolic activation). These chemicals bind to nucleophilic centers in DNA, forming DNA adducts. If the DNA is replicated prior to repair of an adduct, a mutation can be “ fixed” in the DNA of the daughter cell. This mutation essentially primes the cell for later steps in neoplastic development. Most initiators are mutagens and are thus classified as genotoxic carcinogens. Chemicals that act as tumor promoters may not, by themselves, be carcinogenic. However, if given subsequent to an initiating agent, they increase either the number of tumors or decrease the latency period or both. Tumor promoters typically do not bind DNA; rather, they allow for the clonal expansion of initiated cells by providing a selective growth advantage. For this reason, promoters are considered epigenetic (or nongenotoxic) carcinogens. For example, the active ingredients of the first tumor promoter, croton oil, are phorbol esters. These compounds mimic endogenous molecules that trigger cell proliferation, thus allowing initiated cells to proliferate. Progression, the third stage of the experimental carcinogenic process, is less well characterized. In general, progression is thought to involve the accumulation of further genetic alterations in a population of initiated cells that have been provided a growth advantage through promotion. These changes ultimately lead to a malignant tumor (Figure 13.1). A concept that is important in the context of neoplastic progression is that of tumor cell heterogeneity. Investigators studying leukemias and lymphomas have demonstrated that these cancers are almost universally clonal in origin. While there is also evidence of this type of clonal origin in solid tumors (i.e., carcinomas and sarcomas), by the time clinical signs of cancer are evident, the cells that make up tumor have usually developed a certain amount of genotypic and phenotypic diversity. Many researchers believe that the cellular heterogeneity observed in these tumors is the result of genetic instability acquired during tumor progression. Genetic instability suggests that the DNA in tumor cells is mutated at higher rates than the surrounding normal cells rapidly producing subclones. Some of these clones would have adaptations, such as the ability to escape the host’s defense mechanisms or invade surrounding tissue that give them a selective advantage. These clones eventually grow to dominate the tumor population. Multiple rounds of this type of selection lead to populations of cells that are increasingly abnormal on a genotypic and phenotypic level and as a result, more aggressive and invasive. While the initiation-promotion model was first described in a rodent skin model, the process has been experimentally reproduced in other organs such as liver, colon, lung, prostate, and mammary gland. Our experience with both initiators and promoters indicates that many are organ specific. There are, however several features of the initiation–promotion model that remain relatively constant (Figure 13.2).

• Exposure to a sub-threshold dose of an initiator alone results in few, if any tumors. • Exposure to a sub-threshold dose of an initiator followed by repeated exposure to a promoter results in many tumors.

• Exposure to a promoter will produce tumors even if there has been a latent period following exposure to an initiator. Thus, initiation is irreversible.

• In contrast, if initiation is not followed by promotion of sufficient duration, no tumors are produced. Thus, the effects of tumor promoters are reversible in the early stages.


Figure 13.1 Schematic diagram of the development of a malignant tumor from a normal cell.



Figure 13.2 Generalized scheme of initiation-promotion experiments. Initiation is caused by a single dose of an initiating agent such as a carcinogenic polycyclic aromatic hydrocarbon; promotion is carried out by repeated application or chronic dosing with a tumor promotor such as TPA. (I, initiator; P, promoter; solid line indicates continual application of agent; dotted lines indicate the duration of time without exposure to an agent.)

• Initiation must occur prior to promotion. • Repeated exposure to a promoter alone will result in few, if any tumors. Experiments with initiation–promotion models have demonstrated that there are chemicals that possess both initiator and promoter activity. These chemicals are known as complete carcinogens. By the same token, chemicals that cannot by themselves induce cancer in experimental animal models are called incomplete carcinogens. In reality, dosage is a critical factor in determining whether a chemical is a complete carcinogen. At sufficiently low doses, most initiators require subsequent promotion for the development of a tumor, while at very high doses, most carcinogens possess initiating and promoting ability. As you will learn later, this has important implications for the identification and the assessment of risk associated with potential carcinogens.

Electrophilic Theory The nature of the initiation step in chemical carcinogenesis was the subject of much scientific inquiry and debate for decades. Until 1940, the only known chemical carcinogens were aromatic hydrocarbons and amines. Soon afterward, other aliphatic chemicals were also shown to be carcinogenic and by the 1960s the various chemical carcinogens belonged to over a dozen chemical classes (Figure 13.3). Attempting to explain this structural diversity, in 1969 James and Elizabeth Miller hypothesized that “ most, if not all, chemical carcinogens either are, or are converted to, reactive electrophilic derivatives which combine with nucleophilic growth crucial tissue components, such as nucleic acids or proteins.” In what became known as the electrophilic theory of chemical carcinogenesis, the Millers described the metabolic activation of inactive procarcinogens to intermediates they called proximate carcinogens and on to ultimate carcinogens that covalently bind DNA and cause mutations. Examples of chemical carcinogens that require metabolic activation include benzo[a]pyrene and other polycyclic aromatic hydrocarbons, 1,3-butadiene, and 2-acetylaminofluorene. The bioactivation of several chemical carcinogens is illustrated in Figure 13.4.



Carcinogenicity and Mutagenicity The relationship between mutagenicity and carcinogenicity as it related to the effects of ionizing radiation has been known since the early part of the twentieth century. Sufficient doses of ionizing radiation produce large structural alterations in the genetic material. The first chemical mutagens to be identified were the nitrogen and sulfur mustards (mustard gas). These chemicals were studied because their biological effects were similar to those of ionizing radiation. The nitrogen mustards and other radiomimetic compounds have the ability to form crosslinks between strands of DNA or between DNA and proteins. When such lesions are not repaired, large-scale alterations in the DNA can occur. These types of effects are referred to as clastogenic, and chemicals that cause them are known as clastogens. The identification of DNA as the genetic material along with the proposition that many chemical carcinogens are, or become, metabolically activated to reactive electrophiles paved the way for the identification of a strong link between chemical mutagenicity and carcinogenicity. As described in Chapter 12, there are now many short-term tests for the identification of chemical mutagens. The most widely used is the Salmonella mutation assay (Ames assay). It is important to note that the combination of this assay with mammalian metabolic activation systems was a critical step in the identification of a large number of chemical carcinogens as mutagens. Analyses conducted by Tennant and Ashby (1991) indicate that there is an approximately 60 percent concordance between carcinogenicity in rodent bioassays and mutagenicity in the Salmonella assay. In other words, the Salmonella assay predicts whether a chemical will be a carcinogen in rodents approximately 60 percent of the time. These investigators also identified a number of structural “ alerts” that are indicative of potential mutagenicity and carcinogenicity. Initially, it had been hoped that such short-term tests would be highly predictive of carcinogenicity in rodents, and by extension, humans. However, much experimental evidence now suggests that there are a large number of chemicals that do not act via a genotoxic mechanism and thus will not be detected by such short-term tests.

Genotoxic and Epigenetic Carcinogens With the realization that carcinogens elicited their effects via diverse mechanisms, a number of classification schemes based on carcinogenic mechanism were developed. One popular scheme was offered by Williams (Table 13.3). This scheme divides chemical carcinogens into two broad categories based on whether they act in a genotoxic fashion. Accordingly, the two main groups of carcinogenic chemicals are

• Genotoxic carcinogens • Epigenetic (or nongenotoxic) carcinogens1 Genotoxic carcinogens are those chemicals that are capable of modifying the primary sequence of DNA (i.e., initiators). This group includes chemicals that induce mutational and clastogenic changes or changes in the fidelity of DNA replication. Epigenetic carcinogens do not alter the primary sequence of DNA; instead they can affect cell proliferation and differentiation by a number of mechanisms including cytotoxicity and compensatory cell proliferation, receptor mediated events, and by altering the expression or repression of certain genes and cellular events related to cell proliferation and differentiation. It is estimated that at least 40 percent of carcinogens identified by rodent bioassays elicit their affects via an epigenetic mechanism. Many epigenetic agents favor the proliferation of cells with altered genotypes due to an interaction with an initiating carcinogen. While epigenetic carcinogens and tumor promoters share many of the same characteristics, there is some debate regarding whether classic tumor promoters should be considered carcinogens at all. However, since tumor promoters can 1

The term epigenetic carcinogen may seem confusing since many of these chemicals have the ability to alter the regulation of critical genes. This is why some prefer the term non-genotoxic carcinogen because it implies that the chemical does not alter the primary sequence of DNA by direct interaction. The two terms are often used synonymously and for our purposes we will use the term epigenetic carcinogen.



Figure 13.3 Chemical structures of some representative chemical carcinogens. [Adapted from Cancer Biology, Third Edition by Raymond W. Ruddon. copyright © 1981, 1987, 1995 by Oxford University Press, Inc. Used by permission of Oxford University Press, Inc.]


Figure 13.3 (Continued)




Figure 13.4 Metabolic activation of several representative genotoxic carcinogens.



TABLE 13.3 Classification of Carcinogenic Chemicals Type Genotoxic carcinogens Direct-acting or procarcinogenic Inorganic carcinogenic Epigenetic carcinogens Solid-state Hormonal


Cocarcinogenic Promoter


Possible or Probable Mechanism of Action


An electrophile, the compound alters the genetic code via mutagenic or clastogenic processes Alters the fidelity of DNA replication

Bis(chloromethyl) ether, nitrosamines, benzanthracene, epoxides, dimethyl sulfate, nitrosoureas Cadmium, chromium, nickel

Mechanical disruption of tissue Disrupts cellular dedifferentiation, promotes cellular growth

Asbestos, metal foils, plastic Estrogens, androgens, thyroid hormone, tamoxifen, diethylstilbestrol Azathioprine

Depression of the immune system allows the proliferation of initiated cells or tumors Modifies the response of genotoxic carcinogens when coadministered Enhances cell growth, promotes response to initiator or genotoxic carcinogen Increases the rate of spontaneous mutation, promotes regenerative cell growth

Ethanol, solvents, catechol Phorbol esters, catechol, ethanol

Trichlorethylene, carbon tetrachloride, chloroform

Source: Adapted from Weisberger and Williams (1981).

affect the proliferation of cells that have spontaneous as well as chemically induced mutations, we will group them with the epigenetic carcinogens. The mechanism of a chemical’s carcinogenic action affects the manner in which it is treated for regulatory purposes. Based on early theories, it was assumed that even a single molecule of a genotoxic chemical could irreversibly damage DNA and that each additional exposure could add to the damage from previous exposures. Thus, regulators initially assumed that there is no safe level of exposure or “ threshold” below which harmful effects do not occur. Now, however, many carcinogens, particularly epigenetic carcinogens are thought to elicit their effects in a manner consistent with a threshold.

Epigenetic Mechanisms Although the precise mechanisms of carcinogenesis by epigenetic chemicals are unknown, progress is being made toward understanding the organ- and species-specific effects of certain classes of epigenetic carcinogens. Some effects of these chemicals appear to be mediated by cytotoxic insult and a compensatory regenerative response while others have been shown to act via receptor-mediated events resulting in the altered transcription of critical cellular genes. Alteration of DNA methylation status such that critical genes are expressed or are inactivated inappropriately is another epigenetic mechanism that is currently receiving increased attention, although as yet, it is uncertain how chemicals might affect this process. Some chemicals that are not directly genotoxic, but are carcinogenic in chronic rodent bioassays have been shown to exhibit cytotoxic properties. It has been shown that many of these chemicals produce necrosis or cell death due to cytotoxicity at the target organ. This is usually followed by regenerative cell proliferation. The organ- or cell-type-specific effects of carcinogens that induce cytotoxicity may be due to the high concentration of the chemical at that organ or the selective toxicity directed at specific cell populations. Cells in the target tissue could become initiated following an insult



with a cytotoxic chemical due to (1) spontaneous mutations produced by defective mitosis or inefficient repair occurring during multiple rounds of DNA replication during regenerative cell proliferation or (2) generation of DNA damage by oxygen free radicals, produced by lipid peroxidation or by recruited inflammatory cells. Such initiated cells would then have a selective growth advantage due to the production of various stimuli (e.g., growth factors) produced by proliferating cells. Examples of chemicals that may work via these types of mechanisms include chloroform, carbon tetrachloride, and saccharine. There is an increasing amount of evidence that suggests that many epigenetic carcinogens activate receptors and as a result, elicit changes in the expression of critical target genes involved in cellular functions ranging from signal transduction, cell proliferation, and differentiation to apoptosis and cell-to-cell communication. Some of these chemicals and their receptors are shown in Table 13.4. The changes in gene expression induced by certain epigenetic carcinogens may mimic the effects of endogenous growth factors and hormones that similarly affect these cell functions. The effects of natural hormones, growth factors, and dietary constituents on promotion of neoplasia suggest that endogenous tumor promotion and epigenetic carcinogens have common links. The tissue-specific effects of some chemicals may be due to the predominance of particular signaling pathways in a given cell or tissue type. Because of the increased understanding of nongenotoxic mechanisms of action of epigenetic carcinogens, endpoints such as proliferation, differentiation, apoptosis, cell-to-cell communication, and the induction of gene expression have all been used in the assessment of epigenetic carcinogens. Experimental evidence that some of these compounds elicit their effects via receptors provides a solid basis for the contention that there exists a threshold level under which epigenetic carcinogens would not exert carcinogenic activity. Another epigenetic mechanism that received more attention in the late 1990s is the alteration of DNA methylation patterns in neoplastic cells. Most cells in the body contain the same genetic information. Yet somehow, different cell types express only a subset of that genetic code that is required for the cell to function properly. Cell differentiation is almost always achieved without altering the primary sequence of DNA, yet the phenotypic characteristics of the cell are usually stable and can be passed on to daughter cells during cell division. Much of the control of the gene expression that ultimately determines cell phenotype is maintained by the addition of methyl groups to the 5′ carbon of cytosine residues in cellular DNA, particularly at CpG dinucleotide sequences. The promoter and enhancer elements of many genes have regions high in CpG dinucleotides. There is an inverse correlation between gene expression levels and the degree of methylation in these regions. That is, actively transcribed genes have low levels of methylation (hypomethylation) in their promoter regions while transcriptionally silent genes have heavily methylated (hypermethylation) promoter regions. It has been hypothesized that changes in methylation status play an important role in the neoplastic progression of tumor cells. Once in place, changes in methylation status could be passed to daughter as permanent epigenetic changes. There is evidence for this type of mechanism in the inactivation of the p16 tumor suppressor gene in human tumors by hypermethylation of the promoter region. It is still unclear how epigenetic carcinogens might affect this mechanism of gene regulation, but it has become increasingly apparent that alteration of DNA methylation patterns do play a role in the progression of some tumors.

TABLE 13.4 Some Epigenetic Carcinogens and the Receptors They Activate Chemical Tetrachloro dibenzo-p-dioxin (TCDD) 12-o-Tetradecanoylphorbol-13-acetate (TPA) Peroxisome proliferating compounds Estrogenic compounds Okadaic acid

Receptor Ah receptor Protein kinase C Peroxisome proliferator-activated receptor (PPAR) Estrogen receptor Protein phosphatase-2A



13.4 MOLECULAR ASPECTS OF CARCINOGENESIS To this point we have discussed some of the critical observations that have contributed to our understanding of the roles chemicals play in the carcinogenic process. From these observations we know that a majority of chemical carcinogens produce mutations in DNA, presumably in critical genes. We also know that there are other chemicals that alter cell growth and differentiation through epigenetic mechanisms. Until relatively recently, very little was understood about the nature of genes involved in carcinogenesis. As you will learn, the discovery of oncogenes and tumor suppressor genes as positive and negative regulators of cell growth, respectively, unified many of the earlier observations and theories about the nature of the carcinogenic process. These discoveries have been followed by further refinements in our understanding of the molecular basis of cell growth and differentiation and how these processes are subverted in the development of cancer. Oncogenes Thus far, our study of carcinogenesis has focused on information gathered by researchers studying the induction of cancer by chemicals. There was however, another group of cancer researchers who believed that infectious agents, namely viruses were actually the cause of cancer. While we now know that this is not the case for the overwhelming majority of human cancer, the work of these investigators has provided us with some of the most important information on the molecular mechanisms of cancer. Peyton Rous and his coworkers discovered the first known tumor virus in 1909. They demonstrated that virus particles extracted from a sarcoma in one chicken could produce similar tumors when injected into other chickens. The Rous sarcoma virus as it became known was eventually shown to be an oncogenic (from the Greek word onkos—mass or swelling) retrovirus. In the decades that followed, many other oncogenic retroviruses were discovered. These viruses were active in other avian species, rodents, and even primates. Retroviruses encode their genetic material in RNA rather than DNA like other organisms. The genome of the retrovirus is limited to several genes that are critical to the production of more virus particles. Following infection, the viral RNA is transcribed into DNA by the enzyme reverse transcriptase. The newly created double stranded DNA integrates into the DNA of the host cell, where the strong promoter sequences of the virus induce the host cell’s nuclear machinery to express the viral genes and produce new viral particles. Some oncogenic retroviruses, such as the Rous sarcoma virus, have the ability to rapidly “ transform” normal cells into cancer cells. Researchers working with these so called acute transforming retroviruses hypothesized that they contained a one or more genes responsible for their rapid transforming ability. Eventually, a gene responsible for the transforming ability of the Rous sarcoma virus was discovered. It was named src for sarcoma, and it was the first oncogene or gene capable of inducing cancer identified. In the years that followed, oncogenes from other acute transforming retroviruses were identified, each of these genes was named with a three-letter identifier that corresponded to the virus from which they were first isolated: ras from the rat sarcoma virus, myc from the avian myelocytomatosis virus, sis from the simian sarcoma virus, and fes from the feline sarcoma virus. It was ultimately demonstrated that the oncogenes responsible for the transforming ability of the oncogenic retroviruses had normal, highly conserved counterparts in the cells of a wide variety of prokaryotic and eukaryotic organisms. It stood to reason that if these genes had been conserved over so many years of evolutionary history, they must play an absolutely critical role within the cell. These normal cellular genes appear to have been transduced or in essence “ stolen” by retroviruses. Following transduction, the cellular genes became “ activated,” that is, altered in ways that made them oncogenic. The method of activation was not the same for each oncogene. Some like erb B were activated by a deletion of several hundred basepairs of DNA from the coding region of the gene, some like myc were activated by gene amplification, while others like ras were activated by a single-point mutation. The normal cellular counterparts to the oncogenes of the transforming retroviruses were called protooncogenes. Subsequently protooncogenes were shown to be activated in a number of human tumors. Many



lines of experimental evidence converged when it was shown that oncogenes that had been activated by mutation with a carcinogenic chemical could transform normal cells into cancer cells. We now know that the same oncogenes are often activated in tumors of the same cell type, whether the tumors arose spontaneously or were virally or chemically induced. A word about nomenclature should be mentioned here. The student may encounter several notations associated with the names of oncogenes (oncs). The notation v-onc (e.g., v-sis) is used to distinguish an oncogene of viral origin from a similar oncogene of cellular origin (c-onc, e.g., c-sis). It should also be noted that not all oncogenes have been discovered in transforming retroviruses.

Oncogenes and Signal Transduction Pathways To date, approximately 75 cellular oncogenes and their protooncogene counterparts have been identified. The protein products of nearly all of these genes function in one way or another in cellular signal transduction pathways to precisely regulate cell growth and differentiation. Signal transduction pathways are used by cells to receive and process information and ultimately to effect a biological response. These pathways generally consist of external signaling molecules, receptors on the cell surface, transducer proteins, second messenger proteins, amplifier proteins, and effector proteins such as transcription factors, all of which are involved in the regulation of cellular function or gene expression. A generalized signal transduction pathway is shown in Figure 13.5. The protein products of oncogenes have been grouped according to their function in several different categories (Table 13.5). These categories include growth factors, growth factor receptors, membrane-associated GTP binding proteins (G proteins), nonreceptor tyrosine kinases, cytoplasmic serine/threonine kinases, and nuclear transcription factors. Activation of a protooncogene confers a gain of function to the gene product in the sense that its ability to promote cell proliferation is enhanced. Mechanisms of activation of

Figure 13.5 Schematic diagram of growth factor-mediated signal transduction pathways.



TABLE 13.5 Selected Protooncogenes and the Functions of Their Encoded Proteins Oncogene Name

Function Growth Factorsa

sis int-2 tgf-α

Platelet-derived growth factor Fibroblast growth factor Transforming growth factor-α Growth Factor Receptorsb

erbB fms kit met

Epidermal growth factor receptor (tyrosine kinase) Colony stimulating factor receptor (tyrosine kinase) Stem cell receptor (tyrosine kinase) Hepatocyte growth factor receptor (tyrosine kinase) GTP Binding Proteins (G proteins)c


Membrane-associated GTP binding/GTPase Nonreceptor Tyrosine Kinasesd

src yes abl fes

Membrane associated—mediates integrin signaling Membrane associated Cytoplasmic with nuclear translocation ability—DNA binding and DNA transcription activation Cytoplasmic Cytoplasmic Serine/Threonine Kinasese

raf mos

Phosphorylates MAPKK proteins in cell signaling Activates and/or stabilizes maturation promoting factor (MPF) Nuclear Transcription Factors f

myc fos jun ets a

Sequence-specific DNA binding protein (transcription factor) Combines with jun to form AP1 transcription factor Combines with fos to form AP1 transcription factor Transcription factor

These are secreted factors that typically act in an autocrine or paracrine fashion. Normally these receptors are transiently activated by ligand binding. Mutant forms are persistently activated. c Numerous growth factor receptors normally signal through GTP binding proteins. These proteins transiently activated in response to ligand binding at the receptor. Mutant forms are persistently activated. d These are cytoplasmic proteins involved in the relay of signals from growth factor receptors and from the extracellular matrix through cytoskeletal proteins. Activation requires differential, transient phosphorylation of tyrosine residues. Mutant forms are persistently activated. e These are another group of cytoplasmic proteins involved in the relay of signals to the cell nucleus. Activation requires differential, transient phosphorylation at serine and threonine residues. Mutant forms are persistently activated. f These proteins are localized primarily to the cell nucleus; where they function to transcriptionally activate and repress genes associated with cell growth and differentiation. b



protooncogenes that commonly occur in human tumors include point mutation, gene rearrangement, gene amplification, chromosomal translocation, and increased transcription (Table 13.6). Tumor Suppressor Genes Demonstration of the existence of cellular oncogenes and knowledge of their function as positive regulators of cell growth provided an obvious mechanism by which chemicals could induce the carcinogenic process. The thinking was that an activated oncogene could force the cell and its descendants into unneeded rounds of division ultimately resulting in a tumor. However, there was a problem with such a simplistic view. Researchers soon demonstrated that when tumor cells were fused with normal cells, the resulting hybrid cells were usually nontumorigenic. Thus the transforming ability of oncogenes could be reversed or controlled by some other factor produced by normal cells. It was eventually discovered that normal cells carried genes that coded for proteins that function as negative regulators of cell growth. These genes came to be called tumor suppressor genes. There now exists much evidence supporting the existence of tumor suppressor genes and their functions as negative regulators of cell growth. To date, approximately 20 putative tumor suppressor genes have been identified, although, for many of these, a function is not well understood. Like the oncogenes, the products of tumor suppressor genes appear to have diverse functions within the cell. These functions include cell cycle control, transcriptional regulation, regulation of signal transduction, maintenance of cellular structure, and DNA repair. Some tumor suppressor genes and the functions of the proteins they encode are shown in Table 13.7. In contrast to the situation with oncogenes where a mutation in only one allele is often transforming, the inactivation of tumor suppressor genes requires two genetic events, that is, the inactivation of both alleles. The mechanism most commonly invoked in tumorigenesis is a mutation in one allele followed by a subsequent deletion of the second allele or replacement of the second allele with a copy of the mutated allele, resulting in what is commonly known as loss of heterozygosity (LOH). Tumor suppressor genes are often linked to rare, inherited forms of cancer. In fact, the existence of tumor suppressor genes had been suggested as early as 1971 when Knudson forwarded the “ two hit” hypothesis, in which he proposed that the development of retinoblastoma, a rare tumor of the eye in children, required two genetic events. His work eventually led to the cloning of the retinoblastoma

TABLE 13.6 Oncogenes Activated in Human Tumors Oncogene abl erbB-1 erbB-2 (neu) gip gsp myc L-myc N-myc H-ras K-ras N-ras ret K-sam trk

Neoplasm(s) Chronic myelogenous leukemia Squamous cell carcinoma; astrocytoma Adenocarcinoma of the breast, ovary, and stomach Adenocarcinoma of the ovary and adrenal gland Thyroid carcinoma Burkitt’s lymphoma Carcinoma of lung, breast, and cervix Carcinoma of lung Neuroblastoma, small cell carcinoma of lung Carcinoma of colon, lung, and pancreas; melanoma Acute myelogenous and lymphoblastic leukemia; thyroid carcinoma, melanoma Carcinoma of the genitourinary tract and thyroid; melanoma Thyroid carcinoma Carcinoma of stomach Thyroid carcinoma

Lesion Translocation Amplification Amplification Point mutations Point mutations Translocation Amplification Amplification Amplification Point mutations Point mutations Point mutations Rearrangement Amplification Rearrangement



TABLE 13.7 Tumor Suppressor Genes in Human Cancer and Genetic Disease Gene

Consequence of loss


Retinoblastoma and osteosarcoma


Li-Fraumeni syndrome inactivated in >50% of human cancers

p16 Wt1

Familial melanoma, pancreatic cancer Wilms’ tumor/nephroblastoma


Von Hippel–Lindau syndrome renal cell carcinoma Neurofibromatosis type 1 schwannoma and glioma Neurofibromatosis type 2 acoustic nerve tumors and meningiomas Familial and sporadic breast and ovarian cancer, also prostate and colon cancers Breast cancer (female and male) also prostate cancer Colon cancer Familial and sporadic adenomatous polyposis colorectal tumors Hereditary nonpolyposis colorectal cancer


Function of encoded protein Binds and sequesters the transcription factor E2F to maintain cells in G0 of cell cycle Transcription factor with multiple functions, including cell cycle progression, detection of DNA damage, and apoptosis Inhibits CDK4 to block cell cycle progression Transcription factor required for renal development Negative regulation of hypoxia-inducible mRNAs GTPase-activating protein (GAP), which regulates signaling through ras Connects cell membrane proteins with the cytoskeleton Secreted growth factor Unknown function Cell adhesion molecule Interacts with catenins, proteins involved in signaling pathway for tissue differentiation Mediates DNA mismatch repair

gene (Rb) and the discovery that both copies of the gene are inactivated and/or deleted in retinoblastoma tumors. It is now known that a large proportion of persons with retinoblastoma have inherited a defective copy of the Rb gene. Tumors develop when the second copy is inactivated prior to the terminal differentiation of the retinoblasts. Another group of retinoblastoma patients do not have a defective copy of the Rb gene. In this group, two somatic mutations have occurred sometime after conception. Individuals born with a mutated copy of Rb gene are also at a higher risk of developing other cancers, most notably osteosarcoma, later in life. A number of the known or putative tumor suppressor genes appear to be involved in a relatively small subset of tumors specific to certain tissue types. These include Wt-1 (Wilms’ tumor), NF-1 and NF-2 (neurofibromatosis types 1 and 2), APC (adenomatous polyposis coli), and DCC (deleted in colon carcinoma). In contrast to these, the p53 tumor suppressor gene, is inactivated in more than 50 percent of all human tumors. The p53 protein is a remarkable protein that is involved in diverse cell functions including the detection of DNA damage, the regulation of cell cycle progression, and the induction of apoptosis or programmed cell death. Rb and p53 will be discussed briefly below as well as in the context of the cellular functions in which they are involved. The p53 gene and the protein it encodes has been called the “ guardian of the genome” in recognition of the critical role it plays in the life and death of cells. The p53 gene is considered to be the most frequently mutated gene in human tumors. Approximately 40 percent of breast cancers, 70 percent of colon cancers, and 100 percent of small cell lung cancers contain mutations in the p53 gene. The p53 gene encodes a 53-kD nuclear phosoprotein that is active in regulating the transcription of a number of genes relating to cell cycle progression and apoptosis. Levels of p53 are increased in response to several types of cell stress, including DNA damage, hypoxia, and decreases in the levels of nucleotide triphosphates required for DNA replication. The p53 protein has been shown to have a direct role in the detection of DNA damage by some chemical carcinogens and radiation. In the presence of DNA damage p53 has the ability to slow cell cycle progression or bring the cycle to a halt until the damage can be repaired. In the face irreparable damage p53 has been shown to initiate the events leading to apoptosis.



The Rb gene encodes a 107-kD nuclear protein that plays a critical role in the early stage of the cell cycle. When the cell is stimulated by a growth factor, that signal is ultimately relayed to the nucleus. This signal results in the production of proteins that temporarily inactivate Rb. In its active form, the Rb protein is tightly bound to an important transcription factor, E2F. When the cell receives a signal to divide, Rb is hyperphosphorylated, causing a conformational change and the release of E2F. The transcription factor E2F induces the production of other proteins involved in cell cycle progression.

The Cell Cycle and Apoptosis It is important to discuss some of the processes that govern the life and death of cells to better understand how oncogenes and tumor suppressor genes are involved in these processes. As pointed out previously, proto-oncogenes function in various capacities in the transduction of signals for cell growth and differentiation within and between cells. In normal cells, replication of the DNA and cell division is stimulated by the presence of growth factors that bind receptors at the cytoplasmic membrane and initiate a cascade of intracellular signals. Once these signals reach the nucleus they cause the transcription of a complex array of genes, producing proteins that mediate progression of the cell through the cell cycle culminating in mitosis or cell division. The cell cycle is divided into five phases (Figure 13.6). The length of each of these phases can vary depending on factors such as cell type and localized conditions within the tissue. After completing mitosis (M), daughter cells enter the Gap 1 (G1) phase. If conditions are favorable, cells enter the synthesis (S) phase of the cycle, where the entire genome of the cell is replicated during DNA synthesis. Following S phase, cells enter the Gap 2 (G2) phase before proceeding through mitosis again. There is a critical boundary early in G1 called the restriction point. This is the point at which the cell must

Figure 13.6 Schematic diagram of the cell cycle including primary checkpoints.



make a decision to (1) enter the cell cycle again or (2) move into a state of quiescence also known as G0 phase. Once in G0 phase, the cell can either remain in a state of replicative quiescence until it receives a signal to divide again or it can proceed down a path that leads either to terminal differentiation or to apoptosis. Movement of the cell through the cell cycle is controlled by an enormously complex network of proteins many of which are expressed in a phase-specific fashion. Several major groups of these proteins have been studied to date. These include cyclins, cyclin-dependent kinases (CDKs), cyclinactivating kinases (CAKs), and CDK inhibitory proteins. The cyclins and CDK proteins are categorized by the stage of the cell cycle in which they are the primarily active. The binding of the appropriate growth factor at the cell surface starts a signaling cascade that ultimately leads to the expression of the G1 phase cyclins. These cyclins combine with appropriate CDKs to form a complex that inactivates the Rb protein. As mentioned previously, in its active form, RB binds the transcription factor E2F. When Rb is inactivated by a cyclin/CDK complex in G1 phase of the cell cycle E2F is released to transcribe genes necessary for continued progression through the cell cycle. It is important to note that in normal cells, external factors (e.g., growth factors) are absolutely required for the cell to continue past the restriction point. After the restriction point, the cell is committed to DNA replication and cell division. Thus, the interference with normal signal transduction pathways by chemical carcinogens, regardless of mechanism, can force a cell into proliferation that is not governed by normal physiological controls. Even after passing through the restriction point early in G1 phase and committing to replication, there are still multiple mechanisms through which the cell regulates progression through the cell cycle. For example, the cell must pass through what is known as a “ checkpoint” at the G1/S boundary. The G1/S checkpoint serves to insure that DNA has been sufficiently repaired before new DNA is synthesized. The p53 protein plays a critical role at the G1/S checkpoint. There is evidence that p53 is directly involved in the detection of several types DNA damage. Upon detecting damage, p53 regulates the production of proteins that function to bring a halt to the cell cycle. There is also evidence to suggest that p53 actually mediates the repair of certain genetic lesions by DNA repair enzymes. Once the damaged DNA has been sufficiently repaired, the cell proceeds with the synthesis of new DNA. In this phase of the cell cycle, alterations in the fidelity of DNA synthesis or inefficient repair of replication errors could have detrimental effects on the cell. Following S phase, cells pass through another checkpoint to ensure that the DNA has been fully replicated before moving into the G2 phase. During this phase, the cells where prepares for mitosis by checking the DNA for replication errors and ensuring that the cellular machinery needed in mitosis is functioning properly. Following G2, the cells undergo mitosis and a daughter cell is created. Any errors in the made in the replication of the DNA of the original cell are now fixed in the DNA of the daughter cell. If a cell has sustained an unacceptable level of DNA damage, or in situations where the cell receives irregular growth signals, such as in the overexpression of the transcription factor and protooncogene myc, p53 can mediate a process called apoptosis. Simply put, apoptosis is cell suicide. Apoptosis is an extremely important component of many physiological processes relating to growth and development. In the developing embryo, for example, apoptosis is responsible for the elimination of superfluous cells that must be eliminated to ensure proper tissue structure and function (e.g., digit formation in developing limbs). Apoptosis is also responsible for the maintenance of the correct number of cells in differentiated tissues and the elimination of cells that have been irreparably damaged. Apoptosis is an orderly process characterized by several morphological stages, including chromatin condensation, cell shrinkage, and the packaging of cellular material into apoptotic bodies (also known as blebing) that can be consumed by phagocytes in the vicinity of the cell. This orderly and well-regulated process is a distinct contrast to cell death by necrosis. As indicated previously, the p53 protein has been implicated in apoptosis resulting from several different types of cell stress, including DNA damage induced by chemical mutagens (Figure 13.7). The mechanisms by which p53 mediates apoptosis are currently a subject of intensive study for cell biologists. Some functions of p53 in the apoptotic pathway are mediated by the transcription of certain genes (e.g., bax) that regulate apoptosis, while other effects appear to stem from protein–protein interactions with other intercellular mediators of apoptosis.


Figure 13.7 DNA damage leads to p53 accumulation and subsequent changes in gene expression and protein-protein interactions [Adapted from Harris (1996).]


Figure 13.8 A genetic model of the molecular events involved in human colorectal tumor development. [Adapted from Trends in Genetics, Volume 9. Vogelstein and Kinzler, The multistep nature of cancer. p. 140, 1993. with permission from Elsevier Science.]



Clearly apoptosis is a process critical to the balance of cell populations in normal tissue. The loss of the ability for neoplastic cells to undergo apoptosis could tip the scales in favor of cell proliferation and uncontrolled growth. As such, apoptosis is another process that could be detrimentally affected by the loss of p53 function. This has been a brief and necessarily simplistic overview of some of the cellular processes that can be subverted in the course of the carcinogenic process. It should be evident from this discussion that the formation of a malignant tumor is a multistage process that involves multiple molecular mechanisms, including the activation of oncogenes and the inactivation of tumor suppressor genes (Figure 13.8). What is not often clearly articulated to the student of chemical carcinogenesis is the fact that while different types of chemical carcinogens (e.g., genotoxic and epigenetic) differ mechanistically, these mechanisms have impacts on similar molecular pathways. While there is still much work to be done, the knowledge that has been developed over a relatively short time regarding the mechanisms of action of chemical carcinogens and critical cellular targets for these agents is astounding. This knowledge, has given us valuable insights to the origins of human cancer and will lead to the development of better tools with which to fight it.


The General Chronic Animal Bioassay Protocol The National Toxicology Program (NTP) is the agency currently responsible for testing chemicals for carcinogenic activity in the United States, a responsibility originally held by the National Cancer Institute. But while the responsibility for testing chemicals has changed, the general animal testing protocol currently used to evaluate the carcinogenic potential of a chemical has remained essentially the same for more than two decades. The basic procedure is relatively simple in experimental design, given the complexity of the disease process that constitutes the observational endpoint of this test procedure (cancer) and the consequences and importance accorded any positive findings identified by this test procedure. Furthermore, echoing an early recommendation by the FDA that testing be done at doses and under experimental conditions likely to yield maximum tumor incidence, the use of high doses to maximize the sensitivity of the procedure has become an area of considerable controversy. In short, the chemical doses tested, the animal species selected, and the simple, observational nature of this test often later become targets for criticism when the test results are applied in risk or hazard assessments that have a large impact on public health policy. For this reason, this and subsequent sections of this chapter will focus on the basic experimental design of the chronic animal cancer bioassay and the scientific issues commonly raised about these procedures or what interpretation might be given the results. The commonly recommended requirements for a thorough assessment of carcinogenic potential in a test animal that mold the basic experimental design of a chronic animal bioassay are

• That two species of rodents, both sexes of each species, should be tested as a minimum. This • •

helps ensure that false negative responses are not generated by selecting a non-responsive species for the test. That adequate controls are run during the test procedure. Ideally, the tumor incidence in test animals is compared to both historical and concurrent control animal responses. This helps ensure that the observed response is not an aberration of that specific study. A sufficient number of animals should be tested so that a positive response is not likely to be missed. The goal is to test enough animals to have a sufficient statistical basis whereby even a weak carcinogenic response should be observed and to be able to determine whether an observed increase in tumors, or lack thereof, was a chance or real observation. Typically, 50–100 animals of each sex and species are considered to be an appropriate-sized test.



• •

Increasing the number of animals tested might increase the sensitivity of the test, but as the number of animals is increased, the cost of the experiment rises and could render the test cost-prohibitive. The exposure and observation periods should last a lifetime, if possible, so that the latency of the response does not become an issue. At least two doses should be tested. One should be the maximally tolerated dose (MTD), the second dose should be some fraction (usually 50% or 25%) of the MTD. The MTD is defined as the highest dose that can be reasonably administered for the lifetime of the animal without producing serious, life-threatening toxicity to the animal that might compromise completion of the study. In the past the MTD has been defined as a dose that causes no more than a 10 percent decrease in body weight gain and does not lead to lethality over time. A detailed pathologic examination of all tissues should be held at termination of the experiment (and sometimes at 6-month intervals).

In addition to these recommended guidelines, this test is normally performed following good laboratory practice (GLP) procedures. These and other procedures ensure proper animal care during the extended period of the test, that no cross-contamination with other chemicals being tested will occur, and the possibility of having infectious agents or disease affect the outcome of the test is limited. Using these basic guidelines, any positive result obtained in at least one sex of one species is generally considered sufficient evidence to classify the chemical, for regulatory and public health purposes, as a carcinogen. Four different types of tissue response might be observed in a chronic test and considered positive evidence of carcinogenicity: 1. An increase in the incidence of a tumor type that occurs in control animals but at a significantly lower rate 2. The development of tumors at a significantly earlier period than is observed in the control animals 3. The presence of tumor types that are not seen in control animals 4. An increased multiplicity of tumors (although generally speaking, differences in total tumor load between exposed and unexposed animals is not considered reliable evidence) Positive results in a test with a more limited power to detect carcinogenicity (e.g., tests of shorter duration or fewer animals), but where the overall test procedures employed are considered adequate, may also become accepted as sufficient evidence of carcinogenicity, particularly where other relevant evidence (e.g., mechanistic data, structural alerts, structure–activity relationships) are also available. In contrast, because it is well recognized that important species differences exist in regard to response, negative results (an observed lack of a tumorigenic response), might not be considered definitive evidence that a chemical is not a carcinogen in other species that were not tested. The Issue of Generating False-Negative or False-Positive Results Both false-negative and false-positive results are a potential problem in carcinogen bioassays. Ideally, the number of animals required to provide adequate negative evidence should be great enough that even a false-negative test (a test failing to detect existent carcinogenicity) will not allow an excessive risk to go unnoticed. The likelihood that such a risk will not be detected during the evaluation of bioassay data is dependent on two factors (excluding species differences in response): the number of animals tested and the extent to which the test dose exceeds the usual level of human exposure, therefore increasing either parameter tends to lessen the chance of obtaining a false negative response (with respect to humans). The probability that a test will generate a false-negative result is also affected by the background tumor rate in the control animals. As the background incidence of tumorigenesis increases, so does the



number of animals required to detect a small percent increase in tumor incidence above the animal’s background rate. This means that it may be difficult to detect small increases for those tumor types that have large spontaneous background rates in an animal model when the test group contains only 50–100 animals. To increase the safety of the animal-to-human extrapolations, the number of animals tested in a cancer bioassay may need to be increased if (1) the number of humans to be exposed to the chemical is either expected to be large or (2) a small margin of safety exists between the animal dose tested and the expected human exposure. In general, however, resource limitations are such that only 50 animals of each sex are tested at each dose for both species; this limits the total number of animals tested to about 400 animals, plus 200 animals to serve as controls.

Short-Term Cancer Bioassays and Other Measures of Carcinogenic Potential Because of the large number of animals, lengthy timeframe, and expense associated with the chronic carcinogenesis bioassay, there has long been a need for reliable shorter-term tests of carcinogenic potential that could be used to complement standard carcinogenicity testing protocols. In the past, such short-term tests were limited to abbreviated initiation–promotion experiments with defined endpoints and the induction of tumors in susceptible animal models (e.g., lung tumors in strain A mice). Recently however, the tools of molecular biology have made it possible to construct genetically altered (transgenic) animals that may prove to be useful models for predicting the carcinogenic potential of chemicals. The U.S. National Toxicology Program is currently in the process of validating two of these transgenic models, Tg.AC mice and p53+/– mice, with chemicals previously tested in the standard 2-year chronic bioassay. These transgenic models are described briefly below. 1. The Tg.AC line was produced in FVB/N mice by the incorporation of a v-H-ras transgene into the cellular DNA. Mutations in ras oncogenes, which encode a family of GTP binding proteins critical to many growth factor signaling pathways, have been detected in a large proportion of human tumors. Tg.AC mice behave like genetically initiated mice, and rapidly develop epidermal papillomas in response to topical treatment with carcinogens. Researchers have shown that the mutant transgene is overexpressed in the proliferating cells in benign and malignant tumors but is not expressed in normal cells. Interestingly, treatment with initiators or tumor promoters induces the development of skin tumors. While treated mice have a dramatic increase in tumor yield with abbreviated time-to-tumor response, untreated mice have a normal skin histology and do not usually develop spontaneous tumors within the testing period. Treatment with carcinogens results in the production of papillomas in less than 6 months, substantially reducing the period of time for a typical initiation-promotion experiment in mouse skin. 2. Heterozygous p53+/– mice possess only a single functional copy of the p53 gene. As discussed previously, p53 function is lost through mutation or deletion in over 50 percent of all human cancers. With only a single functional allele, p53 mice develop normally but are at an increased susceptibility to the induction of tumors. This situation is analogous to an individual who has inherited a defective copy of a tumor suppressor gene. Upon dosing with mutagenic carcinogens, p53+/– mice rapidly develop tumors compared to normal mice, usually within 6 months. Untreated p53+/– mice do not usually develop tumors within the test period. In the chemicals tested thus far, there has been a high degree of concordance with results from traditional chronic bioassays. Examination of the discordant cases indicates that a number of these are due to the absence of hepatocellular carcinoma in the p53+/– mice. As will be discussed, the B6C3F1 mice typically used in chronic bioassays have a high spontaneous background rate of these tumors, making interpretation of positive carcinogenic responses in this tissue problematic. These and other transgenic animal models hold promise as less expensive and time-consuming adjuncts or replacements for conventional chronic bioassays. In addition, some scientists believe that transgenic animal models



may be more relevant to the humans because they possess alterations in genes known to be involved in many human tumors.

13.6 INTERPRETATION ISSUES RAISED BY CONDITIONS OF THE TEST PROCEDURE Human health hazards and, to allow for some quantitative assessment of the risk, the reliability of the animal-to-human extrapolation of animal cancer data is understandably an important issue. And as is true for any animal test procedure, questions concerning the reliability with which the results of the chronic bioassay can be extrapolated to human exposure conditions are frequently raised. However, in addition to the obvious potential for frank differences to arise in the human response because of species differences, an issue that can be raised with the animal test data for any other toxic endpoint (e.g., liver injury or developmental deficits), a number of interpretation issues have been raised that stem from the experimental conditions of the cancer bioassay test procedure itself. For example, it has long been noted that a number of interpretation problems will simply arise out of the data collected from this procedure because significant species differences may reasonably be anticipated and because of the test’s relatively crude and observational approach. That is, after approximately 3 years of test procedure, tissue collection and histopathological examination, we are largely left with a single, simple observation, specifically, the number of tumors in a tissue following lifetime high-dose exposure. Because the test procedure is arguably a screening test for carcinogenic activity, regardless of dose, when a positive result is observed, little else is provided. For example, typically little or no information is provided on dose–response or the mechanism by which the cancer was induced. Thus, it should perhaps not be surprising that the utility of this procedure continues to foster debate in the scientific community, or that much additional research is routinely required to be able to reliably interpret and extrapolate the results obtained by this procedure. In a seminal article dealing with the problems associated with chronic cancer bioassay tests and their interpretation, Squire listed five experimental design issues that remain relevant today: 1. 2. 3. 4. 5.

Use of the MTD as currently defined The number of doses tested The relevance of findings in certain test species Route of exposure and vehicle Extent of the pathological examination

These and related issues raised by this test procedure are discussed in the following paragraphs because they have a considerable impact on the hazard evaluation of the chemical in question, and because they are frequently raised when debates occur over the significance of the observation or the regulation of a particular chemical. The Doses Used to Test Chemicals Are Too High This has become perhaps the most frequently raised criticism of chronic animal cancer tests. Because the MTD is often selected as the highest dose the test animal can maintain for a lifetime without shortening the length of the test, it is often a dose where chronic toxicity and biological changes occur. The biological arguments against the applicability of positive results that are seen only at high doses, doses that produce chronic toxicity and substantially exceed the expected human exposure level, are as follows: 1. High doses may alter the metabolism and disposition of the chemical such that the types of reactive, toxic metabolites that are responsible for the critical biochemical changes producing cancer



are not present at lower doses. It has been shown with a number of chemicals that a particular metabolic pathway becomes saturated above a certain dose level. Once saturated either the formation of a specific toxic metabolite begins to increase, or a detoxification–protective pathway now begins to become overwhelmed. This leads to a cellular insult and damage that either does not occur at lower doses, or does so at a significantly lower rate. This phenomenon is often referred to as a dose that “ produces zero order kinetics” in an otherwise “ first-order reaction process.” 2. High doses produce irritation or inflammation. These conditions produce the formation of reactive oxygen species that are capable of inducing DNA damage that simply does not occur at lower doses. 3. High doses may produce changes in immune or endocrine systems, disrupt nutrition, or otherwise produce stressors that induce cancer secondary to changes in the background cancer rate. Because these same organ toxicities have thresholds and so do not occur at lower doses, low doses of the chemical are incapable of inducing cancer secondary to these specific biochemical and molecular changes. 4. High doses can produce damage to important DNA repair enzymes, or the DNA damage will overwhelm the cell’s ability to withstand these genetic assaults. 5. High doses produce a recurrent injury, cell death, and cell turnover that are not induced at lower doses. Under these conditions the cytotoxicity that is induced by high doses alters important cellular pools of factors responsible for maintaining genetic integrity within the cell. Thus, high doses may foster conditions within the cell that result in mutations or genetic damage indirectly. In addition, the increase in cell turnover may now cause mutations to become fixed that would normally be repaired. A sustained increase in cell turnover may also increase the rate at which natural errors in DNA replication and spontaneous mutations occur. In short, the criticism of high-dose testing is that the cancers observed may originate secondarily to other important biochemical changes and toxicities that are induced only at high doses. In this situation, where the chemical induces cancer indirectly and is related to conditions unique to high doses, then low-dose conditions would not be carcinogenic. Whether such chemicals should be viewed as a carcinogenic hazard or as chemicals without carcinogenic activity becomes a function of dose; an issue that may raise considerable controversy when the positive carcinogenicity data are used to regulate the exposures of such chemicals. The possibility that the carcinogenicity of a particular chemical may be a high-dose phenomenon has been assumed or hypothesized for a number of different chemicals and different mechanisms. For example, a number of different chemicals cause a chronic reduction in the circulating levels of thyroid hormone at high doses. This, in turn, causes a chronic elevation in blood levels of thyroid-stimulating hormone, a normal response to low thyroid levels, that results in a chronic overstimulation of thyroid follicular cells and eventually the development of thyroid follicular cell tumors. This high-dose phenomenon has a threshold (lower doses will not decrease thyroid hormone levels), and no risk of cancer would be associated with lower doses. Similarly, the bladder tumors observed with very high doses of compounds such as vitamin C, saccharin, glycine, melamine, and uracil are believed to be induced only when an excessive dose produces the depositing of insoluble calculi or crystals in the urinary bladder. The occurrence of these physical agents produce a chronic irritation or inflammation, thereby providing a stimulus for the proliferation of the bladder epithelium and ultimately the formation of bladder tumors. Since none of these changes are produced at lower doses, there are clear thresholds for these carcinogens, and their “ carcinogenic hazard” can be induced only at unrealistically high exposure levels. As the evidence accrued that use of the MTD in bioassays frequently produced dose-dependent results, scientists within the NTP assessed the use of the MTD and the long-held view that responses obtained at the MTD could be extrapolated in a linear fashion to lower doses. This assessment concluded that the following implicit assumptions underlie the current use of the MTD, and the associated use by regulatory agencies of a linear extrapolation of the results obtained with it:



• • • • •

The pharmacokinetics of the chemical are not dose-dependent. The dose–response relationship is linear. DNA repair is not dependent on dose. The response is not dependent on the age of the animal. The test dose need not bear a relationship to human exposure.

Following a review of these assumptions, however, the Board of Scientific Counselors within the NTP concluded that the implicit assumptions underlying linear extrapolation from the MTD do not appear to be valid for many chemicals, and that both the criteria for selecting doses tested in the chronic bioassay, as well as the method for extrapolating these results, should be reevaluated. Regarding the issue of alternative criteria for selecting the highest doses to be tested in a chronic bioassay, the following criteria recommended earlier by Squire seem to address a number of the issues that are raised by implicit assumptions associated with the MTD: 1. The MTD should induce no overt toxicity, that is, no appreciable death, organ pathology, organ dysfunction, or cellular toxicity. 2. The MTD induces no toxic manifestation predicted to shorten lifespan. 3. The MTD does not retard body weight by 10 percent. 4. The MTD is a dose that in two-generational studies is not detrimental to conception, fetal or neonatal development, and postnatal development or survival. 5. Takes into consideration important metabolic and pharmacokinetic data. There are a number of attractive features of this proposed definition for the MTD. Because the ultimate goal of all toxicity testing is to identify all potential hazards we should be guarding against, and to develop exposure limits that will prevent all toxicities, the criteria listed above allow other toxicological considerations, under specific circumstances, to set a reasonable upper limit on the doses tested for carcinogenicity. If high doses produce biochemical changes not seen at lower doses, and if at these doses the chemical produces other toxicities we have already identified and must prevent by limiting exposure, then these toxic endpoints may set a reasonable upper limit on the dose range we should employ to test for other toxicities (e.g., cancer). While a number of additional arguments and example compounds can be cited in support of changing the doses tested in chronic animal cancer bioassays, especially if we are going to use the results in the risk assessment and risk management areas, this particular feature of the test protocol has placed regulatory agencies on the horns of a dilemma that is difficult to escape. It seems only logical to attempt to maximize the sensitivity (ability to detect carcinogens) of this test by using the highest dose possible. Testing the maximal dose helps eliminate the chance of producing false negative responses. Testing the maximal dose helps ensure the statistical significance of small but important changes, and helps set a manageable limit on the number of animals that must be tested to be able to statistically identify a positive response. On the other hand, by maximizing the dose that is tested, we seem to be incurring a considerable number of positive responses, the results of which, after further testing for mechanisms at considerable additional expense, do not seem to be relevant, serious human hazards, at least at the doses to which humans are exposed. The possibility that this may be a substantial problem with the current testing scheme is indicated by analyses showing some 44 percent of the positive test results observed in NTP bioassays as positive (carcinogenic) only at the highest dose tested and not at a dose that is 25–50% of the MTD. Thus, it would appear that for almost one-half of the chemicals tested thus far, the carcinogenicity of the chemical is strictly a function of the high doses being tested.



Number of Doses Tested Once a chemical has been identified as capable of producing cancer in at least one animal species the results of that test are frequently used to develop exposure guidelines or regulatory standards via the development of a cancer slope factor or benchmark dose from these same data. Because in general only two or a very few doses are tested, and as these doses are usually relatively close in magnitude, the chronic bioassay frequently provides a poor database from which the human risk must be modeled. In rare instances both doses are positive at a maximal rate (i.e., 100 cancer incidence is seen at both doses). In this situation no judgement can be made as to the shape of the dose–response curve or how far one must go down in dose before the response begins to decline in a dose-dependent fashion. In other instances, one dose is positive and the second dose is not. In this situation there is again no information concerning the slope of the dose–response curve discernible from the data. Furthermore, in this situation modeling the single positive dose would also appear to inflate the cancer risks associated with low doses as the second dose, which is also a relatively high dose, produced no discernable activity. Both of these problems might be eliminated with the use of more doses, particularly where the doses are selected with the intent of developing usable dose–response data. A related problem is caused when the doses tested are both positive, and yield some information concerning the shape of the dose–response curve at doses where the increase in the cancer incidence is observable, but both doses are above that point where metabolic processes become saturated and now significant changes are seen in key biochemical pathways responsible for the tumorigenic response (e.g., metabolism, disposition, endocrine, immune or DNA damage–repair responses). In these instances it would be helpful to have tested doses at those points where the biochemical changes believed to be key to the carcinogenic process are either not saturated or do not occur so as to assess their mechanistic significance directly. The problem, however, with changing the protocol to include more doses is that it will dramatically increase the cost of performing a cancer bioassay (i.e., a 50 percent increase with each additional dose). So, once again regulatory agencies and public health officials are faced with the dilemma of either improving the test results at the expense of having the financial resources to test more compounds, or maximizing the number of tests performed within a specific budget at the possible expense of limiting the interpretation of the data.

The Route of Administration and Vehicle Issues Because the route of administration of a compound may alter the metabolism and disposition of a chemical, and because local damage at the site of application may induce certain changes necessary for the carcinogenic response, the route of administration tested in animals should mimic that of the intended or most likely route of human exposure. For example, some metals induce sarcomas at the site of injection when injected into the muscles of animals, apparently in response to local inflammatory and other responses, but are not carcinogenic by any other route of exposure. What importance should be attached to these responses? The issue of route specific differences in response has become so well recognized that agencies like the U.S. Environmental Protection Agency now calculate separate cancer slope factors for the inhalation and oral routes of exposure. In so doing they use route-specific animal test data in order to avoid making a route-to-route extrapolation from a single animal experiment. In some instances a vehicle is used to administer the test compound that is capable or either altering the pharmacokinetics of the compound (absorption, metabolism, etc.) or may cause changes (e.g., inflammation) that potentially influence the tumorigenic response of the chemical being tested. Where the vehicle produces either a qualitative or quantitative change in the response (compared to when no, or another, vehicle is used), the results should be interpreted with the appropriate caution. For example, corn oil has been used as a vehicle to administer chemicals not readily soluble in water. But distinct preneoplastic changes have been observed in organs like the pancreas in the corn-oil-only treatment group. So, tests producing pancreatic tumors when the chemical is administered in corn oil should be



evaluated for cocarcinogenic responses rather than attributing all of the activity to the chemical being tested. Issues Associated with the Histopathological Examination In some instances, perhaps more so in years past, the histopathological examination of the slides taken from the control animals have not been examined as rigorously as those slides taken from the animals administered the test compound. While at first it might seem that more attention should be paid to those slides where the potential change is anticipated, this can lead to results that are an artifact of the examination. For example, if all animals during the test became infected by a viral organism, and if this infection affected the background cancer incidence in a particular organ of the animal, then placing a greater emphasis on the “ exposed” slides might lead one to reach erroneous conclusions. In this situation the pathologist might identify more tumors in exposed animals simply because of the more extensive microscopic search of the exposed tissues even though equivalent numbers of infectioninduced tumors might exist in both control and exposed animals. Other aspects of the histopathological examination may affect the outcome of the study. For example, what organs should be examined? Should we evaluate organs like the Zymbal glands of rats if humans have no anatomic correlate? What relevance should be attached to results where only benign tumors, or tumors that behave benignly, are elicited? What relevance should be attached to a chemical that increases the tumor incidence in one organ while decreasing the tumor incidences in other organs, particularly if the total cancer/tumor risk of the animal group does not increase? Should we attach the same significance to these results? (Note: Here the extrapolation to humans would essentially be no net changes in the population’s risk of cancer.) As only one chemical example of this phenomenon, PCBs, a chemical of considerable regulatory restriction and interest, has been observed in several studies to produce liver tumors in rats, and relatively low exposure guidelines have been developed for this chemical on the basis of such data. However, two general findings in these studies were that the total tumor incidence in exposed animals was not increased (because the prevalence of other tumor types were decreased) and that these tumors did not behave like malignant masses; in fact, the exposed animals lived on average longer that did the control animals. One final facet of this issue is the fact that over the years the pathological descriptions (criteria for classifying pathological changes as tumors) have evolved. This means that chemicals using more modern descriptions might be viewed as having lower tumor incidences than they would if their evaluation occurred in years past. While this difference does not affect whether the test was considered to have produced a positive finding for carcinogenicity, it does affect the tumor incidence reported in the test, which, in turn, affects the perceived potency of the chemical as measured by the cancer slope factors derived from the tumor incidence that was reported. Thus, the perceived potency of a chemical carcinogen, as measured by its cancer slope factor, may differ according to which pathological criteria were used. Dietary and Caloric Restrictions Over the years we have come to realize that nutritional status and caloric intake during the test can affect the test results. In most test protocols the rats are fed ad libitum; that is, they are given constant access to food in the cage. Given the already restricted activities that can occur within these cages and the propensity of animals to eat as often as allowed, the animals tested under these conditions are generally obese animals during much of their lifetime. Studies with a number of different chemicals have shown that obesity can inflate the final tumor incidence that is observed; that is, there are a number of chemicals that, when administered at the same dose, will result in those animals placed on a normal caloric or restricted caloric intake to have significantly lower tumor rates than observed in animals fed ad libitum. Since ad libitum feeding is the general rule in chronic animal test procedures, the results of many studies have been inflated or possibly made statistically significant by the mere fact that the animals were allowed to ingest more food than their bodies need. Similarly, some chemicals might



induce nutritional changes in the animal secondary to organ toxicity, which, if ameliorated, may significantly alter the outcome of the bioassay. What Animal Species Represents the Most Relevant Animal Model? While it may be prudent for regulatory purposes to use animal data to predict what the human response might be when human data are unavailable, it should be remembered that when one makes an animal-to-human extrapolation, the basic assumption of that extrapolation is that the animal response is both qualitatively and quantitatively the same as the human response. However, because two different species may respond differently, either qualitatively and quantitatively, to the same dosage of a particular chemical, any animal-to-human extrapolation should be considered a catch-22 situation. That is, to know whether it is valid to extrapolate between a particular animal species and humans in a sense requires prior knowledge of both outcomes. So, even though toxicologists frequently use animal data to predict possible human outcomes, the potential for significant qualitative and quantitative differences to exist among species requires that the human response first be known before an appropriate animal model can be selected for testing and extrapolation purposes. But the selection of the appropriate animal model is complicated by the fact that innumerable and vast species differences exist. These differences are related primarily to the anatomical, physiological, and biochemical specificity of each species; these differences may produce significant wide variation in the metabolism, pharmacokinetics, or target organ concentrations of a chemical between species. When these differences are then combined with species-related differences in the physiology or biochemistry of the target organ, it is not surprising that significantly different responses may be achieved when one moves to a different test species. The major point of interest here, however, is that because these differences exist, the extrapolation of animal responses to humans should be viewed as being fraught with considerable difficulty and uncertainty. Important species differences encompass, but are not limited to, the following: 1. 2. 3. 4.

Basal metabolic rates Anatomy and organ structure Physiology and cellular biochemistry The distribution of chemicals in tissues (toxicodynamics); pharmacokinetics, absorption, elimination, excretion, and other factors 5. The metabolism, bioactivation, and detoxification of chemicals and their metabolic intermediates A few well-known examples that illustrate the magnitude of these differences are discussed below. Anatomic Differences Laboratory animals possess some anatomic structures that humans lack, and when cancer is observed in one of these structures, the particular relevance to humans is unknown and cannot be assumed with any scientific reliability. For example, the Zymbal gland, or auditory sebaceous gland, is a specialized sebaceous gland associated with the ears in Fischer rats. This gland secretes a product known as sebum. Although there is little information about the specific function of the secretion of the Zymbal gland, there is no known human structural correlate. Thus, the fact that dibromopropanol can cause squamous cell papillomas of the Zymbal gland in Fischer rats might be argued as providing no information relevant to discerning the carcinogenic potential of this chemical in humans. Another such problem exists with rodent species because they also possess an additional structure with no known human correlate: the forestomach. The esophagus empties into this organ, and it is here that ingested materials are stored before passing to the glandular stomach. The forestomach of rodents has a high pH, as opposed to the low pH of the human stomach, and high digestive enzyme activity.



In rats, hyperplastic and neoplastic changes in the forestomach may result from the chronic administration of compounds like butylated. Once again, however, the relevance to humans of such responses is not known.

Physiologic Differences Male rats produce a protein known as α-2-microglobulin, which, in combination with certain chemicals or their metabolites, causes a repeated cell injury response in the proximal tubules of the kidney. However, significant levels of α-2-microglobulin are not found in female rats, mice, or humans. Thus, the mechanism believed responsible for the repeated cell injury and tumors formation observed in male rats does not exist in these species. The male rat kidney tumors observed after chronic gasoline exposure, or exposure to certain aliphatic compounds, such as d-limonene, are notable examples of this phenomena. The scientific community has concluded that the positive male rat data for such chemicals is not relevant for predicting human cancer risk.

Cellular and Biochemical Differences The B6C3F1 mouse routinely used in cancer bioassays has a genetically programmed high background incidence of hepatocellular cancer. Approximately 20–30 percent of untreated animals develop this type of cancer. The B6C3F1 mouse is a genetic cross between the C3H mouse, which has almost a 60 percent background rate of liver cancer, and a C57BL mouse, which has a very low incidence rate of liver. Because the B6C3F1 mouse was bred to exhibit a genetic predisposition for developing liver cancer, tests using this animal model have subsequently identified a number of chemicals that are only liver carcinogens in this mouse strain and not the rat. In turn, the relevance of the liver tumors which are so commonly induced in this mouse are frequently questioned when extrapolated to humans, especially in light of the relatively low incidence with which human hepatocellular cancer occurs (3–5 cases per 100,000) in the United States. The molecular mechanism for the high background cancer incidence in the B6C3F1 mouse appears to be related to its propensity for oncogene activation in the liver. For example, the DNA of the B6C3F1 mouse H-ras oncogene is hypomethylated, or deficient in methylation. Methylation of DNA serves to block transcription of a gene. And since the mouse H-ras oncogene is not adequately methylated (i.e., not “ blocked” ), it may be inappropriately expressed more easily, thus providing a mechanistic foundation for the higher background incidence of liver tumors in this mouse strain. Further, certain types of hepatotoxicity may exacerbate the hypomethylation of the H-ras gene in this sensitive species, but have no significant effect on the gene methylation rates in less sensitive species. Thus, the relevance to humans of liver tumor development in this test species, or any other animal species which has a propensity for the spontaneous development of the tumor, is questionable. To summarize, the use of mice and rats is generally a compromise aimed at decreased costs. While primates or dogs might better represent the human response to some chemicals, they cannot be used routinely because of the additional costs incurred and other reasons. In general, the use of rodents as a surrogate animal model for humans might be criticized because rodents typically have a faster rate of metabolism than do humans. So, at high doses the metabolic pattern and percentage of compound ultimately metabolized may be significantly different than that of humans. If the active form of the carcinogen is a metabolite, then the animal surrogate may be more sensitive to the chemical because it generates more of the metabolite per unit of dose. Alternatively, the problem of false negatives also applies in that the selection of an insensitive species may yield a conclusion of noncarcinogenicity whereas further testing would uncover the actual tumorigenic activity. Because significant species differences exist in key aspects of all areas relevant to carcinogenesis (metabolism, DNA repair, etc.), and as these differences are the rule rather than the exception, extrapolating the response in any species to humans without good mechanistic data should be done with caution. In addition, developing mechanistic data that will allow comparisons to be made between humans and both a responsive and



nonresponsive species would appear to be the only way to improve our use (extrapolation) of chronic cancer bioassay data.

Are Some Test Species Too Sensitive? A number of strains or species have a significantly higher tumor incidence in a particular tissue than do humans. The incidence of liver tumors in B6C3F1 mice was discussed earlier. Another example is the strain A mouse, a mouse strain sometimes used to test a chemical’s potential to induce lung tumors. In this particular mouse strain the incidence of lung tumors in the control (unexposed) animals will reach 100 percent by the time the animals have reached old age. In fact, because all animals will at some point develop lung tumors, a shortening of the latency (time to tumor) or the number of tumors at an early age are used, rather than the final tumor incidence measured at the end of the animals’ lives. The use of positive data from an animal species with a particularly high background tumor incidence poses several problems. For example, are the mechanisms of cancer initiation or promotion the same for this chemical in humans? Can the potency of the chemical be estimated or even ranked when it might not be clear if the enhanced animal response is just a promotional effect of high background rate or the added effect of a complete carcinogen? Where the biology of the test animal clearly differs from that of humans is a positive response meaningful without corroboration in another species?


What is the Reliability of the Species Extrapolation? To test the reliability of making interspecies extrapolations, scientists have analyzed the results of a large number of chronic animal bioassays to ascertain the consistency with which a response in one species is also observed in another species. In one of the largest analyses performed to date, scientists analyzed the results for 266 chemicals tested in both sexes of rats and mice. The data forming this analysis is presented in Table 13.8. From the findings discussed above, after defining concordance to be species agreement for both positive and negative results, the authors of this analysis concluded the following:

• The intersex correlations are stronger than the interspecies correlations. • If only the male rat and female mouse had been tested, positive evidence of carcinogenicity would have led to the same conclusions regarding carcinogenicity/noncarcinogenicity in 96 percent of the chemicals tested in both sexes of both species (i.e., 255/266 correct responses). TABLE 13.8 Correlations in Tumor Response in NCI/NTP Carcinogenicity Studies Observed Outcome Comparisona






% Concordant (++ or ––) Responses

Male rats vs. female rats Male rats vs. male mice Male rats vs. female mice Female rats vs. male mice Female rats vs. female mice Male mice vs. female mice Rats vs. mice

74 46 29 46 57 78 67

25 43 33 32 23 10 32

12 36 36 37 39 23 36

181 145 145 156 156 177 131

292 270 273 271 275 288 266

87.3 70.7 74.7 74.5 77.5 88.5 74.4

Source: Adapted from Haseman and Huff (1987).



TABLE 13.9 Correlations across Species of Positive Cancer Bioassays Observed Outcome Comparison





Percent concordance (++ or ––)

12 23

111 111

67% 70%

36 36 37 39 148

125 128 115 119 487

37% 46% 40% 48% 43%

Intraspecies Comparisons Male rats vs. female rats Male mice vs. female mice Male rats vs. male mice Male rats vs. female mice Female rats vs. male mice Female rats vs. female mice Rats vs. mice

74 25 78 10 Interspecies Comparisons 46 59 46 57 208

43 33 32 23 131

This, in turn, suggests that the number of animals tested might be reduced (i.e., eliminate the testing of male mice and female rats).

• The high concordance between rats and mice supports the view that extrapolation of carcinogenicity outcomes to other species (humans) is appropriate. However, the high degree of concordance in this analysis stems from the fact that about half of the studies are negative and the chemical being tested manifested no carcinogenic activity. When a slightly different questions is asked—regarding how reliably positive test results can be extrapolated across species—a much different answer is reached. In Table 13.9 the noncarcinogens have been removed and the comparisons across sexes and species have been reanalyzed. Figure 13.9 contains the same

Figure 13.9



TABLE 13.10 The Poor Correlation in Organ Sites among Positive Rodent Tests Site of Cancer

N Rats/Mice


N Mice/Rats


Liver Lung Hematopoietic system Kidney (tubular cells) Mammary gland Forestomach Thyroid gland Zymbal gland Urinary bladder Skin Clitoral/Preputial gland Circulatory system Adrenal medulla

25/33 2/7 3/14 3/21 4/18 8/14 7/16 2/12 2/12 3/11 0/7 2/4 0/4

75 29 21 14 22 57 44 17 17 27 — 50 —

25/78 2/18 3/11 3/4 4/7 8/15 7/9 2/2 2/3 3/3 0/3 2/10 0/4

32 11 27 75 57 53 78 100 67 100 — 20 —






Source: Adapted from Haseman and Lockhart (1993).

analysis but compares the data from a subsequent update of the original study as well, illustrating that as the number of chemicals tested expands, the agreement in results across species does not seem to be changing. From this analysis it is evident that when a chemical induces cancer in one of these two rodent species, it is also carcinogenic in the other species less than 50 percent of the time. This lack of concordance between these two phylogenetically similar species raises a concern voiced by many scientists when such data are extrapolated to humans without also considering mechanistic and pharmacokinetic data from both species that might help explain why such large differences exist. A similar problem arises when the issue of identifying the correct target organ is considered. A recent analysis of the predictivity of the target organ for a carcinogen when extrapolating across two rodent species found one could predict the correct target organ about only about 37 percent of the time (Table 13.10). So, it would appear that not only is the assumption that a positive response in animals can be assumed to predict the human response, but the likelihood that the correct target has been identified would also seem to be of some question.

13.8 OCCUPATIONAL CARCINOGENS Although the first occupational carcinogen was identified by Sir Percival Pott in 1775, it was not until 1970 with the passage of the Occupational Safety and Health Act and establishment of the Occupational Safety and Health Administration (OSHA) that the United States had enforcement authority granted to an agency to regulate the use of substances that were considered carcinogenic in the workplace. Prior to 1970, the source that was widely considered the most authoritative was the American Conference of Governmental Industrial Hygienists (ACGIH) and industry relied on this organization to regulate worker exposure to chemicals and agents. The other event occurring about this time that has shaped our current view of occupational carcinogens was the emergence of the cancer bioassay. The development and continued use of this bioassay over the years has identified many hundreds of industrial chemicals as having carcinogenic activity, at least in high-dose animal tests, many of which had never before been suspected of human carcinogenic activity. As certain chemicals or groups of chemicals became identified as carcinogens, this, in turn, brought to bear new pressures on industries as lower exposure levels or alternative chemicals were sought to reduce the possible risks associated



with exposure to chemicals, many of which, before these new data were developed, were believed to be very safe and industrially useful chemicals. Since the mid-1970s, several organizations—both private and public—have attempted to identify occupational carcinogens, or possible carcinogens, in an effort to reduce workplace exposure since logically, occupational exposures to carcinogenic chemicals would potentially be their gravest threat to human health because of their duration (a working lifetime) and the magnitude of occupational exposures. For example, the ACGIH ranks the known carcinogenic hazard of the compounds for which it provides TLVs in their annual listing (Table 13.11). Similarly, OSHA has identified its own list of chemical carcinogens that it regulates (Table 13.12), and the National Institute for Occupational Safety and Health (NIOSH), which is often referred to as the “ research arm” of OSHA, provides a separate listing of what it considers to be the known or probable carcinogens that might be encountered in the workplace. Additional lists of known human carcinogens and chemicals known to be carcinogenic in animal tests include lists by the National Toxicology Program (Table 13.13) and the International Agency for Research on Cancers (IARC) which publishes a monograph series that evaluates the animal and human data for widely used chemicals and chemical processes (Table 13.14). In reviewing these different lists, it is of interest to note that rather than being identical, as one might expect, there can be significant differences in what is viewed as a possible carcinogen depending upon the agency promulgating the listing.

TABLE 13.11 Known or Suspected Carcinogens Identified by the ACGIHa Confirmed Human Carcinogen (A1) 4-Aminodiphenyl Arsenic Asbestos Benzene Benzidine Beryllium Bis(chloromethyl)ether Chromite ore processing Chromium(VI)

Coal tar pitch volatiles β-Naphthylamine Nickel, insoluble Nickel subsulfide Uranium (natural) Vinyl chloride Wood dust (hard or mixed hard/soft woods) Zinc chromates Suspected Human Carcinogen (A2)

Acrylonitirile Antimony trioxide Benz[a]anthracene Benzo[b]fluoranthene Benzo[a]pyrene Benzotrichloride 1,3-Butadiene Cadmium Calcium chromate Carbon tetrachloride Chloromethyl methyl ether Coal dust Diesel exhaust a

Diazomethane 1,4-Dichloro-2-butene Dimethyl carbamoyl chloride Ethylene oxide Formaldehyde Lead chromate 4,4′-Methylene bis(2-chloroaniline) 4-Nitrodiphenyl Oil mist, mineral Strontium chromite Sulfuric acid Vinyl bromide Vinyl fluoride

Including agents identified as carcinogens A1 or A2 in the Notice of Intended Changes for the TLVs.



TABLE 13.12 Potential Occupational Carcinogens Listed by NIOSH Acetaldehyde 2-Acetylaminofluorene Acrylamide Acrylonitrile Aldrin 4-Aminodiphenyl Amitrole Aniline o-Anisidine Arsenic Arsine Asbestos Benzene Benzidine Benzidine dyes Benzo[a]pyrene Beryllium 1,3-Butadiene tert-Butylchromate Cadmium (dust and fume) Calcium arsenate Captafol Captan Carbon black Carbon tetrachloride Chlordane Chlorinated camphene Chloroform Bis(chloromethyl) ether Chloromethyl methyl ether β-Chloroprene Chromic acid Chromates Chromyl chloride Coal tar pitch volatiles Coke oven emissions DDT 2,4-Diaminoanisole o-Dianisidine o-Dianisidine-based dyes 1,2-Dibromo-3-chloropropane Dichloroacetylene p-Dichlorobenzene 3,3′-Dichlorobenzidine Dichloroethyl ether 1,3-Dichloropropane Dieldrin Diesel exhaust Diglycidyl ether 4-Dimethylaminoazobenzene

Formaldehyde Gallium arsenide Gasoline Heptachlor Hexachlorobutadiene Hexachloroethane Hexamethyl phosphoramide Hydrazine Kepone Malonaldehyde Methoxychlor Methyl bromide Methyl chloride 4,4′-Methylenebis(2-chloroaniline) Methylene chloride 4,4′-Methylenedianiline Methyl hydrazine Methyl iodide α-Naphthylamine β-Naphthylamine Nickel carbonyl Nickel (insoluble, and soluble compounds) Nickel subsulfides (and roasting operations) 4-Nitrobiphenyl p-Nitrochlorobenzene 2-Nitronaphthalene 2-Nitropropane N-Nitrosodimethylamine Phenylglycidyl ether Phenylhydrazine N-Phenyl-β-naphthylamine Polychlorinated biphenyl Propane sultone β-Propiolactone Propylene dichloride Propylene imine Propylene oxide Rosin core solder pyrolysis products Silica, crystalline Silica, Christobolite Silica, quartz Silica, Tridymite Silica, Tripoli Talc, asbestiform 2,3,7,8-Tetrachlorodibenzo-p-dioxin 1,1,2,2-Tetrachloroethane Tetrachloroethylene Titanium dioxide Toluene-2,4-diisocyanate Toluenediamine (continued)



TABLE 13.12 Continued Dimethyl carbamoyl chloride 1,1-Dimethylhydrazine Dimethyl sulfate Dinitrotoluenes Di-sec-octyl phthalate Dioxane Environmental tobacco smoke Epichlorohydrin Ethyl acrylate Ethylene dibromide Ethylene dichloride Ethyleneimine Ethylene oxide Ethylene thiourea

o-Toluidine p-Toluidine 1,1,2-Trichloroethane Trichloroethylene 1,2,3-Trichloropropane Uranium Vinyl bromide Vinyl chloride Vinyl cyclohexene dioxide Vinylidene chloride Welding fumes Wood dust Zince chromates

Source: NIOSH Pocket Guide, 1999.

13.9 CANCER AND OUR ENVIRONMENT: FACTORS THAT MODULATE OUR RISKS TO OCCUPATIONAL HAZARDS Increased awareness of the ubiquity of synthetic, industrial chemicals in our environment has led a number of scientists to try to determine what role environmental exposures play in cancer causation. The USEPA devotes a great deal of its resources to this question as do other federal, international and private agencies such as the Agency for Toxic Substances and Disease Registry (ATSDR) of the Centers for Disease Control (CDC), the American Cancer Society (ACS), and the World Health Organization’s (WHO) International Agency for Research on Cancer (IARC) (see Table 13.14). While each organization researching the impact of our occupations, lifestyles, diets, and environmental exposures on cancer have differing agendas and views as to the predicted cancer risks associated with environmental exposures or our daily routines, there is widespread agreement that the most substantial risks, and the greatest causes of cancer, are those factors that are controlled by the individual (e.g., diet, smoking, alcohol intake). The importance of this fact is twofold: (1) it should be recognized that cancer is a phenomenon associated with normal biologic processes, and is therefore impacted by those factors that may affect our normal biologic processes (e.g., diet); and (2) many environmental risk factors exist, and these, in combination with hereditary risk factors, may frequently provide overwhelming influences in epidemiological studies of occupational hazards. Thus, the risk factors not being studied (and so frequently not controlled for) may mask or exacerbate the response being studied and so confound any study that is not normalized in a manner that removes all potential influences from the association being studied. Estimates of the contribution of various factors to the rate of cancer in humans were perhaps first put forth by Doll and Peto, who produced the results plotted in Figure 13.10. As can easily be seen in Figure 13.10, the vast majority of the cancers were thought to be related to lifestyle factors; tobacco and alcohol use, diet, and sexual behavior accounted for 75 percent of all cancers in this initial analysis. Conversely, industrial products, pollution, and occupation were thought to be related to only 7 percent of all cancers. Currently, the contributions of diet, disease, and viral agents are still being researched as perhaps the most common causes of cancer. In the years following Doll and Peto’s initial assertions, some scientists have questioned whether such a large proportion of the cancers in humans had such clearly defined causal associations. However, the most recent evidence accumulated by researchers in this area indicates that less than 1 percent of today’s cancers result from exposure to environmental pollution, and diet has since been identified as a key risk factor for cancer in nearly 200 epidemiologic studies. More importantly, the view that there



TABLE 13.13 Agents Listed in the Report on Carcinogens (8th Edition) from the National Toxicology Program, as Known or Suspected Human Carcinogens Known Human Carcinogens Aminobiphenyl (4-aminodiphenyl) Analgesic mixtures containing phenacetin Arsenic compounds, inorganic Asbestos Azathioprine Benzene Benzidine Bis(chloromethyl) ether 1,4-Butanediol dimethylsulfonate (Myleran) Chlorambucil 1-(2-Chloroethyl)-3-(4-methylcyclohexyl)-1nitrosourea Chloromethyl methyl ether Chromium hexavalent Coal tar Coke oven emissions Creosote (coal) Creosote (wood) Cyclophosphamide Cyclosporin A (cyclosporine A; ciclosporin) Diethylstilbestrol

Erionite Lead chromate Melphalan Methoxsalen [with ultraviolet A (UVA) therapy] Mineral oils Mustard gas 2-Naphthylamine (β-naphthylamine) Piperazine Estrone Sulfate Radon Sodium equilin sulfate Sodium estrone sulfate Soots Strontium chromate Tars Thiotepa [tris(1-aziridinyl)phosphine sulfide] Thorium dioxide Tris(1-aziridinyl)phosphine sulfide (thiotepa) Vinyl chloride Zinc chromate

Agents Reasonably Anticipated to be Human Carcinogens Acetaldehyde 2-Acetylaminofluorene Acrylamide Acrylonitrile Adriamycin (doxorubicin hydrochloride) 2-Aminoanthraquinone o-Aminoazotoluene 1-Amino-2-methylanthraquinone Amitrole o-Anisidine hydrochloride Azacitidine (5-azacytidine) Benz[a]anthracene Benzo[b]fluoranthene Benzo[j]fluoranthene Benzo[k]fluoranthene Benzo[a]pyrene Benzotrichloride Beryllium aluminum alloy Beryllium chloride Beryllium fluoride Beryllium hydroxide Beryllium oxide Beryllium phosphate Beryllium sulfate tetrahydrate

Beryllium zinc silicate Beryl ore Bis(chloroethyl) nitrosourea (BCNU) Bis(dimethylamino)benzophenone Bromodichloromethane 1,-Butadiene Butylated hydroxyanisole (BHA) Cadmium Cadmium chloride Cadmium oxide Cadmium sulfate Cadmium sulfide Carbon tetrachloride Ceramic fibers Chlorendic acid Chlorinated paraffins (C12, 60% chlorine) 1-(2-Chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU) Chloroform 3-Chloro-2-methylpropene 4-Chloro-o-phenylenediamine p-Chloro-o-toluidine p-Chloro-o-toluidine hydrochloride Chlorozotocin (continued)



TABLE 13.13 Continued CIa Basic Red 9 monohydrochloride Cisplatin p-Cresidine Cristobalite [under “ Silica, crystalline (respirable size)” ] Cupferron Dacarbazine 2,4-Diaminoanisole sulfate 2,4-Diaminotoluene Dibenz[a,h]acridine Dibenz[a,j]acridine Dibenz[a,h]anthracene 7H-Dibenzo[c,g]carbazole Dibenzo[a,e]pyrene Dibenzo[a,h]pyrene Dibenzo[a,i]pyrene Dibenzo[a,l]pyrene 1,2-Dibromo-3-chloropropane 1,2-Dibromoethane [ethylene dibromide (EDB)] 1,4-Dichlorobenzene (p-dichlorobenzene) 3,3-Dichlorobenzidine 3,3-Dichlorobenzidine dihydrochloride Dichlorodiphenyltrichloroethane (DDT) 1,2-Dichloroethane (ethylene dichloride) 1,3-Dichloropropene (technical-grade) Diepoxybutane N,N-Diethyldithiocarbamic acid 2-chloroallyl esterDEHP; bis(2-ethylhexyl phthalate)] Diethylnitrosamine Diethyl sulfate Diglycidyl resorcinol ether 1,8-Dihydroxyanthraquinone [Danthron] 3,3-Dimethoxybenzidine 4-Dimethylaminoazobenzene 3,3-Dimethylbenzidine Dimethylcarbamoyl chloride 1,1-Dimethylhydrazine (UDMH) Dimethylnitrosamine Dimethyl sulfate Dimethylvinyl chloride 1,6-Dinitropyrene 1,8-Dinitropyrene 1,4-Dioxane Direct Black 38 Direct Blue 6 Disperse Blue 1 Epichlorohydrin Estradiol-17b Estrone Ethinylestradiol

Ethyl acrylate Ethylene oxide Ethylene thiourea Ethyl methanesulfonate Formaldehyde (gas) Furan Glasswool Glycidol hexachlorobenzene α-Hexachlorocyclohexane β-Hexachlorocyclohexane γ-Hexachlorocyclohexane Hexachlorocyclohexane Hexachloroethane Hexamethylphosphoramide Hydrazine Hydrazine sulfate Hydrazobenzene Indeno[1,2,3-cd]pyrene Iron dextran complex Kepone (chlordecone) Lead acetate Lead phosphate Lindane Mestranol 2-Methylaziridine (propylenimine) 5-Methylchrysene 4,4-Methylenebis(2-chloraniline) 4,4-Methylenebis(N,N-dimethylbenzenamine) Methylene chloride 4,4-Methylenedianiline 4,4-Methylenedianiline dihydrochloride Methylmethanesulfonate N-Methyl-N-nitro-N-nitrosoguanidine Metronidazole Mirex Nickel Nickel acetate Nickel carbonate Nickel carbonyl Nickel hydroxide Nickel hydroxide Nickelocene Nickel oxide Nickel subsulfide Nitrilotriacetic acid o-Nitroanisole 6-Nitrochrysene Nitrofen Nitrogen mustard hydrochloride 2-Nitropropane (continued)



TABLE 13.13 Continued 1-Nitropyrene 4-Nitropyrene N-Nitroso-n-butyl-N-(3-carboxypropyl)amine N-Nitroso-n-butyl-N-(4-hydroxybutyl)amine N-Nitrosodi-n-butylamine N-Nitrosodiethanolamine N-Nitrosodi-n-propylamine N-Nitroso-N-ethylurea (N-ethyl-N-nitrosourea (ENU) 4-(N-Nitrosomethylamino)-1-(3-pyridyl)1-butanone N-Nitroso-N-methylurea N-Nitrosomethylvinylamine N-Nitrosomorpholine N-Nitrosonornicotine N-Nitrosopiperidine N-Nitrosopyrrolidine N-Nitrososarcosine Norethisterone Ochratoxin A 4,4-Oxydianiline Oxymetholone Phenacetin Phenazopyridine hydrochloride Phenoxybenzamine hydrochloride Phenytoin Polybrominated biphenyls (PBBs) Polychlorinated biphenyls (PCBs) Polycyclic aromatic hydrocarbons (PAHs) Procarbazine hydrochloride Progesterone

1,3-Propane sultone β-propiolactone Propylene oxide Propylthiouracil Quartz [under “ silica, crystalline (respirable size)” ] Reserpine Saccharin Safrole Selenium sulfide Silica, crystalline (respirable size) Streptozotocin 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) Tetrachloroethylene (perchloroethylene) Tetranitromethane Thioacetamide Thiourea Toluene diisocyanate o-Toluidine o-Toluidine hydrochloride Toxaphene 2,4,6-Trichlorophenol 1,2,3-Trichloropropane Tridymite Tris(2,3-dibromopropyl) phosphate Urethane (Urethan; ethyl carbamate) 4-Vinyl-1-cyclohexene diepoxide


Color Index.

was a “ cancer epidemic” in this nation attributable to environmental exposure to pollutants shown to cause cancer in animals has been found to be inaccurate. In the absence of large percentages of cancers attributable to environmental contaminants or occupational exposures, then, we are faced with determining how much of our cancer risk is inevitable (due to aging processes or perhaps genetic predisposition) or could be offset by changes to lifestyle factors such as smoking and diet.

Genetic Makeup of Individuals The understanding of the role that genetics plays in carcinogenesis increased greatly in the 1990s and the relationship between genetic makeup and carcinogenesis is rapidly becoming a dominant area of cancer research. To date there have been more than 600 genetic traits associated with an increased risk of neoplasia. This relatively recent area of research is focused on how changes in the phenotypic expression of certain enzymes may alter the activation, detoxification, or repair mechanisms and thereby enhance the genetic damage produced by a particular chemical exposure. Genetic predisposition now accounts for perhaps 5–10 percent of all cancers, and it has been identified as a component



TABLE 13.14 IARC Carcinogens Group 1: Carcinogenic to Humans (75) Exposure circumstances Helicobacter pylori (infection with) Aluminum production Hepatitis B virus (chronic infection with) Auramine, manufacture of Hepatitis C virus (chronic infection with) Boot and shoe manufacture and repair Human immunodeficiency virus type 1 (infection with) Coal gasification Human papillomavirus type 16 Coke production Human papillomavirus type 18 Furniture/cabinetmaking Human T-cell lymphotropic virus type I Hematite mining with exposure to radon Melphalan Iron and steel founding 8-Methoxypsoralen (methoxsalen) Isopropanol manufacture (strong-acid process) MOPP and other combined chemotherapy, including Magenta, manufacture of alkylating agents Painter Mustard gas (sulfur mustard) Rubber industry 2-Naphthylamine Strong-inorganic-acid mists containing sulfuric acid Nickel compounds Agents and groups of agents Opisthorchis viverrini (infection with) Aflatoxins, naturally occurring Oral contraceptives, combined 4-Aminobiphenyl Oral contraceptives, sequential Arsenic and arsenic compounds Radon and its decay products Asbestos Schistosoma haematobium (infection with) Azathioprine Silica, crystalline Benzene Solar radiation Benzidine Talc containing asbestiform fibers Beryllium and beryllium compounds Tamoxifen N,N-Bis(2-chloroethyl)-2-naphthylamine 2,3,7,8-Tetrachlorodibenzo-para-dioxin (Chlomaphazine) Thiotepa Bis(chloromethyl) ether and chloromethyl Treosulfan methyl ether Vinyl chloride 1,4-Butanediol dimethanesulfonate (Busulphan; Myleran) Mixtures Cadmium and cadmium compounds Alcoholic beverages Chlorambucil Analgesic mixtures containing phenacetin 1-(2-Chloroethyl)-3-(4-methylcyclohexyl)-1nitrosourea (methyl-CCNU; Semustine) Betel quid with tobacco Chromium VI compounds Coal tar pitches Ciclosporin Coal tars Cyclophosphamide Mineral oils, untreated and mildly treated Diethylstilboestrol (DES) Salted fish (Chinese style) Epstein–Barr virus Shale oils Erionite Soots Ethylene oxide Tobacco products, smokeless Estrogen therapy, postmenopausal Tobacco smoke Estrogens, nonsteroidal Wood dust Estrogens, steriodal (continued)



TABLE 13.14 Continued Group 2A: Probably Carcinogenic to Humans (59) Agents and groups of agents Acrylamide Adriamycin Androgenic (anabolic) steroids Azacitidine Benz[a]anthracene Benzidine-based dyes Benzo[a]pyrene

4,4′-Methylene bis(2-chloroaniline) Methyl methanesulfonate N-Methyl-N′-nitro-N-nitrosoguanidine (MNNG) N-Methyl-N-nitrosourea (nitrogen mustard) N-Nitrosodiethylamine N-Nitrosodimethylamine Phenacetin Procarbazine hydrochloride Bischloroethyl nitrosourea (BCNU) Styrene-7,8-oxide 1,3-Butadiene Tetrachloroethylene Captafol Trichloroethylene Chloramphenicol 1,2,3-Trichloropropane Chlorinated toluenes (benzyl chloride), benzotrichloride, benzyl chloride and benzoyl chloride Tris(2,3-dibromopropyl) phosphate Ultraviolet radiation A 1-(2-Chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU) Ultraviolet radiation B para-Chloro-ortho-toluidine and its strong-acid salts Ultraviolet radiation C Chlorozotocin Vinyl bromide Cisplatin Vinyl fluoride Clonorchis sinensis (infection with) Mixtures Dibenz[a,h]anthracene Creosotes Diethyl sulfate Diesel engine exhaust Dimethylcarbarnoyl chloride Hot mate 1,2-Dimethylhydrazine Polychlorinated biphenyls Dimethyl sulfate Exposure circumstances Epichlorohydrin Art glass, glass containers and pressed ware (manufacture of) Ethylene dibromide Hairdresser or barber (occupational exposure as a) N-Ethyl-N-nitrosourea Nonarsenical insecticides (occupational exposures Formaldehyde in spraying and application of) Human papillomavirus type 31 Petroleum refining (occupational exposure in) Human papillomavirus type 33 Sunlamps and sunbeds (use of) IQ (2-Amino-3-methylimidazo[4,5-f]quinoline) Kaposi’s sarcoma herpesvirus/human herpesvirus 8 5-Methoxypsoralen Group 2B: Possibly Carcinogenic to Humans (227) Agents and groups of agents A-a-C(2-amino-9H-pyrido[2,3-b]indol) Acetaldehyde Acetamide Acrylonitrile A-F-2[2-(2-Furyl)-3-(5-nitro-2-furyl)acrylamide] Aflatoxin M1 para-Aminoazobenzene ortho-Aminoazotoluene

2-Amino-5-(5-nitro-2-furyl)-1,3,4-thiadiazole Amitrole ortho-Anisidine Antimony trioxide Aramite Auramine Azaserine Aziridine Benzo[b]fluoranthene (continued)



TABLE 13.14 Continued Benzo[j]fluoranthene Benzo[k]fluoranthene Benzofuran Benzyl violet 4B Bleomycins Bracken fern Bromodichloromethane Butylated hydroxyanisole (BHA) β-Betyrolactone Caffeic acid Carbon black Carbon tetrachloride Catechol Ceramic fibres Chlordane Chlordecone (Kepone) Chlorendic acid para-Chloroaniline Chloroform 1-Chloro-2-methylpropene Chlorophenoxy herbicides 4-Chloro-ortho-phenylenediamine Chloroprene Chlorothalonil CI Acid Red 114 CI Basic Red 9 CI Direct Blue 15 Citrus Red 2 Cobalt para-Cresidine Cycasin Dacarbazine Dantron (1,8-Dihydroxyanthraquinone) Daunomycin DDT (p,p′-DDT) N,N′-Diacetylbenzidine 2,4-Diaminoanisole 4,4′-Diaminodiphenyl ether 2,4-Diaminotoluene Dibenz[a,h]acridine Dibenz[a,j]acridine 7H-Dibenzo[c,g]carbazole Dibenzo[a,e]pyrene Dibenzo[a,h]pyrene Dibenzo[a,i]pyrene Dibenzo[a,l]pyrene 1,2-Dibromo-3-chloropropane para-Dichlorobenzene

3,3′-Dichlorobenzidine 3,3′-Dichloro-4,4′-diaminodiphenyl ether 1,2-Dichloroethane Dichloromethane (methylene chloride) 1,3-Dichloropropene Dichlorvos Di(2-ethylhexyl)phthalate 1,2-Diethylhydrazine Diglycidyl resorcinol ether Dihydrosafrole Diisopropyl sulfate 3,3′-Dimethoxybenzidine (ortho-dianisidine) para-Dimethylaminoazobenzene trans-2-[(Dimethylarnino)methylimino]-5-[2-(5nitro-2-furyl)-vinyl]-1,3,4-oxadiazole 2,6-Dimethylaniline (2,6-Xylidine) 3,3′-Dimethylbenzidine (ortho-tolidine) 1,1-Dimethylhydrazine 3,7-Dinitrofluoranthene 3,9-Dinitrofluoranthene 1,6-Dinitropyrene 1,8-Dinitropyrene 2,4-Dinitrotoluene 2,6-Dinitrotoluene 1,4-Dioxane Disperse Blue 1 1,2-Epoxybutane Estrogen–progestogen therapy, postmenopausal Ethyl acrylate Ethylene thiourea Ethyl methanesulfonate 2-(2-Formylhydrazino)-4-(5-nitro-2-furyl)thiazole Furan Glasswool Glu-P-1 (2-Amino-6-methyldipyrido[1,2-a:3′,2′d]imidazole) Glu-P-2(2-Aminodipyrido[1,2-a:3′,2′-d]imidazole) Glycidaldehyde Griseofulvin HC Blue No. 1 Heptachlor Hexachlorobenzene Hexachloroethane Hexachlorocyclohexanes Hexamethylphosphoramide Human immunodeficiency virus type 2 (infection with)




TABLE 13.14 Continued Human papillomaviruses: some types other than 16, 18, 31 and 33 Hydrazine Indeno[ 1,2,3-cd]pyrene Iron–dextran complex Isoprene Lasiocarpine Lead Magenta MeA-a-C(2-amino-3-methyl-9H-pyrido[2,3-b]indol) Medroxyprogesterone acetate MeIQ MeIQx merphalan 2-Methylaziridine (Propyleneimine) Methylazoxymethanol acetate 5-Methylchrysene 4,4′-Methylene bis(2-methylaniline) 4,4′-Methylenedianiline Methyl mercury compounds 2-Methyl-1-nitroanthraquinone N-Methyl-N-nitrosourethane Methylthiouracil Metronidazole Mirex Mitomycin C Monocrotaline 5-(Morpholinomethyl)-3-[(5nitrofurfurylidene)amino]-2-oxazolidinone Nafenopin Nickel, metallic Niridazole Nitrilotriacetic acid 5-Nitroacenaphthene 2-Nitroanisole Nitrobenzene 6-Nitrochrysene Nitrofen 2-Nitrofluorene 1-[(5-Nitrofurfurylidene)amino]-2-imidazolidinone N-[4-(5-Nitro-2-furyl)-2-thiazolyl]acetamide Nitrogen mustard N-oxide 2-Nitropropane 1-Nitropyrene 4-Nitropyrene N-Nitrosodi-n-butylamine N-Nitrosodiethanolamine N-Nitrosodi-n-propylamine 3-(N-Nitrosomethylamino)propionitrile

4-(N-Nitrosomethylamino)-1-(3-pyridyl)-1butanone (NNK) N-Nitrosomethylethylamine N-Nitrosomethylvinylamine N-Nitrosomorpholine N-Nitrosonornicotine N-Nitrosopiperidine N-Nitrosopyrrolidine N-Nitrososarcosine Ochratoxin A Oil Orange SS Oxazepam Palygorskite (attapulgite) Panfuran S Phenazopyridine hydrochloride Phenobarbital Phenoxybenzamine hydrochloride Phenyl glycidyl ether Phenytoin PhIP (2-amino-1-methyl-6-phenylimidazo[4,5b]pyridine) Polychlorophenols and their sodium salts (mixed exposure) Ponceau MX Ponceau 3R Potassium bromate Progestins Progestogen-only contraceptives 1,3-Propane sultone β-Propiolactone Propylene oxide Propylthiouracil Rockwool Safrole Schistosoma japonicum (infection with) Slagwool Sodium ortho-phenylphenate Sterigmatocystin Streptozotocin Styrene Sulfallate Tetrafluoroethylene Tetranitromethane Thioacetamide 4,4′-Thiodianiline Thiourea Toluene diisocyanates ortho-Toluidine Toxins derived from Fusarium moniliforme




TABLE 13.14 Continued Trichlormethine (trimustine hydrochloride) Diesel fuel, marine Trp-P-1 (3-Amino-1,4-dimethyl-5H-pyridol[4,3Engine exhaust, gasoline b]indole) Fuel oils Trp-P-2 (3-Amino-1-methyl-5H-pyrido[4,3-b]indole) Gasoline Trypan blue Pickled vegetables Uracil mustard Polybrominated biphenyls Urethane Toxaphene Vinyl acetate Welding fumes 4-Vinylcyclohexene Exposure circumstances 4-Vinylcyclohexene diepoxide Carpentry and joinery Mixtures Dry cleaning Bitumens Printing processes Carrageenan Textile manufacturing industry Chlorinated paraffins (C12 and 60% Cl) Coffee

Figure 13.10 Cancer factors: approximate percent contribution (Doll and Peto, 1981)



in lung, colorectal, and breast cancers (among the major cancer types) as well as being a key factor in many rarer forms of cancer such as nevoid basal cell carcinoma. This area of research may well change the way in which we view certain chemical exposures, as the risk of cancer may ultimately be shown to be more a function of an individual’s or groups of individuals’ unique susceptibility to a given chemical. Such information would not only improve our understanding of the carcinogenic process, but it may alter chemical exposure regulation by allowing screening tests to eliminate potentially susceptible persons from future potentially adverse exposures. For example, El-Zein et al. report that the inheritance of variant polymorphic genes such as CYP2D6 and CYP2E1 for the activation of certain chemicals, and GSTM1 and GSTT1 for the detoxification of certain chemicals, may predispose smokers with these traits to lung cancer. The importance of identifying the range of phenotypic expression among specific genes is clearly manifest in the impact that such changes may frequently make in the ultimate outcome of chemical exposure. In the future, identifying gene variants have a large impact on epidemiological research, cancer prevention, and the development of more effective intervention and treatment modalities. In addition, the ability to identify those genetic traits that influence certain types of cancer might become useful biomarkers that enable employers to place persons in positions that do not expose them to agents that would otherwise place them at a greater risk than the normal population. Because of the cell transformation that occurs in carcinogenesis, there is some “ genetic” component to every cancer. However, the traits referred to as one’s “ genetic makeup” are only a portion of the many factors that might occur in the progression from a healthy cell to an immortal, cancerous one. The role of environmental factors, as they might impact or augment hereditary or genetic elements of carcinogenesis are illustrated in Figure 13.11. The “ all environmental risks” box in this diagram is intended to represent the sum of all possible environmental insults; these might come from occupational exposures, lower-level environmental chemical exposures (indoor air, drinking water, diet), diets and dietary insufficiency, viruses and other infectious diseases, and important lifestyle factors (e.g., inactivity, smoking, drinking, illicit drug use).

Smoking The American Cancer Society (ACS) has compiled statistical data for the incidence of cancers in the U.S. population (Figure 13.12). For six major cancer sites in males in the United States, only lung cancers, which are far and away associated with tobacco smoking (perhaps 87 percent of all lung cancer deaths), have shown any demonstrable increase in the last 65+ years. The data for female cancers were similar. Lung cancer in females, driven by smoking, has now outstripped breast cancer as the leading cause of cancer death among U.S. women. The ACS stated:

Figure 13.11 Interactions of environmental (lifestyle, diet viral, occupational) exposures and genes.



Figure 13.12 Cancer incidence rates for U.S. males, annual trends. (From Cancer Facts and Figures—1999, American Cancer Society.)

Lung cancer mortality rates are about 23 times higher for current male smokers and 13 times higher for current female smokers compared to lifelong never-smokers. In addition to being responsible for 87 percent of lung cancers, smoking is also associated with cancers of the mouth, pharynx, larynx, esophagus, pancreas, uterine cervix, kidney, and bladder. Smoking accounts for at least 30 percent of all cancer deaths, is a major cause of heart disease, and is associated with conditions ranging from colds and gastric ulcers to chronic bronchitis, emphysema, and cerebrovascular disease. The data surrounding smoking is particularly distressing for persons who might be occupationally exposed to other substances as well. Asbestos-exposed workers who smoke reportedly contract lung cancer at a rate that is 60 times that of persons not exposed to either substance. Other risk factors for lung cancer may include exposure to arsenic, some organic chemicals, radon, radiation exposure from occupational, medical, and environmental sources. Smokers who incur such exposures should be aware of the increased risks they face compared to their nonsmoking co-workers. Research has identified more than 40 carcinogenic substances emitted in tobacco smoke. Many of these substances are initiating agents (genotoxic) and are capable of inducing cancer by themselves at sufficient doses, others are recognized as promoters or cocarcinogens and act to enhance the activity of chemicals initiating the key genetic change. With so many different chemical carcinogens contained in cigarette smoke, it seems logical to ask if cigarette smoking is largely a phenomenon of initiation or promotion. If lung cancer due to cigarette smoking was the result of initiating carcinogens, the observed risk should arguably be proportional to cumulative lifetime exposure, and the cessation of cigarette smoking would not alter the already accumulated pack/year risk (i.e., one’s risk of cancer, once achieved, could not be decreased with abstinence). Current data, however, is contradictory to this



suggestion, and studies indicate that as the duration of abstinence from smoking increases, a person’s lung cancer risk actually becomes lower until it eventually approaches the risk faced by a nonsmoker. For this reason, many have argued that the affect of cigarette smoking is largely one of promotion. Regardless of whether smoking is largely due to promotion or initiation, it is clearly an avoidable health hazard and after factoring in the increased risk from cerebrovascular disease due to smoking is arguably society’s greatest contributor to preventable causes of death. Alcohol Alcohol is another clearly avoidable cancer risk. Alcohol consumption is causally related to cancers of the oral cavity, pharynx, larynx, esophagus, and liver. The combined use of alcohol and tobacco products also leads to an increased incidence of oral cavity, esophagus, and larynx cancers. Associations between alcohol and breast cancer have also been proposed. Estimates of the contribution of alcohol to cancer in the United States range as high as 5 percent; however, it is estimated that there are some 10 million problem drinkers in the United States, and so, the influence ultimately exerted upon the national cancer incidence by alcohol might not be fully determined at the present time. There are several theories regarding the carcinogenic activity of alcohol. Alcohol is known to induce specific oxidative enzymes and so is suspected of potentially enhancing the initiation activity of certain carcinogens. It has also been proposed to make tissues more responsive to the action of a carcinogen by increasing cell permeability or by increasing the effective concentration of a carcinogen intracellularly. Ethanol is cytotoxic chemical at high doses, and recurrent cellular injury has been suggested as another possible mechanism for ethanol-induced or enhanced carcinogenesis. The fact that the development of cirrhosis often precedes and frequently ends in primary liver cancer would tend to support this hypothesis. Other possible mechanisms include the generation of free radicals (via lipid peroxidation), and possibly some immunosuppressive effect. Regardless of the mechanism or mechanisms by which chronic alcohol intake induces cancer or enhances the response of other carcinogens, it clearly remains as a clearly important, but avoidable, cancer risk factor. Diet When Doll and Peto released their statistical analysis of the causes of cancer, many authors noted the impact that diet had on cancer incidence was as yet unknown, or at best, very much debated. Diet, via the intake of high quantities of animal fats, can have a decidedly negative impact on a person’s health and such diets are clearly linked to higher incidences of cancers. However, diet is a double-edged sword in that it can also be an important moderating influence by providing antioxidants, anticarcinogens, and other nutritional benefit that helps the body’s detoxification and repair mechanisms to fight off tumorigenic activity. So, with the possible exception of the cessation of smoking, the improvement of our diet can have the greatest impact on our own health and the national cancer rate. It is now well recognized that the plants we consume as part of our diet contain their own natural pesticides. In fact, certain strains of plants have been cultivated with the purpose of enhancing these natural defense mechanisms and so require less maintenance and care. However, as was seen with the increased use of synthetic chemicals, this can enhance the toxicity of the foods we consume. As with the synthetic chemicals tested in the chronic animal cancer bioassay, the carcinogenic activity of the “ natural” pesticides normally contained in vegetables and fruits is running at roughly 50 percent for the chemicals tested. Thus, it has been argued that when chemicals are tested in high-dose animal cancer bioassays one can expect approximately half of the chemicals tested, human-made (synthetic) or natural, to elicit carcinogenic activity. Based on these projections and on the currently available data, it has been estimated that 99.9 percent of our total pesticide intake is via the ingestion of natural, plant-produced pesticides. In fact, it would appear we ingest as much as 1.5 g (1500 mg) of plant-produced, natural pesticides each day. Recently, the National Research Council’s (NRC) Board of Environmental Studies and Toxicology Committee on Comparative Toxicity of Naturally Occurring Carcinogens published a conclusion



similar to that of Ames. Although the committee admitted that more research was needed before definitive conclusions could be drawn, it stated that natural components of the diet were likely to be more significant with respect to cancer risk than were synthetic chemicals found in food. The committee’s conclusion was based on the amounts of foods consumed by the typical U.S. citizen and the levels of natural or synthetic pesticides present in those foods. The committee refers to various studies, including the National Health and Nutrition Examination Surveys (NHANES, the recent study of pesticides in the diets of infants and children, and the Nationwide Food Consumption survey performed by the US Department of Agriculture (USDA) as sources of data for their analysis. The NRC committee interpreted from these different studies that Americans consume a large number of natural and synthetic carcinogens in their diets. The committee also based its conclusion regarding the potential significance of dietary carcinogens on the fact that the natural dietary substances studied to date have, on average, a greater carcinogenic potency than the synthetic chemicals found in food. A diet high in animal fats has been implicated in numerous epidemiologic and case-control studies as being a factor in colorectal and possibly prostate cancer. Excess dietary fat is thought to induce cancer by a number of potential mechanisms, including the alteration of hormone levels, a change in the composition of cellular membranes, an increase in fatty acids (which may inhibit immune responses or serve as precursors to prostaglandins, which may then act as promoters), and a stimulation of the production of liver bile acids, some of which can act as promoters. Diet has been linked to numerous other cancers as well (Table 13.15). Microorganisms normally found in foods, such as fungi, are another potential source of carcinogens. For example, mycotoxins are prominently distributed in the food chain, and the prevalence of Aspergillus in the environment, a producer of dietary aflatoxins, appears to contribute significantly to the higher risk of liver cancer that is observed in some third world countries. Fusarium monilifome is ubiquitous in corn and produces fumonisins B1, B2, and fusarin C, all of which have been implicated in human esophageal cancer. Cooking is another factor that may alter the dietary carcinogen load. Cooking alters the chemical structure of foods, and cooking has long been known to produce cyclic compounds, a number of which

TABLE 13.15 Cancer Sites and Associated Risks/Benefits of Diets Probable Site of Cancer Colorectum

Breast Lung Stomach

Increases Risk

Salt, pickled and preserved food

Prostate Cervix Esophagus Pancreas

Bladder Liver

Decreases Risk

Red meat, processed Vegetables, meat nonstarch, polysaccharides Alcohol, red meat, Vegetables fried meat


Possible Increases Risk Alcohol, fat

Alcohol, meat Fruit and vegetables, vitamin C Vitamin E (Red) meat, fat Fruit and vegetables, vitamin C Fruit and vegetables Red meat

Fruit and vegetables Alcohol

Source: Adapted from Cummings and Bingham (1998).

Decreases Risk Folate

Fruit, phytoestrogens Fruit and vegetables Carotenoids Vegetables Folate, vitamin A

Fruit and vegetables, vitamin C, nonstarch, polysaccharides



are mutagens and carcinogens. For example, polycyclic heterocyclic amines (PHAs) are produced when any amino acid is pyrolyzed (e.g., in broiling a beefsteak), and many of these are highly mutagenic. Broiling and charring foods may also increase the presence of polyaromatic hydrocarbons (PAHs). Other carcinogens are among those chemicals that are frequently found as natural or added constituents of the foods that make up our diet (Tables 13.16 and 13.17) or as synthetic chemical pesticide or other residues (Table 13.18). For example, caffeic acid occurs in higher plants and has produced tumors in both male and female rats. The rodent carcinogen (rabbits, hamsters and mice) n-nitrosodimethylamine is found in cheeses, bacon, frankfurters, soybean oil, smoked or cured meats, fish, and some alcoholic beverages, including beer. Nitrates and nitrites occur naturally and are introduced to foods in curing and preserving processes. It has been argued that nitrites may form carcinogenic nitrosamines in the acid environment of the stomach by combining with amines of the aminoacids that form the protein in our diets. Thus, cooking, curing processes, applied chemicals (fertilizers, pesticides, soil or water contamination, etc.), and the selective growth of insect resistant plants are ways in which the carcinogenic load or potential of the foods we ingest may be altered. Typically, these sources outweigh the contributions by the application of synthetic pesticide by perhaps as much as 10,000-fold. So, although it is clear that naturally occurring chemicals outweigh the synthetic chemicals we are exposed to in our diet. However, the relative contribution to the incidence of cancer by these exposures is generally considered to be far less than is caused by the intake of excess calories via animal fat ingestion. Finally, diets deficient in iron, selenium, and vitamin C have all been associated with increased cancer rates. Vitamin C has been shown to inhibit the formation of certain initiating carcinogens, vitamin E appears to prevent promotion, and vitamin A appears to decrease the susceptibility of epithelial tissue to carcinogens. Overall, the evidence indicates diet can have a profound effect on the incidence of cancer, and estimates that have diet contributing to as high as 70 percent of the total cancer incidence [perhaps as much as 80 percent of large bowel (colon) and breast cancers] can be found in the scientific literature. In addition, differences in diet may explain some regional geographic differences in the distribution and frequency of the cancer types observed. Like drinking alcohol and smoking, diet can also have a unknown impact on the results of epidemiologic investigations, an impact that is often inadequately investigated.

TABLE 13.16 Natural Pesticides and Metabolites Found in Cabbage Glucosinolates: 2-propenyl glucosinolate (sinigrin),a 3-methylthiopropyl glucosinolate, 3methylsulfinylpropyl glucosinolate, 3-butenyl glucosinolate, 2-hydroxy-3-butenyl glucosinolate, 4methylsulfinylbutyl glucosinolate, 4-methylsulfonylbutyl glucosinolate, benzyl glucosinolate, 2-phenylethyl glucosinolate, propyl glucosinolate, butyl glucosinolate Indole glucosinolate and related indoles: 3-indolylmethyl glucosinolate (glucobrassicin), 1-methoxy-3indolylmethyl glucosinolate (neoglucobrassicin), indole-3-carbinol,a indole-3-acetonitrile, bis(3indolyl)methane Isothiocyanates and goitrin: allyl isothiocyanate,a 3-methylthiopropyl isothiocyanate, 3-methylsulfinylpropyl isothiocyanate, 3-butenyl isothiocyanate, 5-vinyloxazolidine-2-thione (goitrin), 4-methylthiobutyl isothiocyanate, 4-methylsulfinylbutyl isothiocyanate, 4-methylsulfonylbutyl isothiocyanate, 4-pentenyl isothiocyanate, benzyl isothiocyanate, phenylethyl isothiocyanate Cyanides: 1-cyano-2,3-epithiopropane, 1-cyano-3,4-epithiobutane, 1-cyano-3,4-epithiopentane, threo--1cyano-2-hydroxy-3,4-epitiobutane, erythro--1 -cyano-2-hydroxy-3,4-epithiobutane, 2-phenylpropionitrile, allyl cyanide,a 1-cyano-2-hydroxy-3-butene, 1-cyano-3-methylsulfinylpropane, 1-cyano-4methylsulfinylbutane Terpenes: menthol, neomenthol, isomenthol carvonea a

Indicates data on mutagenicity or carcinogenicity (see Ames et al. 1990 for discussion of data); others untested. Source: Adapted from Ames et al. (1990)



TABLE 13.17 Naturally Occurring Carcinogens Potentially Present in U.S. Diets Constitutive: acetaldehyde, benzene, caffeic acid, cobalt, estradiol 17β, estrone, ethyl acrylate, (with UV light exposure), 8-methoxypsoralen (xanthotoxin) (with UV light exposure), progesterone, safrole, styrene, testosterone Derived: A-alpha-C, acetaldehyde, benz(a)anthracene, benzene, benzo(a)pyrene, benzo(b)fluoranthene, benzo(j)fluoranthene, benzo(k)fluoranthene, dibenz(a,h)acridine, dibenz(a,j)acridine, dibenz(a,h)anthracene, formaldehyde, glu-P1, glu-P2, glycidaldehyde, IQ, Me-A-alpha-C, MEIQ, MeIQx, methyl mercury compounds, N-methyl-N′-nitro-nitrosoquanidine, N-nitroso-N-dibutylamine, N-nitorosodiethylamine, Nnitrosodimethylamine, N-nitrosodi-N-propylamine, N-nitorosomehtylethylamine, N-nitrosopiperidine, Nnitrosopyrrolidine, N-nitrososarcosine, PhIP, Trp-P1, Trp-P2, urethane Acquired: aflatoxin B1, aflatoxin M1, ochratoxin A, sterigmatocystin, toxins derived from Fusarium moniliforme Pass through: arsenic, benz(a)anthracene, benzo(a)pyrene, beryllium, cadmium, chromium, cobalt, indeno(1,2,3)pyrene, lead, nickel Added: Contaminant introduced through tap water: arsenic, asbestos, benzene, beryllium, cadmium, hexavalent chromium, dibenzo(a,l)pyrene, indenol(1,2,3,-cd)pyrene, radon Indirect through use as a drug or in packaging: i) veterinary drugs—estradiol 17β, progesterone, reserpine, testosterone, ii) food packaging material—benzene, cobalt, ethyl acrylate, formaldehyde, nickel Direct food additives: acetaldehyde, ethyl acrylate, formaldehyde Traditional foods and beverages: alcoholic beverages, betel liquid, bracken fern, hot mate, pickled vegetables, salted fish (Chinese style) Source: Adapted from Table 5-1, NRC (1996). Reprinted with permission from Carcinogens and Anticarcinogens in the Human Diet. Copyright 1996 by the National Academy of Sciences. Courtesy of the National Academy Press, Washington, D.C.

Iatrogenic Cancer The use of drugs that might impact the cancer incidence in a given population, is rarely addressed in the mortality studies of occupational cohorts from which we derive much of our knowledge regarding chemical carcinogenicity. No chemical has only one effect and pharmaceutical medications are no exception to this rule. Pharmaceuticals are known to be capable of producing side effects other than the desired therapeutic effect. A surprising number of drugs are known to have carcinogenic effects. Perhaps the most well-known class of agents with such effects is, of course, the potent chemotherapeuTABLE 13.18 Synthetic Carcinogens that Might Be Present in Foods Pesticide residues: acrylonitrile, amitrole, aramite, atrazine, benzotrichloride, 1,3-butadiene, captafol, carbon tetrachloride, chlordane, Kepone, chloroform, 3-chloro-2-methylpropene, p-chloro-o-toluidine, chlorophenoxy herbicides, creosotes, DDD, DDE, DDT, Dichlorvos, 1,2-dibromo-3-chloropropane, pdichlorobenzene, 1,2-dichloroethane, 2-dichloroethane, dichloromethane, 1,3-dichloropropene, dimethylcarbamoyl chloride, 1,1-dimethylhydrazine, ethylene dibromide, ethylene thiourea, heptachlor, hexachlorobenzene, hexachlorocyclohexane, mirex, N-nitrosodiethanolamine, pentachlorophenol, ophenylphenate, nitrofen, 1,3-propane sulfone, propylene oxide, styrene oxide, sulfallate, tetrachlorodibenzop-dioxin, thiourea (past), toxaphene, 2,4,6-trichlorophenol Potential animal drug residues: diethylstilbesterol (now banned), ethinyl estradiol, medroxyprogesterone acetate, methylthiouracil, N-[4-(5-nitro-2-furyl)-2-thiazolyl]acetamide, nortestosterone, propylthiouracil Packaging or storage container migrants: acrylamide, acrylonitrile, 2-aminoanthraquinone, BHA, 1,3butadiene, chlorinated paraffins, carbon tetrachloride, chloroform, 2-diaminotoluene, di(2ethylhexyl)phthalate, dimethylformamide, diethyl sulfate, dimethyl sulfate, 1,4-dioxane, ethyl acrylate, epichlorohydrin, ethylene oxide, ethylene thiourea, 2-methylaziridine, 4,4′-methylenedianiline, 4,4′methylene bis(2-chloroaniline) (now prohibited), 2-nitropropane, 1-nitropyrene, phenyl glycidyl ether, propylene oxide, sodium phenyl phenate, sodium saccharin, styrene, styrene oxide, tetrachloroethylene, toluene diisocyanate, vinyl chloride Residues from food processing: dichloromethane, epichlorohydrin, NTA trisodium salt monohydrate Source: Adapted from Table 5-3, NRC (1996). Reprinted with permission from Carcinogens and Anticarcinogens in the Human Diet. Copyright 1996 by the National Academy of Sciences. Courtesy of the National Academy Press, Washington, D.C.



TABLE 13.19 Pharmaceutical Agentsa with Carcinogenic Effectsb Generic Name

Therapeutic Use

Daily Dosage (mg/day)

Rifampin Isoniazid Clofibrate Disulfiram Phenobarbital Acetaminophen Metronidazole Sulfisoxazole Dapsone

Antibiotic: tuberculosis Antibiotic: tuberculosis Lowers cholesterol Discourages alcohol abuse Antiepileptic Pain relief (OTC) Antibiotic, antiparasitic Antibiotic, urinary tract Antibacterial, AIDS, leprosy, etc.

600 300 2000 125–500 100–200 2000–4000 500 8000 300








Water retention in disease states


Tumor Site; Species Liver; mice Lung; mice Liver; mice Liver; rats Liver; mice Liver; mice Lung; rats/mice Spleen, thyroid, and peritoneum; rats Thyroid and pituitary tumors; rats Thyroid, testes, prostate; rats/mice liver; mice


List adapted from Waddell (1996). Cancer effects as listed in the Physicians Desk Reference (PDR), 1996, or Ames and Gold (1991).


tic agents used to treat cancer. Many antineoplastic drugs are potent genotoxic chemicals, and their damage to DNA in rapidly dividing cells like cancer cells is a primary feature of both their therapeutic effects and toxicities. Admittedly, it may be well worth a theoretical risk of developing cancer 20 years after taking medication to cure a current case of cancer, however, a number of drugs whose therapeutic benefits are directed at less serious health conditions are also known to have carcinogenic effects in humans or in animal cancer bioassays. Some potentially carcinogenic pharmaceuticals are listed in Table 13.19. Not only are many of the drugs listed in Table 13.19 commonly prescribed, but the single daily doses of these chemicals are large relative to the doses of chemicals one is typically concerned with when evaluating environmental pollutants. Thus, the theoretical risks associated with even limited therapy may approach or exceed the theoretical risks posed by the environmental contamination we are often concerned about when remediating sites that contain these contaminants.


Human Cancer Trends in the United States As mentioned regarding smoking, the incidence of cancer in this nation has remained stable, or declined, for most types of cancer according to the American Cancer Society. The greatest exception is, of course, lung cancer in both males and females. A 1998 report from the National Cancer Institute (NCI) (see Table 13.20) indicated that after increasing 1.2 percent per year from 1973 to 1990, incidence for all cancers combined declined in the United States an average of 0.7 percent from 1990 to 1995. Cancer mortality similarly declined about 0.5 percent per annum for the same period (1990–1995). Cancers of the lung, breast, prostate, and colon–rectum accounted for over half of the new cases. Cancer of the lung, both incidence and mortality, is actually showing a slight decline while in women, such cancers (and the resultant mortality) are still on the increase. Incidence and mortality



TABLE 13.20 Estimated New Cancer Cases and Deaths by Sex for All Sites, United States, 1999a Estimated New Cases Cancer Sites All sites Oral cavity and pharynx Digestive system Esophagus Stomach Small intestine Colon Rectum, anus, etc. Liver and intrahepatic bile duct Gallbladder and other biliary Pancreas Other digestive organs Larynx Lung and bronchus Other respiratory organs Bones and joints Soft tissue (including heart) Skin (no basal and squamous) Breast Genital system Uterine corpus Ovary Vulva Vagina and other female genitalia Prostate Testis Penis and other genitalia, male Urinary bladder Kidney and renal pelvis Ureter and other urinary Eye and orbit Brain and other nervous system Endocrine system Hodgkin’s disease Non-Hodgkin’s lymphoma Multiple myeloma All leukemias Other primary sites

Both Sexes 1,221,800 29,800 226,300 12,500 21,900 4,800 94,700 38,000 14,500 7,200 28,600 4,100 10,600 171,600 5,400 2,600 7,800 54,000 176,300 269,100 37,400 25,200 3,300 2,300 179,300 7,400 1,400 54,200 30,000 2,300 2,200 16,800 19,800 7,200 56,800 13,700 30,200 35,100

Male 623,800 20,000 117,200 9,400 13,700 2,500 43,000 19,400 9,600 3,000 14,000 1,200 8,600 94,000 4,200 1,400 4,200 33,400 1,300 188,100

179,300 7,400 1,400 39,100 17,800 1,500 1,200 9,500 5,400 3,800 32,600 7,300 16,800 16,400

Female 598,000 9,800 109,100 3,100 8,200 2,300 51,700 15,300 4,900 4,200 14,600 2,900 2,000 77,600 1,200 1,200 3,600 20,600 175,000 81,000 37,400 25,200 3,300 2,300

15,100 12,200 800 1,000 7,300 14,400 3,400 24,200 6,400 13,400 18,700

Estimated Deaths Both Sexes 563,100 8,100 131,000 12,200 13,500 1,200 47,900 8,700 13,600 3,600 28,600 1,200 4,200 158,900 1,100 1,400 4,400 9,200 43,700 64,700 6,400 14,500 900 600 37,000 300 200 12,100 11,900 500 200 13,100 2,000 1,300 25,700 11,400 22,100 36,100

Male 291,100 5,400 69,900 9,400 7,900 600 23,000 4,800 8,400 1,300 13,900 400 3,300 90,900 700 800 2,100 5,800 400 37,500

37,000 300 200 8,100 7,200 300 100 7,200 900 700 13,400 5,800 12,400 18,200

Female 272,000 2,700 61,100 2,800 5,600 600 24,900 3,900 5,200 2,300 14,700 800 900 68,000 400 600 2,300 3,400 43,300 27,200 6,400 14,500 900 600

4,000 4,700 200 100 5,900 1,100 600 12,300 5,600 9,700 17,900

a Excludes basal and squamous cell skin cancers and in situ carcinomas except urinary bladder. Carcinoma in situ of the breast accounts for about 39,900 new cases annually, and melanoma carcinoma in situ accounts for about 23,200 new cases annually. Estimates of new cases are based on incidence rates from the NCI SEER program, 1979–1995. American Cancer Society, Surveillance Research, 1999.



from non-Hodgkin’s lymphoma and from melanoma are also increasing. These data were confirmed in the 1999 joint release from the CDC, NCI, and ACS. As awareness of environmental contamination and the ubiquity of synthetic chemicals arose in the 1960s, specifically after the release of Rachel Carson’s Silent Spring in 1962, speculation persisted that we were awash in a “ sea of carcinogens” and that after an appropriate latency interval, a cancer epidemic would hit. As has been shown in Figure 13.1, the hypothesized epidemic of cancer has never arrived, and considering the data indicating decreasing cancers through 1995, it would seem that perhaps our current reductions in smoking, food consumption, and alcohol might be starting to impact the incidence of new cancers in the United States. Considering that the stability of the cancer incidence (aside from lung cancer due primarily to smoking) occurred during a period of great industrialization in the United States, the impact of occupation and environmental pollution on cancer incidence is probably less than what was postulated 30 years ago. That is not to say, however, that exposure reductions are not still warranted in these areas, but merely to point out that the current data indicate that our future, with its concomitant exposure to new synthetic chemicals, is not a dire one.

13.11 SUMMARY Chemical-induced carcinogenesis represents a unique and complex area within toxicology. The difficulty in assessing the carcinogenic hazards and human risks of chemicals stems from the following characteristics of chemical carcinogenesis:

• It is a multistage process involving at least two distinct stages: initiation, which converts the genetic expression of the cell from a normal to aberrant cell line; and promotion, in which the aberrant cell is stimulated in some fashion to grow, thereby expressing its altered state.

• Since chemicals may increase cancer incidence at various stages and by different mechanisms, the term carcinogen by itself is somewhat limiting and a number of descriptive labels are applied to the chemical carcinogens that define or describe these differences, such as cocarcinogens, initiators, promoters, and epigenetic.

• Chemicals may produce or affect only a single stage or a single aspect of carcinogenesis that leads to a number of important differences and considerations about the potential health impacts of chemical carcinogens. Perhaps the most important considerations are the concept of thresholds and that qualitative differences do exist among carcinogens.

• Carcinogenicity testing raises many questions about interpretations of results. Considerations such as mechanism (genotoxic vs. epigenetic), dose, and relevant test species, are important in determining probable human risk; thus, many additional toxicity test data are needed to improve the extrapolation of cancer bioassay data from test species to humans.

• A number of lifestyle-related factors influence carcinogenesis, altering the risks posed by carcinogenic chemicals and acting to confound epidemiological evidence. Considering the complexities involved in (1) determining the mechanism of cancer causation, (2) using animal and human data to identify carcinogenic substances, and (3) using these data to extrapolate risks with the aim of reducing or eliminating environmental risk factors, it should be clear to the reader that the best approach to occupational carcinogenesis is an interdisciplinary one. As depicted in Figure 13.13, identifying and reducing occupational cancer requires the interfacing of several scientific disciplines and several kinds of health professionals:

• The toxicologist is responsible for testing and identifying chemical carcinogens; through animal testing the toxicologist attempts to provide information about carcino-



Figure 13.13 Identifying and reducing chemical carcinogens requires an interdisciplinary approach, in which health professions interface with other scientific disciplines.

genic mechanisms, and about species differences or similarities that can aid in assessing the human risk.

• Epidemiologists add human evidence to risk evaluations or ascertain if a chemical should or should not be considered a human carcinogen for various reasons (it may have weak or undetectable activity).

• Specialists in occupational medicine provide health surveillance programs to protect the health status of the worker and attempt to prevent those exposures that could lead to serious, chronic health problems.

• Industrial hygienists help design better methods for evaluating and preventing worker exposures; and biometrists and computer scientists aid in risk analysis, data storage, and data analysis. As long as these disciplines are utilized jointly and their relationship to the occupational carcinogenesis problem is understood, occupational health and safety professionals can have good reason to hope for improved success in the prevention of occupational carcinogenesis.



REFERENCES AND SUGGESTED READING ACS (American Cancer Society), Cancer Facts and Figures. New York: American Cancer Society (1999). Ames, B. N., M. Profet, and L. S. Gold, Dietary pesticides (99.99% all natural). Proceedings of the National Academy of Science, 87: 7777–7781 (1990). Ames, B. N., L. S. Gold, and M. K. Shigenaga, Cancer prevention, rodent high-dose cancer tests, and risk assessment. Risk Analysis, 16: 613–617 (1996). Ames, B. N., and L. S. Gold, The causes and prevention of cancer: Gaining perspective. Environmental Health Perspectives, 105(Sup4): 865–873 (1997). Arcos, J. C., M. F. Argus, and Y. T. Woo, Chemical Induction Of Cancer: Modulation and Combination Effects, Birauser, Boston, MA, (1995). Ashby, J., F. R. Johannsen, G. K. Raabe, N. G. Doerrer, S. C. Lewis, R. C. Reynolds, F. G. Flamm, N. D. Krivanek, J. M. Smith, J. T. Stevens, J. E. Harris, J.