Ebook Cancer epidemiology and prevention (4/E): Part 1

(BQ) Part 1 book “Cancer epidemiology and prevention” has contents: Morphological and molecular classification of human cancer, genetic epidemiology of cancer, application of biomarkers in cancer epidemiology, causal inference in cancer epidemiology,… and other contents.

Trang 2

Schottenfeld and Fraumeni

Cancer Epidemiology and Prevention

Trang 4

Schottenfeld and Fraumeni

Cancer Epidemiology and

Prevention

Fourth Edition

Lead Editor

MICHAEL J THUN, MD, MS

Epidemiology and Surveillance Research (Retired)

American Cancer Society

Atlanta, Georgia

Co- Editors

MARTHA S LINET, MD, MPH

Division of Cancer Epidemiology and Genetics

National Cancer Institute

DAVID SCHOTTENFELD, MD, MSC

Department of Epidemiology (Retired) University of Michigan School of Public Health Ann Arbor, Michigan

Project Manager

ANNELIE M. LANDGREN, MPH, PMP

Trang 5

1Oxford University Press is a department of the University of Oxford It furthersthe University’s objective of excellence in research, scholarship, and education

by publishing worldwide Oxford is a registered trade mark of Oxford UniversityPress in the UK and certain other countries

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America

a retrieval system, or transmitted, in any form or by any means, without theprior permission in writing of Oxford University Press, or as expressly permitted

by law, by license, or under terms agreed with the appropriate reproductionrights organization Inquiries concerning reproduction outside the scope of theabove should be sent to the Rights Department, Oxford University Press, at theaddress above

You must not circulate this work in any other formand you must impose this same condition on any acquirer

Library of Congress Cataloging-in-Publication Data Names: Thun, Michael J., editor | Linet, Martha S., editor | Cerhan, James R., editor | Haiman, Christopher, editor | Schottenfeld, David, editor Title: Schottenfeld and Fraumeni Cancer Epidemiology and Prevention / lead editor, Michael J Thun ; co-editors, Martha S Linet, James R Cerhan,

Christopher Haiman, David Schottenfeld ; project manager, Annelie M Landgren Other titles: Cancer epidemiology and prevention

Description: Fourth edition | New York, NY : Oxford University Press, [2018] | Preceded by Cancer epidemiology and prevention / edited by David Schottenfeld, Joseph F Fraumeni Jr 3rd ed 2006 | Includes bibliographical references and index Identifiers: LCCN 2017038170 | ISBN 9780190238667 (hardcover : alk paper) Subjects: | MESH: Neoplasms—epidemiology | Neoplasms—prevention & control Classification: LCC RA645.C3 | NLM QZ 220.1 |

DDC 614.5/999—dc23

LC record available at https://lccn.loc.gov/2017038170

9 8 7 6 5 4 3 2 1Printed by Sheridan Books, Inc., United States of America

Trang 6

Michael Dean and Karobi Moitra

Mark E Sherman, Melissa A Troester, Katherine A Hoadley, and William F Anderson

Christopher J Maher and Elaine R. Mardis

Kathryn L Penney, Kyriaki Michailidou, Deanna Alexis Carere, Chenan Zhang, Brandon Pierce, Sara Lindström, and Peter Kraft

Roel Vermeulen, Douglas A Bell, Dean P Jones, Montserrat Garcia- Closas, Avrum Spira, Teresa W Wang, Martyn T Smith, Qing Lan, and Nathaniel Rothman

Steven N Goodman and Jonathan M. Samet

II THE MAGNITUDE OF CANCER

Ahmedin Jemal, D Maxwell Parkin, and Freddie Bray

Candyce Kroenke and Ichiro Kawachi

K Robin Yabroff, Gery P Guy Jr., Matthew P Banegas, and Donatus U Ekwueme

Trang 7

vi Contents

III THE CAUSES OF CANCER

Michael J Thun and Neal D Freedman

Susan M Gapstur and Philip John Brooks

Marjorie L McCullough and Walter C Willett

NaNa Keum, Mingyang Song, Edward L Giovannucci, and A Heather Eliassen

Steven C Moore, Charles E Matthews, Sarah Keadle, Alpa V Patel, and I- Min Lee

Robert N Hoover, Amanda Black, and Rebecca Troisi

Marie C Bradley, Michael A O’Rorke, Janine A Cooper, Søren Friis, and Laurel A. Habel

Silvia Franceschi, Hashem B El- Serag, David Forman, Robert Newton, and Martyn Plummer

Eric A Engels and Allan Hildesheim

IV CANCERS BY TISSUE OF ORIGIN

Ellen T Chang and Allan Hildesheim

Andrew F Olshan and Mia Hashibe

Michael J Thun, S Jane Henley, and William D. Travis

Mia Hashibe, Erich M Sturgis, Jacques Ferlay, and Deborah M. Winn

William J Blot and Robert E. Tarone

Trang 8

Contents vii

Catherine de Martel and Julie Parsonnet

Samuel O Antwi, Rick J Jansen, and Gloria M Petersen

W Thomas London, Jessica L Petrick, and Katherine A McGlynn

Jill Koshiol, Catterina Ferreccio, Susan S Devesa, Juan Carlos Roa, and Joseph F Fraumeni, Jr.

Jennifer L Beebe- Dimmer, Fawn D Vigneau, and David Schottenfeld

Kana Wu, NaNa Keum, Reiko Nishihara, and Edward L.Giovannucci

Henrik Hjalgrim, Ellen T Chang, and Sally L. Glaser

James R Cerhan, Claire M Vajdic, and John J Spinelli

Mark P Purdue, Jonathan N Hofmann, Elizabeth E Brown, and Celine M. Vachon

Lisa Mirabello, Rochelle E Curtis, and Sharon A. Savage

Marianne Berwick and Charles Wiggins

Rolando Herrero and Raul Murillo

Margaret M Madeleine and Lisa G Johnson

Trang 9

Eve Roman, Tracy Lightfoot, Susan Picton, and Sally Kinsey

Lindsay M Morton, Sharon A Savage, and Smita Bhatia

V CANCER PREVENTION AND CONTROL

Michael J Thun, Christopher P Wild, and Graham Colditz

Jeffrey Drope, Clifford E Douglas, and Brian D. Carter

Ambika Satija and Frank B. Hu

Marc Bulterys, Julia Brotherton, and Ding- Shinn Chen

Robyn M Lucas, Rachel E Neale, Peter Gies, and Terry Slevin

Trang 10

Acknowledgments

We are indebted to the more than 190 chapter authors who generously contributed their time,

labor, and expertise to produce this comprehensively updated fourth edition The multi- authored

text reflects the increasingly interdisciplinary and collaborative nature of the field; it provides a

resource for researchers seeking to harness the unprecedented advances in genetic and

molecu-lar research into molecu-large- scale population studies of cancer etiology, and ultimately into effective

preventive interventions We owe special thanks to Ms Annelie Landgren, whose energy,

enthu-siasm, and organizational expertise as project manager have been invaluable in bringing this

text to completion We also thank Dr. Stephen Chanock for his early and unfailing

encourage-ment and for supporting the critical infrastructure necessary for such a collaborative enterprise

This book would not have been possible without the generous forbearance of our spouses and

families Finally, Michael Thun thanks Dr. Lynne Moody for her insights as a sounding board

throughout this process.

Trang 12

Contributors

E Susan Amirian, PhD

Dan L Duncan Cancer Center

Baylor College of Medicine

Houston, Texas

William F Anderson, MD, MPH

Division of Cancer Epidemiology and Genetics

National Cancer Institute

Bethesda, Maryland

Samuel O Antwi, PhD

Department of Health Sciences Research

Mayo Clinic College of Medicine

Rochester, Minnesota

Bruce K Armstrong, MD, PhD*

School of Public Health

The University of Sydney

Sydney, New South Wales, Australia

Matthew P Banegas, PhD, MPH

Center for Health Research

Kaiser Permanente

Portland, Oregon

Jill Barnholtz- Sloan, PhD

Case Comprehensive Cancer Center

Case Western Reserve University School of Medicine

Cleveland, Ohio

Dorothea T Barton, MD

Department of Surgery

Dartmouth- Hitchcock Medical Center

Lebanon, New Hampshire

Laura E Beane Freeman, PhD

Jennifer L Beebe- Dimmer, PhD, MPH

Wayne State University School of Medicine Karmanos Cancer Institute

Amy Berrington de González, DPhil

Division of Cancer Epidemiology and Genetics National Cancer Institute

Marianne Berwick, PhD

Department of Internal Medicine University of New Mexico Albuquerque, New Mexico

Smita Bhatia, MD, MPH

Institute of Cancer Outcomes and Survivorship University of Alabama at Birmingham, School of Medicine Birmingham, Alabama

Trang 13

xii Contributors

Melissa L Bondy, PhD

Department of Medicine, Section of Epidemiology and Population Sciences

Houston, Texas

AndrÉ Bouville, PhD*

Marie C Bradley, PhD, MScPH

Division of Cancer Control and Population Sciences

Freddie Bray, PhD

Section of Cancer Surveillance

International Agency for Research on Cancer

Lyon, France

Alina V Brenner, MD, PhD

Louise A Brinton, PhD

National HPV Vaccination Program Register

Victorian Cytology Service

East Melbourne, Victoria, Australia

HIV/ Hepatitis Department

World Health Organization

Geneva, Switzerland

Kenneth P Cantor, PhD, MPH*

Deanna Alexis Carere, ScD, CGC

Department of Pathology and Molecular Medicine

McMaster University

Hamilton, Ontario, Canada

Brian D Carter, MPH

Epidemiology Research Program

American Cancer Society

Atlanta, Georgia

James R Cerhan, MD, PhD, (Editor)

Department of Health Sciences Research Mayo Clinic

Rochester, Minnesota

Ellen T Chang, ScD

Center for Health Sciences Exponent Inc.

Menlo Park, California

Ding- Shinn Chen, MD

Hepatitis Research Center National Taiwan University Hospital Taipei, Taiwan

Wong- Ho Chow, PhD

Department of Epidemiology The University of Texas

MD Anderson Cancer Center Houston, Texas

Janine A Cooper, PhD

School of Pharmacy Queen’s University Belfast Belfast, Northern Ireland

Catherine de Martel, MD, PhD

Infections and Cancer Epidemiology Group International Agency for Research on Cancer Lyon, France

‡ Consultant/ Contractor

* Retired

Trang 14

Contributors xiii

Michael Dean, PhD

Susan S Devesa, PhD*

Clifford E Douglas, JD

Center for Tobacco Control

Atlanta, Georgia

Jeffrey Drope, PhD

Economic & Health Policy Research

Atlanta, Georgia

Donatus U Ekwueme, PhD, MS

National Center for Chronic Disease Prevention and Health Promotion

Centers for Disease Control and Prevention

Atlanta, Georgia

A Heather Eliassen, ScD

Brigham & Women’s Hospital and Harvard Medical School

Harvard TH Chan School of Public Health

Boston, Massachusetts

Hashem B El- Serag, MD, MPH

Gastroenterology and Hepatology

Houston, Texas

Eric A Engels, MD

Jacques Ferlay, MSc

Section of Cancer Surveillance

Lyon, France

Catterina Ferreccio, MD, MPH

Division of Public Health and Family Medicine

School of Medicine, Pontificia Universidad Católica de Chile

Morten Frisch, MD, PhD, DrSci(Med)

Department of Epidemiology Research Statens Serum Institut

Copenhagen, Denmark

Susan M Gapstur, PhD, MPH

Epidemiology Research Program American Cancer Society Atlanta, Georgia

Montserrat Garcia- Closas, MD, DrPH

Mia M Gaudet, PhD

Epidemiology Research Program American Cancer Society Atlanta, Georgia

Trang 15

xiv Contributors

Steven N Goodman, MD, PhD

Department of Medicine, Clinical and Translational Research

Stanford University School of Medicine

Population Health Division

QIMR Berghofer Medical Research Institute

Brisbane, Queensland, Australia

Andrew E Grulich, PhD

Kirby Institute

The University of New South Wales

Christopher A Haiman, ScD, (Editor)

Department of Preventive Medicine

Keck School of Medicine of University of Southern California

Los Angeles, California

Lineberger Comprehensive Cancer Center

University of North Carolina School of Medicine

Chapel Hill, North Carolina

Mia Hashibe, PhD

Department of Family and Preventive Medicine

Huntsman Cancer Institute, University of Utah School of Medicine

Salt Lake City, Utah

S Jane Henley, MSPH

Division of Cancer Prevention and Control

US Centers for Disease Control and Prevention

Atlanta, Georgia

Rolando Herrero, MD, PhD

Prevention and Implementation Group

Lyon, France

Allan Hildesheim, PhD

Henrik Hjalgrim, MD, PhD, DrSci(med)

Department of Epidemiology Research Statens Serum Institut

Rick J Jansen, MS, PhD

Department of Public Health North Dakota State University Fargo, North Dakota

Lisa G Johnson, PhD, MPH

Division of Public Health Sciences Fred Hutchinson Cancer Research Center Seattle, Washington

Dean P Jones, PhD

Department of Medicine Emory University Atlanta, Georgia

Margaret R Karagas, PhD

Department of Epidemiology Geisel School of Medicine at Dartmouth Hanover, New Hampshire

§ adjunct

Trang 16

Department of Pediatric Hematology

Leeds Teaching Hospitals NHS Trust

Leeds, United Kingdom

Cari M Kitahara, PhD

Jill Koshiol, PhD

Stella Koutros, PhD

Peter Kraft, PhD

Departments of Epidemiology and Biostatistics

Barnett S Kramer, MD, MPH

Division of Cancer Prevention

Martha S Linet, MD, MPH, (Editor)

Robyn M Lucas, MBChB, MPH&TM, PhD

Radiation Health Services Branch The Australian National University Canberra, The Australian Capital Territory, Australia

Margaret M Madeleine, PhD

Trang 17

xvi Contributors

Steven C Moore, PhD

Lindsay M Morton, PhD

Raul Murillo, MD

Prevention and Implementation Group

Lyon, France

Rachel E Neale, PhD

Population Health Division

QIMR Berghofer Medical Research Institute

Brisbane, Queensland, Australia

Heather H Nelson, MPH, PhD

Division of Epidemiology, Masonic Cancer Center

University of Minnesota, Twin Cities

Minneapolis, Minnesota

Marian L Neuhouser, PhD

Division of Public Health Sciences

Fred Hutchinson Cancer Research Center

School of Medicine Dentistry and Biomedical Sciences

Queen’s University Belfast

Belfast, Northern Ireland

Andrew F Olshan, PhD

Department of Epidemiology

Gillings School of Global Public Health

University of North Carolina

Chapel Hill, North Carolina

Quinn T Ostrom, MA, MPH

Case Comprehensive Cancer Center

Case Western Reserve University School of Medicine

Nuffield Department of Public Health University of Oxford

Oxford, United Kingdom

Kathryn L. Penney, ScD

ScD Department of Medicine Brigham and Women’s Hospital / Harvard Medical School Boston, Massachusetts

Gloria M Petersen, PhD

Department of Health Sciences Research Mayo Clinic College of Medicine Rochester, Minnesota

I Mary Poynten, PhD Kirby Institute

The University of New South Wales Sydney, New South Wales, Australia

Ludmila Prokunina- Olsson, PhD

Ewa Rajpert- De Meyts, MD, PhD

Department of Growth & Reproduction Copenhagen University Hospital Rigshospitalet Copenhagen, Denmark

Judy R Rees, BM, BCh, PhD

Department of Epidemiology Geisel School of Medicine at Dartmouth Hanover, New Hampshire

Juan Carlos Roa, MD, Msc

Department of Pathology School of Medicine, Pontificia Universidad Católica de Chile Santiago, Chile

Trang 18

Jonathan M Samet, MD

Department of Preventive Medicine

Keck School of Medicine of University of Southern California

Los Angeles, California

Section of Endocrinology, Diabetes and Metabolism

University of Illinois at Chicago College of Medicine

Chicago, Illinois

David Schottenfeld, MD, MSc (Editor)*

Department of Epidemiology

University of Michigan School of Public Health

Ann Arbor, Michigan

Mary Schubauer- Berigan, PhD

Division of Surveillance Hazard Evaluation and Field Studies

Centers for Disease Control and Prevention

Atlanta, Georgia

Joachim Schüz, PhD

Section of Environment and Radiation

Lyon, France

Amy L Shafrir, ScD

Division of Adolescent and Young Adult Medicine

Boston Children’s Hospital

Mark E Sherman, MD

Health Sciences Research

Mayo Clinic College of Medicine

Jacksonville, Florida

Debra T Silverman, ScD, ScM

Terry Slevin, MPH

Cancer Council Western Australia

Perth, Western Australia, Australia

Martyn T Smith, PhD

School of Public Health University of California at Berkeley Berkeley, California

Mingyang Song, MD, ScD

Clinical and Translational Epidemiology Unit and Division

of Gastroenterology Massachusetts General Hospital and Harvard Medical School Boston, Massachusetts

John J Spinelli, PhD Cancer Control Research

British Columbia Cancer Agency Vancouver, British Columbia, Canada

Avrum Spira, MD, MSc

Division of Computational Biomedicine Boston University School of Medicine Boston, Massachusetts

Janet L Stanford, PhD

Erich M Sturgis, MD, MPH

Department of Head and Neck Surgery The University of Texas MD Anderson Cancer Center Houston, Texas

Catherine M Tangen, DrPH

Robert E Tarone, PhD*

International Epidemiology Institute Rockville, Maryland

Michael J Thun, MD, MS (Editor)*

Epidemiology and Surveillance Research American Cancer Society

Atlanta, Georgia

William D Travis, MD

Department of Pathology Memorial Sloan Kettering Cancer Center New York, New York

Melissa A Troester, PhD

Department of Epidemiology Lineberger Comprehensive Cancer Center University of North Carolina at Chapel Hill Chapel Hill, North Carolina

* Retired

Trang 19

xviii Contributors

Rebecca Troisi, ScD, MA

Shelley S Tworoger, PhD

Harvard Medical School and the Brigham and Women’s Hospital

Centre for Big Data Research in Health

University of New South Wales

Wayne State University School of Medicine

Karmanos Cancer Institute

Detroit, Michigan

Teresa W Wang, PhD

Division of Computational Biomedicine

Boston University School of Medicine

Mary H Ward, PhD

Charles Wiggins, PhD, MPH

Department of Internal Medicine University of New Mexico Albuquerque, New Mexico

Christopher P Wild, PhD

Director’s Office International Agency for Research on Cancer Lyon, France

Walter C Willett, MD, DrPH

Department of Nutrition Harvard TH Chan School of Public Health Boston, Massachusetts

K Robin Yabroff, PhD

Division of Cancer Control and Population Sciences National Cancer Institute

Chenan Zhang, PhD

Department of Epidemiology and Biostatistics University of California, San Francisco San Francisco, California

‡ Consultant/ Contractor

Trang 20

Preface

The Schottenfeld and Fraumeni text on Cancer Epidemiology and Prevention has served as the

premier reference text for population research on the causes and prevention of cancers since the

publication of the first edition in 1982 (Schottenfeld and Fraumeni, 1982) It is written for

col-leagues pursuing careers in research in cancer epidemiology and, more broadly, in preventive

oncology The founding editors, Dr. David Schottenfeld, now emeritus professor of

epidemiol-ogy at the University of Michigan, and Dr. Joseph Fraumeni, recently retired as the director of the

Division of Cancer Epidemiology and Genetics at the National Cancer Institute (NCI), updated

their landmark text in 1996 and 2006 (Schottenfeld and Fraumeni, 1996, 2006).

The current edition again provides a comprehensive update of research advances in cancer

epidemiology, prevention, and related fields in the past 10– 15 years, and honors the founding

editors in the title The new editorial team is led by Dr. Michael Thun (editor- in- chief), formerly

with the American Cancer Society, and includes four senior co- editors: Drs Martha Linet from

NCI, James Cerhan from the Mayo Clinic, Christopher Haiman from the University of Southern

California, and David Schottenfeld We are also deeply indebted to the internationally

recog-nized experts who authored the 63 chapters Without their generous effort and commitment, this

updated synthesis would not be possible.

Trang 22

MICHAEL J THUN, MARTHA S LINET, JAMES R CERHAN, CHRISTOPHER A HAIMAN, AND DAVID SCHOTTENFELD

In this introduction, we provide an overview of the text and highlight

cross- cutting developments and new opportunities that are

trans-forming our understanding of the causes and prevention of cancer As

in previous editions, the text is grouped into five major parts: “Basic

Concepts,” “The Magnitude of Cancer,” “The Causes of Cancer,”

“Cancers by Tissue of Origin,” and “Cancer Prevention and Control.”

Part I first describes research advances in understanding “the

biol-ogy of neoplasia,” including the progressive disruption of genetic and

epigenetic controls that regulate cell growth, division, and survival

(Chapter 2) Advances in high- throughput technologies have greatly

expanded the ability to identify germline and somatic mutations and

to relate these to etiology, prognosis, and treatment Tumor

classifi-cation is also changing for certain cancers, as data on the molecular

features and lineage of the neoplastic cells is combined with

infor-mation on the primary anatomic location and the morphologic,

histo-pathologic and clinical characteristics of the tumor (Chapter 3) The

“landscape” of genomic and epigenomic alterations in tumor tissue

has been cataloged for multiple human cancers (Chapter 4), revealing

both the singularity of individual cancer genomes and the

commonal-ity of genetic alterations that drive cancer in different tissues Chapter

5 describes advances in research on inherited genomic variants that

affect cancer risk Genome- wide association studies (GWAS) have

identified more than 700 germline loci associated with increased or

decreased risk for various types of cancer, although the risk estimates

for almost all are small to modest Innovations in genomics and other

“OMIC” technologies are identifying biomarkers that reflect internal

exposures, biological processes, and intermediate outcomes in large

population studies (Chapter 6) While research in many of these areas

is still in its infancy, mechanistic and molecular insights are extending

the traditional criteria for inferring causation in epidemiologic studies

of cancer (Chapter 7)

Part 2 of the book discusses the global public health impact of

can-cer and its relationship to demographic trends, changing risk factors,

socioeconomic disparities, and economic development It considers

the direct and indirect costs of cancer in the United States to illustrate

the economic burden in a high income country Parts 3– 5 of the book

discuss the growing list of exposures known to affect cancer risk, the

epidemiology of over 30 types of cancer by tissue of origin, and the

encouraging progress in cancer prevention and control Major

devel-opments in these areas are discussed below, beginning with those that

affect the public health impact of cancer

MAJOR NEW DEVELOPMENTS

Global Trends in Cancer Risk and Burden

Part II, “The Magnitude of Cancer,” provides a global public health

perspective on cancer The human and economic costs of cancer are

increasing worldwide (http:// globocan.iarc.fr) The World Health

Organization (WHO) estimates that 14 million new cases and 8.2

mil-lion deaths from cancer occurred in 2012 This burden is projected to

increase to 24 million cases and 13 million deaths annually by 2035

(Ferlay et al., 2013) Chapter 8 decribes the disproportionate increase

in the cancer burden in low- and middle- income countries (LMICs),

which can least afford additional health- related, social and financial

costs In 2012, these countries accounted for over half (57%) of all incident cancers; this is projected to increase to nearly two- thirds (65%) by 2035 Much of the increase will result from the growth and aging of populations, since LMICs currently comprise about 80% of the world’s population, and large numbers of young adults are now surviving to older ages, when cancer becomes more common In addition to the effect of demographic changes, cancer incidence and mortality rates are increasing in LMICs because of the widespread adoption of Western patterns of diet, physical inactivity, excess body fat, delayed reproduction, and tobacco smoking, especially of manu-factured cigarettes As countries advance economically, the incidence rates of cancers traditionally associated with Westernization (e.g., breast, colorectum, lung, and prostate) increase more rapidly than the decrease in cancers caused partly or wholly by infectious agents (e.g., stomach, liver, uterine cervix) Survival after a diagnosis of cancer is also lower in LMICs than in high- resource countries, because of later stage at diagnosis, a higher proportion of tumors diagnosed clinically rather than incidentally, and limited access to standard and state- of- the- art treatment protocols

In economically developed countries, the incidence rates of most

cancers are either stabilizing at a high level or decreasing, ing on the temporal trends of underlying risk factors and utilization of

depend-cancer screening Despite the decreasing rates, the disease burden, or

number of cancer cases and deaths, continues to increase The ing burden results from the aging and growth of populations, and the decline in competing causes of death from circulatory and infectious diseases Mortality rates are decreasing more rapidly than incidence rates for many cancer sites due to a combination of prevention, early detection, and improvements in treatment

increas-Part III, “The Causes of Cancer,” discusses 15 broad categories

of exposure that affect cancer risk These include exposures that are typically considered “environmental” by the public (chemical carcino-gens, ionizing radiation, occupational exposures, pollutants in air and drinking water), as well as exposures that are less widely recognized as carcinogenic (infectious agents, metabolic factors, body composition, reproductive and other hormones, pharmaceutical drugs, and immu-nological conditions) All of these exposures are “environmental” in the sense that they are acquired after conception rather than inherited Some are genotoxic and damage the structure of DNA or alter DNA repair; others modify gene expression, induce oxidative stress and/

or chronic inflammation, suppress host immunity, immortalize cells, modulate receptors, and/ or alter cell proliferation, cell death, or nutri-ent supply (Smith et al., 2016)

Although some exposures are conventionally perceived as style choices”, they are by no means entirely voluntary For example, behavioral risk factors such as tobacco smoking, energy imbal-ance, and physical inactivity are strongly influenced by factors in the social, economic, and cultural environment, beginning in early childhood Physiologic addiction is a major driver of tobacco use at all ages

“life-Part IV of the book describes “Cancers by Tissue of Origin” for

33 anatomic sites, multiple primary tumors, and cancers in children Rapid advances in discovering the molecular events that drive certain forms of cancer are transforming clinical diagnoses and treatment, and affecting tumor classification This will influence future endpoints in etiologic studies and population- based cancer surveillance

Trang 23

2 INTRODUCTION

Part V, “Cancer Prevention and Control,” discusses the impact of

interventions that effectively reduce carcinogenic exposures or disrupt

the multistage progression of tumors It focuses on interventions that

demonstrably reduce cancer risk in the general population, rather than

in special circumstances or high- risk subgroups Examples of these

are discussed in Chapters 61– 63 In all cases, the design and

imple-mentation of preventive measures require translational research to

ensure safety, optimize feasibility and impact, and critically evaluate

all stages of the process

Cancer Prevention and Control

A growing number of population- level preventive interventions are

proving to be highly effective, as confirmed by the decreases in

inci-dence as well as mortality rates from certain cancers (Chapters 61– 63)

Tobacco control has reduced the age- standardized incidence rate of

lung cancer by up to 40% among men in high- and middle- income

countries Increased screening for colorectal cancer and removal of

precursor lesions is credited for the 30% decrease in the incidence

rates at this site in the United States Universal neonatal vaccination

against hepatitis B virus (HBV) has markedly decreased the

preva-lence of chronic HBV infection and liver cancer at younger ages in

high- risk areas of East Asia and will yield maximal benefits against

cancer in the future The development of safe and effective vaccines

against human papillomavirus (HPV) and less expensive and less

oner-ous screening tests for cervical cancer have greatly expanded

opportu-nities to prevent HPV- related cancers among women in many LMICs

Increased funding is becoming available for application research and

cancer preventive services in LMICs Cancer prevention presents both

opportunities and challenges, as discussed in Part V of the text The

best practices developed for tobacco control provide an

encourag-ing model of how health- related policies can address the behavioral

causes of cancer However, these must be tailored to fit the particular

social, economic, and other considerations that affect the exposure

(Chapter 61)

Advances in Genomics and other OMICs

Technological advances in high- throughput genotyping/

sequenc-ing and gene expression arrays have transformed research on both

inherited (germline) susceptibility variants and the largely acquired

(somatic) mutations in tumor tissue Epidemiologic studies of cancer

genetics have focused mainly on germline variants associated with

cancer risk and etiology, whereas clinical and basic researchers have

characterized the landscape of somatic alterations in tumor cells that

drive the development and progression of cancer

Germline Susceptibility Variants

The tools to identify inherited genetic susceptibility variants have

advanced enormously since publication of the previous edition of

this text in 2006 At that time, studies involved either high- risk

fami-lies or the evaluation of a small number of pre- specified “candidate

genes” in case-control studies of sporadic cancers in the general

pop-ulation The candidate gene approach was largely unsuccessful in

identifying robust associations for several reasons, including small

sample size, limited statistical power, failure to account for

multi-ple testing (generating negative and false positive results,

respec-tively), and limited biologic knowledge to inform the selection of

candidate genes Following the completion of the Human Genome

Project in 2003, genome- wide maps of single nucleotide

polymor-phisms (SNPs) became available Advances in high- throughput

genotyping technology, combined with knowledge about the

struc-ture of genetic linkage disequilibrium, created opportunities to

con-duct exploratory (hypotheis- free or “agnostic”) surveys across the

entire genome Over the past decade, GWAS have robustly

identi-fied more than 700 common (i.e., minor allele frequency >5%)

sus-ceptibility loci associated with cancer risk, as discussed for specific

sites in Part IV, “Cancers by Tissue of Origin.” Because GWAS test

millions of alleles across the genome, they require stringent criteria

(“genome- wide significance”), large sample size, and replication

in more than one study to exclude chance associations Most of the associations identified through GWAS are modest (per allele ORs: 1.5– 2.0) or weak (ORs <1.5), but in aggregate these loci can distin-guish a wide range of risk in the population, thus providing oppor-tunities for targeted screening and prevention While our current knowledge regarding germline risk comes from studies in popula-tions of European ancestry, the identification of population- specific risk loci highlights the importance of conducting GWAS in diverse racial and ethnic populations

Statistical modeling suggests that, for many cancers, additional variants remain to be identified, yet the search for variants with smaller effect sizes, as well as less common variants, drives the need for even larger studies With the recent development of next- generation sequence technology, it is now practical to sequence whole exomes (coding regions plus regulatory regions) and whole genomes in population- and family- based studies in the search for heritability not identified through common variation in GWAS

An important limitation of GWAS is biological interpretation,

as the vast majority of risk variants revealed through GWAS are in non- coding genetic sequences Functional analyses are underway to address this issue The process is time- consuming, however, since

it incorporates new bioinformatics tools and a comparison of gene expression in tumor and normal tissue to localize the functional SNP and ultimately the affected gene

There has as yet been little progress in identifying interactions between inherited germline loci identified through GWAS and acquired

“environmental” risk factors Candidate gene studies have mented gene– environment interactions between tobacco smoking and the slow NAT- 2 acetylation phenotype for bladder cancer (Chapter 52)

docu-and between alcohol consumption docu-and slow ADH1B metabolizers for

esophageal cancer (Chapter 30) However, much larger GWAS with more precise measures of exposure and risk will be needed to assess other, subtler gene– environment interactions

Somatic Genomic Alterations

Most of the somatic genomic alterations, including mutations, indels, copy number alterations, and chromosomal rearrangements, that drive neoplastic progression in tumor tissue are acquired rather than inherited As mentioned, the Human Cancer Genome Project and other international laboratory and clinical collaborations have characterized so- called driver mutations (i.e., those which confer growth advantage to a mutated cell line) for multiple types of human cancer (Chapter 4) These mutations represent the events involved in the multistage development of particular forms of cancer (Armitage and Doll, 2004; Hornsby et al., 2007; Wu et al., 2016) It is notewor-thy that discoveries in somatic mutations over the past three decades provide strong support for the theory of multistage carcinogenesis that was proposed by Armitage and Doll, 10 years before elucidation

of the structure of DNA, and over 30 years before the identification

of the first proto- oncogenes and tumor suppression genes (Armitage and Doll, 1954) Sequencing studies have also implicated epigenetic modification as a major source of alterations in cancer (Chapters 2 and 4)

Variable combinations of genetic and epigenetic abnormalities account for the phenotypic heterogeneity within and among cancers Molecular characterization of tumors is increasingly used to predict prognosis and to guide the use of targeted therapies for individual can-cer patients These markers are only beginning to be evaluated and integrated into large- scale epidemiologic studies, yet they are already changing the taxonomy of some types of cancers and are likely to pro-foundly affect future etiologic studies (Chapter 3) There has been some progress in efforts to link specific classes of somatic muta-tions, such as the mutational signatures of ultraviolet (UV) radiation, tobacco smoke, and oncogenic viruses, to established carcinogenic exposures (Chapters 2 and 4) These molecular signatures may, in the future, identify the causal exposure(s) for cancer in individuals as well as populations Hopefully this goal will motivate interdisciplinary collaborations between epidemiologists, cancer prevention scientists, geneticists, cancer biologists, and clinicians

Trang 24

Introduction 3

Other OMICs

While genomic research is the poster child for the value of

agnos-tic, comprehensive explorations of germline variants associated with

cancer, other areas of OMIC research are moving toward this goal

(Chapter 6) The development of technologies to screen many

thou-sands of analytes related to gene expression (e.g., RNAseq),

epi-genetics (methylation; ChIP- Seq), metabolomics, and the microbiome

will open new opportunities to identify the connections between

expo-sures and the biologic effects that mediate carcinogenesis Various

OMIC technologies are at different stages of development One of the

more advanced efforts along these lines is the identification of

hor-mone metabolites that influence breast cancer risk, illustrating the

potential of these new technologies (Chapter 22)

OUTCOMES AND EXPOSURES

Changing Taxonomy of Cancer

Accurate and reproducible classification of neoplastic diseases is

essential for advances in diagnosis and treatment, for quantifying

geographic, temporal, and demographic variations in incidence, and

for identifying etiologic relationships and mechanisms Tumor

clas-sification has historically been based on the primary anatomic site and

morphology for most solid tumors and on histologic characteristics for

leukemias Classification systems have evolved to incorporate

infor-mation on morphology, genetics, cell lineage, developmental

char-acteristics, and an array of molecular, clinical, and etiologic factors

(ICD- O- 3, 2013) (Fritz et al., 2000)

While the refinement of cancer endpoints based on molecular or

other characteristics will potentially increase the ability of etiologic

studies to detect associations with specific tumor subtypes, it also

poses serious challenges Very large studies will be needed for both

discovery and replication Even well- established tumor markers are

not measured uniformly in all patients Newer classification systems

based on molecular features have generally been evaluated in only a

few hundred sporadic cancers, with little consideration of patient or

population characteristics Clonal heterogeneity within tumors and

changes in tumor pathology during treatment further complicate

clas-sification (Norum et al., 2014) While the use of automated algorithms

and computer- based image analysis is increasing among pathologists,

these methods and the assessment of reproducibility and validity may

not be reported to clinicians, epidemiologists, and others using the data

Population- based cancer registries are already challenged by efforts to

keep abreast of changing tumor classifications, especially when the

new criteria are not uniformly applied in a standardized manner

Exposures and Exposure Measurement

More than 100 different agents and exposures are now designated as

causally related to cancer in humans (Group I) by the International

Agency for Research on Cancer (IARC) Variations in the prevalence

and intensity of these exposures account for the striking geographic

and temporal variations in the occurrence of many types of cancer

While many exposures such as tobacco, alcohol, and numerous

indus-trial chemicals have long been classified as human carcinogens,

expo-sure patterns change, new agents are introduced, the ability to meaexpo-sure

exposures or outcomes progresses, and the quality, quantity, and/ or

specificity of evidence improves For example, the contining global

increase in obesity, metabolic syndrome, and type II diabetes,

com-bined with improvements in laboratory assays to measure hormones

in large population studies, has created new opportunities to study

metabolic and hormonal effects on cancer Thus, the discussion of

“Hormones and Cancer” (Chapter 22) has been expanded to consider

endogenous as well as exogenous exposures and peptide hormones

(insulin, insulin- like growth factors, growth hormone, leptin,

adipo-nectin, resistin, ghrelin, etc.) in addition to steroidal sex hormones

Several agents recently classified as Group 1 human carcinogens

affect massive numbers of people These include outdoor air pollution

(Chapter 17), the combustion of coal as household fuel (Chapters 16,

17, and 28), diesel engine exhaust (Chapter 17), the consumption

of red and processed meat (Chapter 19), and to a lesser extent, UV- emitting tanning beds (Chapter 14) The search for other modifiable causes of cancer continues New associations have been reported with shift work, sedentary behavior (as distinct from physical inactivity), computed tomography scans during childhood, sun- sensitizing phar-maceutical drugs, and others While the studies may be methodologi-cally strong, the evidence for causality is not yet considered definitive There is a continuing need to monitor both the immediate and long- term effects of more recently implemented medical technologies and newly developed drugs, products such as cellular phones and e- cigarettes, nuclear accidents, exposures occurring in war zones, and other exposures potentially related to cancer

Technological advancements in biomarker studies will allow more comprehensive examination of an individual’s metabolome, microbi-ome, genome, epigenome, and exposome The use of specific biomark-ers to assess internal exposures and identify children, adolescents, and other subsets of individuals who may be particularly susceptible to the factor being investigated (for example, dietary exposure, pesticide act-ing as a hormonal disruptor, or medication) could increase our ability

to detect complex exposure– disease relationships

ESTIMATES OF ATTRIBUTABLE FRACTION

Epidemiologists have long debated the fraction of cancer cases or deaths that could be avoided by preventive interventions (Chapter 61) Estimates of the percent of cancer deaths that could theoretically be avoided, if the exposures were eliminated, range from 50% to 80%, although the potential for primary prevention differs for incidence and mortality, by geographic region, gender, and attained age (Whiteman

and Wilson, 2016) About half of the deaths that could be avoided in

principle relate to 11 potentially avoidable risk factors, including the behavioral risk factors discussed above The attributable fraction esti-mates are predicated on the idea that cancer risk is largely acquired, rather than inherited Inherited factors do contribute to the variation in risk among individuals, but they cannot account for the large tempo-ral changes in risk within countries, or the differences in risk among migrants who move from one country to another

Chapter 19, “Diet and Nutrition,” provides the first estimates of the fraction of all cancers attributable to diet, in combination with or sepa-rate from overweight and physical inactivity The authors estimated the total as about 20% overall, which is weaker than previously esti-mated The proportion contributed by dietary composition indepen-dent of adiposity is less clear because some dietary factors have yet to

be identified or established with sufficient certainty, and associations are likely underestimated because of measurement error or misspecifi-cation of temporal relationships The authors estimate an etiologic contribution of 5%– 12% for dietary composition alone, but suggest that this could be appreciably higher when considering nutrient and genetic interactions

ONGOING CHALLENGES Tumor Diagnosis and Classification

In LMICs, the completeness and specificity of tumor diagnosis ies depending on economic resources and medical infrastructure Less than 10% of people in Africa and South America are covered

var-by population- based tumor registries (Chapter 8) In high- income countries, tumor classification is more advanced because of earlier application of diagnostic innovations and revised classification sys-tems Classification systems evolve with the introduction of new histo-chemical and molecular markers and advances in understanding tumor biology Even in high- income countries, there is wide variability in the proportion of tumors incompletely or inadequately characterized Molecular profiling at major cancer centers may include a range of established tumor markers, exome or whole genome sequencing, copy

Trang 25

4 INTRODUCTION

number, messenger and micro RNA sequencing, DNA methylation,

and proteomics analysis (Hoadley et al., 2014) While this information

may be useful clinically, it is not yet available for most cancer patients,

nor are the data routinely incorporated into cancer surveillance

sys-tems (Chapter 8) Even cancers that have undergone intensive

multi-disciplinary review to improve classification, such as hematopoietic

and lymphoproliferative malignancies, include subtypes characterized

as “not otherwise specified” or provisionally classified (Swerdlow

et al., 2016) Thus epidemiologic studies of cancer must collect and

archive tumor tissue in order to ensure a uniform, more complete, and

contemporary approach to molecular testing

Over- diagnosis

“Over- diagnosis” refers to the incidental detection of small and/ or

indolent cancers that otherwise might not cause clinical problems

during the patient’s lifetime (Chapter 8) Extreme examples of this

have been the sudden increase in prostate cancer diagnoses

follow-ing the introduction of PSA screenfollow-ing (Chapter 53), and the increase

in thyroid cancer diagnoses due to screening programs using

ultra-sound (Chapter 44) Overdiagnosis is most problematic if it leads to

unnecessary treatment and serious adverse effects The introduction

of new screening tests can also distort temporal trends in incidence

and bias observational studies of the impact of screening (Chapter 63)

The likelihood of overdiagnosis varies by cancer site and depends

on the baseline incidence and risk characteristics of a population

An estimated 10– 30% of newly diagnosed breast cancers identified

with screening mammography may reflect overdiagnosis (Bleyer &

Welch, 2012; Loeb et al., 2014; Vickers et al., 2014) In the absence of

molecular markers that reliably distinguish indolent from aggressive

tumors, clinicians must grapple with the potential for “overtreatment”

(Chapter 3) Over- diagnosis poses a greater clinical dilemma for early

stage cancers in internal organs than for premalignant lesions detected

by colorectal or cervical screening, because the treatment is more

invasive

Exposure Measurement Issues

Regardless of the epidemiologic study design used, it is challenging

to characterize accurately many types of acquired exposures due to

a lack of comprehensive measurements during the relevant exposure

window This presents a greater problem for some exposures than

oth-ers For example, as described in Chapters 19– 21, misclassification of

exposure is of great concern in characterizing patterns of nutrition and

physical activity, especially at earlier stages of life It presents less of a

problem in studying cigarette smoking, menopausal hormonal therapy,

or exposure to microbial agents, since these exposures can be

reason-ably well defined qualitatively, and biomarkers exist to supplement

questionnaire data Certain exposures are experienced in multiple

settings, including workplace, residential, recreational, medical, war

zone, and other settings Chemical exposures often occur as mixtures

in air or water Surrogate measures, such as job titles for occupational

exposures and administrative databases for residential exposures, do

not capture variations among individuals or over time

The timing of exposure is an area of special interest and challenge

Exposures that occur during a particular time window or a susceptible

stage of development may have adverse effects that are not evident

when exposure occurs later in life A classic example of this involved

in utero exposure to high doses of the hormone di- ethyl stilbesterol

(DES) in the daughters of mothers treated to prevent pregnancy

com-plications, who subsequently developed vaginal carcinoma in

adoles-cence (Chapter 49) For breast cancer, it is hypothesized that hormonal

exposures received in utero or immediately postnatally may affect

early stages in tumor development, or that breast tissue may be more

susceptible to certain exposures (e.g., ionizing radiation, cigarette

smoking, or alcohol consumption) during the period between

men-arche and first full- term pregnancy, when cells are proliferating but not

fully differentiated (Chapter 45) Similar hypotheses have been

pro-posed for nasopharyngeal cancer and Hodgkin lymphoma in relation

to early childhood infection with Epstein- Barr virus (Chapters 26 and 39) Although these hypotheses are important, they are difficult to test without biomarkers or experiments of nature that demarcate the timing

of exposure Serial acquisition of biological samples over many years

of follow- up would be informative, although this approach must be tempered by feasibility and cost considerations

FUTURE RESEARCH DIRECTIONS

While individual chapters outline future research directions for cific exposures or cancers, we highlight here selected cross- cutting issues that apply broadly to many cancers

spe-Team Science

One of the benefits of GWAS and pooled risk factor studies has been the formation of multi- institutional international consortia to share biospecimens and primary data in order to maximize statistical power for discovery and robust replication (Boffetta et al., 2007) These con-sortia have been instrumental in creating new models of funding, lead-ership, and authorship, and in sharing and harmonizing primary data across studies The formation of larger and more complex data sets has stimulated innovations in informatics and the development and appli-cation of novel analytic methods Further advances in high- throughput laboratory technology will continue to create new opportunities to explore the complex biology of genomics, metabolomics, proteomics, and so on, in large population studies This will generate vast amounts

of data to be analyzed It will also require insights from diverse plines to plan analyses and to interpret the results

disci-“Transdisciplinary research” signifies a level of collaboration in which researchers from different scientific disciplines come together

to identify the most important research questions, design the optimal approach, and collect, analyze, and publish the study results jointly (Rosenfield, 1992) This level of team science transcends disciplinary boundaries and benefits all parties For example, the application of haplotype analyses based on the structure of genetic linkage disequi-librium, initially proposed by geneticists, greatly accelerated explor-atory analyses across the entire genome in large population studies Similarly, the involvement of epidemiologists in tumor genomics has brought a population science perspective to this field, increasing atten-tion to population sampling, sex, and racial/ ethnic differences, and exposures such as smoking that can affect the mutational spectrum

of various cancers There are many opportunities for collaborations involving laboratory scientists, analytic chemists, epidemiologists, and others to accelerate research on etiologic issues related to can-cer One example would be transdisciplinary research to understand hormonal carcinogenesis at the molecular level in human populations (Chapter 22)

Cancer Survivorship

The number of people surviving after a diagnosis of cancer has increased rapidly in high- income and many middle- income countries due to improvements in treatment and the effects of screening on both early diagnosis and more complete ascertainment In the United States, the estimated number of people alive for at least 5 years after a diagno-sis of cancer increased from 4 million in 1978 (~1.8% of the popula-tion) to 13.7 million in 2012 (~4% of the population), and is predicted

to approach 18 million by 2022 (de Moor et al., 2013; Harrop et al., 2011) A substantial proportion of these are long- term survivors Forty percent of individuals who survived for at least 5 years were alive

10 or more years after diagnosis, and 15% had survived 20 or more years (Howlader et al., 2015) To address the rapidly increasing pop-ulation of cancer survivors, the NCI Office of Cancer Survivorship and a 2006 publication from the Institute of Medicine (Committee

on Cancer Survivorship, 2006) provided a framework for identifying and addressing the unmet needs of this growing population Disease and treatment affect multiple health domains (e.g., medical, physical

Trang 26

Introduction 5

function, psychiatric and psychosocial, cognitive, work, sexual, and

reproductive) These factors and heath- related quality of life should

be assessed at discrete intervals following diagnosis Studies should

examine whether quality of life and survivorship issues differ for

can-cers in children from those in later life Research on survivorship is

largely still in its infancy (de Moor et al., 2013; Harrop et al., 2011;

Richardson et al., 2011) The occurrence of multiple primary and

sec-ondary cancers following treatment is an important area of research

(Chapter 60) Cancer epidemiologists should have a major research

role in cancer survivorship because of their expertise in observational

study designs and exposure assessment

Risk Prediction Models

Risk prediction models estimate the absolute risk of being diagnosed

with or dying from a specific condition during a defined time period

for individuals with defined demographic characteristics and risk

fac-tors These models are widely used to predict individual risk and to

guide treatment decisions in cardiovascular medicine Unfortunately,

the models currently available for cancer provide reasonably

accu-rate predictions for groups of people but not for individuals (Amir

et al., 2010)

Site- specific chapters discuss research on risk prediction models

for cancers of the lung (Chapter 28), colorectum (Chapter 36), breast

(Chapter 45), and multiple primary cancers (Chapter 60) Brinton

and colleagues in Chapter 45 review the research on factors that may

improve the discriminatory accuracy of existing models of breast

can-cer These include endogenous hormone levels, proliferative benign

biopsy diagnoses, behavioral risk factors, mammographic density, and

genetic polymorphisms Although the associations with genetic

poly-morphisms identified by GWAS are modest (per allele ORs: 1.5– 2.0)

or weak (ORs <1.5), as discussed earlier, it is hoped that in the

aggre-gate the addition of these loci to the model can more accurately predict

risk for individuals, thereby improving targeted approaches for

screen-ing and prophylactic treatment (Garcia- Closas et al., 2014)

Risk prediction models of colorectal cancer have been used to

esti-mate the independent and combined effects of behavioral, medical,

familial and genetic risk factors on attributable risks One such study

estimated that a population with the optimal distribution of established

modifiable risk factors for colorectal cancer would have a 43% percent

lower (95% CI: 0.14, 0.65) incidence of disease (Lee et al., 2016)

Meester et al (2015) estimated the effect of screening on deaths from

colorectal cancer and concluded that the nonuse of screening methods

among people aged ≥ 50 years in the United States accounted for more

than 50% of colorectal cancer deaths

The Microbiome

Remarkable advances have been achieved in identifying microbial

pathogens that influence cancer risk (Chapter 24) This research has

focused historically on disease- causing organisms, rather than on the

trillions of “commensal” organisms that live in or on the human body

Increased interest in the human microbiome has broadened the scope

of inquiry The greatest concentration of commensal

microorgan-isms is in the upper and lower gastrointestinal tract, although there

are well- characterized symbiotic communities in the skin, mouth,

sinonasal cavities, and vagina Diverse species of bacteria, archaea,

viruses (including bacteriophages), and fungi can be identified Some

species have coevolved with humans for millennia and maintain

essen-tial physiologic enzymatic functions, such as the synthesis of essenessen-tial

vitamins and the metabolism of carbohydrates, bile acids, and

xeno-biotics Commensal flora may have detrimental as well as beneficial

effects on cancer The colonic microbiome is thought to account for

the approximately 40- fold higher incidence of adenocarcinoma in the

colon than in the small intestine (Chapter 35) While the small

intes-tine is not sterile, it has comparatively few commensal organisms The

difference in cancer risk between the two sites is remarkable, given

that both organs undergo rapid cell replication The small intestine

comprises about 90% of the mucosal absorptive surface area of the

gastrointestinal system, yet in the United States accounts for only 3.2% of digestive system cancer cases (Howlader et al., 2015).The tools for studying complex microbial communities are only now becoming available There is great interest, however, in characterizing alterations in the microbiome caused by changes in diet, antibiotic use, and other exposures, and in the inflammatory responses that result from the influx of pathogenic organisms into organ- specific microbi-omes (Dale and Moran, 2006) Future epidemiologic and experimental research on the human microbiome will likely continue to focus on the carcinogenic effects of disturbed homeostasis and the disruption

of the normal diversity of the microbial flora (Bultman, 2016), and

as well as on potential role(s) of the microbiome on the pathogenesis

of obesity, diabetes (Okeke et al., 2014), and autoimmune diseases (Hooper et al., 2012)

Development of New Cohorts

There is an ongoing need to develop new cohorts to complement ing cohorts, with a particular focus on collecting and banking biologi-cal samples needed for broader OMICs studies (e.g., tumor tissue, blood, urine, stool), the inclusion of state- of- the- art exposure mea-surements and data collection strategies, data on new and emerging exposures, and repeated assessments over time to capture important changes in exposures, behaviors, or physiological factors Besides enrolling participants from more recent birth cohorts, there is also a critical need to increase the racial/ ethnic and socioeconomic diversity

exist-of participants Careful consideration must be given to cost- effective data sources and methods of data collection to address these needs in the long- term follow- up of future cohorts Data access and data shar-ing also need to be considered since data collected from more recent cohorts (such as the UK Biobank, the NCI Cohort Consortium, and the Childhood Cancer Survivor Study) are increasingly made broadly available as a resource for qualified investigators Finally, new models

of governance and participant engagement and interaction will need to

be developed for the next generation of cohort studies

New and Non- Traditional Data Sources

A variety of new data sources are enriching the opportunities for descriptive and analytic epidemiologic studies The expanding use of electronic health records, along with a movement toward interoperabil-ity and standardized content, can potentially provide individual- level medical data on tumor characteristics, comorbidity, and treatment Data linkage between population- based cancer registries, vital status and mortality records, hospital and other medical records, and admin-istrative data sets provides an important resource for studies of cancer survivorship Personal devices such as accelerometers or smart phones

to track movement can improve the measurement of physical activity

In parallel, many other types of digital data are increasingly able Potentially linkable data from diverse fields of science (e.g., OMICs; environmental and geographic sciences) and areas outside

avail-of science (e.g., business, finance, telecommunications, social media, and the Internet of things) are being promoted under the umbrella of

“Big Data.” Non- traditional data sources have been touted as a olutionary development in the future of epidemiology (Khoury and Ioannidis, 2014; Mooney et al., 2015) It should be recognized that the use of non- traditional data sources involves much more than working with large volumes of data These sources have been characterized as high variety (secondary use from many diverse sources), high volume (both in numbers of observations and/ or variables per observation) and high velocity (real time) (Douglas, 2012) It is important that epide-miologists collaborate with informaticians and computer scientists

rev-to explore the promise of Big Data Expertise in observational study design, along with efforts to assess validity and reproducibility, will

be essential to avoid “big error” driven by chance, bias (e.g., selection bias, measurement error, confounding) and lack of external validity Cancer epidemiologists bring the subject matter expertise needed to frame and interpret Big Data results for clinical and public health rel-evance While the long- term impact of Big Data on cancer surveillance

Trang 27

6 INTRODUCTION

and analytical epidemiology is not yet clear, we anticipate that this

may have a major impact on the field over the next decade

References

Amir E, Freedman OC, Seruga B, and Evans DG 2010 Assessing women at

high risk of breast cancer: a review of risk assessment models J Natl

Cancer Inst, 102(10), 680– 691 PMID: 20427433

Armitage P, and Doll R 1954 The age distribution of cancer and a multi-

stage theory of carcinogenesis Br J Cancer, 8(1), 1– 12 PMCID:

PMC2007940

Bleyer A, and Welch HG 2012 Effect of three decades of screening

mammog-raphy on breast- cancer incidence N Engl J Med, 367(21), 1998– 2005.

Boffetta P, Armstrong B, Linet M, et al 2007 Consortia in cancer

epidemiol-ogy: lessons from Interlymph Cancer Epidemiol Biomarkers Prev, 16(2),

197– 199

Bultman SJ 2016 The microbiome and its potential as a cancer preventive

intervention Semin Oncol, 43(1), 97– 106 PMCID: PMC4789109.

Committee on Cancer Survivorship 2006 Improving Care and Quality

of Life, National Cancer Policy Board From Cancer Patient to

Cancer Survivor: Lost in Transition Washington, DC: National

Academies Press

Dale C, and Moran NA 2006 Molecular interactions between bacterial

symbi-onts and their hosts Cell, 126(3), 453– 465.

De Moor JS, Mariotto AB, Parry C, et al 2013 Cancer survivors in the

United States: prevalence across the survivorship trajectory and

impli-cations for care Cancer Epidemiol Biomarkers Prev, 22(4), 561– 570

PMCID: PMC3654837

Douglas L 2012 The Importance of Big Data: A Definition Stamford,

CT: Gartner

Ferlay J, Soerjomataram I, Ervik M, et al 2013 GLOBOCAN 2012 V1.0,

Cancer Incidence and Mortality Worldwide: IARC Cancerbase No 11

[Internet] Available from: http:// globocan.iarc.fr

Fritz A, Percy C, Jack A, et al 2000 International Classification of Diseases

for Oncology, 3rd ed Geneva: World Health Organization

Garcia- Closas M, Gunsoy NB, and Chatterjee N 2014 Combined associations

of genetic and environmental risk factors: implications for prevention of

breast cancer J Natl Cancer Inst, 106(11) PMCID: PMC4271030.

Harrop JP, Dean JA, and Paskett ED 2011 Cancer survivorship research: a

review of the literature and summary of current NCI- designated cancer

center projects Cancer Epidemiol Biomarkers Prev, 20(10), 2042– 2047

PMCID: PMC3191883

Hoadley KA, Yau C, Wolf DM, et al 2014 Multiplatform analysis of 12 cancer

types reveals molecular classification within and across tissues of origin

Cell, 158(4), 929– 944 PMCID: PMC4152462

Hooper LV, Littman DR, and Macpherson AJ 2012 Interactions between the

microbiota and the immune system Science, 336(6086), 1268– 1273

PMCID: PMC4420145

Hornsby C, Page KM, and Tomlinson IP 2007 What can we learn from the

population incidence of cancer? Armitage and Doll revisited Lancet

Oncol, 8(11), 1030– 1038

Howlader N, Noone AM, Krapcho M, et al 2015 SEER Cancer Statistics Review, 1975– 2012, National Cancer Institute Bethesda, MD, http:// seer.cancer.gov/ csr/ 1975_ 2012/ , based on November 2014 SEER data submis-sion, posted to the SEER Web site, April 2015

Khoury MJ, and Ioannidis JP 2014 Medicine: Big Data meets public health

Science, 346(6213), 1054– 1055 PMCID: PMC4684636

Lee DH, Keum N, and Giovannucci EL 2016 Colorectal cancer epidemiology

in the Nurses’ Health Study Am J Public Health, 106(9), 1599– 1607

PMCID: PMC4981802

Loeb S, Bjurlin MA, Nicholson J, et al 2014 Overdiagnosis and

overtreatment of prostate cancer Eur Urol, 65(6), 1046– 1055 PMCID: PMC4113338

Meester RG, Doubeni Ca, Lansdorp- Vogelaar I, et al 2015 Colorectal

can-cer deaths attributable to nonuse of screening in the United States Ann

Epidemiol, 25(3), 208– 213 E201 PMCID: PMC4554530

Mooney SJ, Westreich DJ, and El- Sayed AM 2015 Commentary:

epi-demiology in the era of big data Epiepi-demiology, 26(3), 390– 394 PMCID: PMC4385465

Norum JH, Andersen K, and Sorlie T 2014 Lessons learned from the intrinsic

subtypes of breast cancer in the quest for precision therapy Br J Surg,

101(8), 925– 938

Okeke F, Roland BC, and Mullin GE 2014 The role of the gut microbiome in

the pathogenesis and treatment of obesity Glob Adv Health Med, 3(3),

Rosenfield PL 1992 The potential of transdisciplinary research for sustaining

and extending linkages between the health and social sciences Soc Sci

Med, 35(11), 1343– 1357

Schottenfeld D, and Fraumeni JF 1982 Cancer Epidemiology and Prevention

1st ed Philadelphia: Saunders

2nd ed Oxford: Oxford University Press

3rd ed Oxford: Oxford University Press

Smith MT, Guyton KZ, Gibbons CF, et al 2016 Key characteristics of cinogens as a basis for organizing data on mechanisms of carcinogenesis

car-Environ Health Perspect, 124(6), 713– 721 PMCID: PMC4892922.Swerdlow SH, Campo E, Pileri SA, et al 2016 The 2016 revision of the

World Health Organization classification of lymphoid neoplasms Blood,

127(20), 2375– 2390 PMCID: PMC4874220

Vickers AJ, Sjoberg DD, Ulmert D, et al 2014 Empirical estimates of prostate

cancer overdiagnosis by age and prostate- specific antigen BMC Med, 12,

26 PMCID: PMC3922189

Whiteman DC, and Wilson LF 2016 The fractions of cancer attributable to

modifiable factors: a global review Cancer Epidemiol, 44, 203– 221.

Wu S, Powers S, Zhu W, and Hannun YA 2016 Substantial contribution of

extrinsic risk factors to cancer development Nature, 529(7584), 43– 47

PMCID: PMC4836858

Trang 28

BASIC CONCEPTS

Trang 30

MICHAEL DEAN AND KAROBI MOITRA

OVERVIEW

The biology of tumors is being understood in unprecedented detail in

terms of the development, growth, survival, and spread of cancer cells

in the body, and their interaction with the host Analyses of familial

cancer syndromes and in vitro and animal assays have revealed the

major tumor suppressor genes and oncogenes that regulate cell growth

and survival More recent scans of large numbers of cancer cases and

controls with single nucleotide polymorphism (SNP) arrays, known as

genome- wide association studies (GWAS), have identified additional

genetic loci, mostly in regulatory regions that influence cancer risk

The genomic characterization of tumors through exome and whole

genome sequencing, transcriptome, and DNA methylation analyses

are identifying further complexity in the somatic alterations present

in tumor cells A new field of tumor signature analysis seeks to

cor-relate exposures (tobacco, carcinogens, radiation) to specific classes

of mutations in the tumor genome Alterations in the DNA repair

pro-cesses of the cell also can be shown to leave a signature in the point

mutations and structural alterations found in cancer A major new class

of somatically altered genes are those encoding proteins that modify

histones and other components of chromatin and/ or alter the

methyla-tion of DNA Such epigenetic alteramethyla-tions appear to be central to late

events in cancer, such as metastasis and therapy resistance, and further

understanding of these alterations will be the key to effective therapy

Single- cell analyses are identifying complex patterns of cancer cell

heterogeneity and evolution and have supported the concept of the

tumor cell as a multicellular entity Finally, the cancer cell arises from

the body’s normal cells, and interacts throughout the entire lifetime of

the tumor with surrounding host cells, blood supply, and the immune

system Inflammation is both a major cause and consequence of

can-cer, and increasingly modulators of the immune system are being used

in cancer therapy

INTRODUCTION

The Odyssey of a Tumor

In the epic poem The Odyssey, Homer details the long, adventurous

journey of Odysseus as he returns home from the Trojan Wars Over

the course of this decade- long, arduous journey, Odysseus must use

all his talents, physical strength, prowess as a soldier, and intelligence

to survive At times the gods are against him, and at other times they

actually come to his aid In the end, he is the only surviving member

of his army

There is a similar long and difficult path of a tumor from a single

abnormal cell to a mature and often lethal growth (Figure 2–1) In the

odyssey of the tumor, though, our own cells are the villains of the epic,

and it is science and medicine (rather than the gods) that are called

upon to wage the epic battle against our own cells The initial tumor

can arise at virtually any time of our life, from just after birth until old

age, and in fact cancer risk continues to increase at least until 80 years

of age (Harding et al., 2008; Pedersen et al., 2016)

Tumors can arise from virtually any cell type or tissue in the body

Despite this diversity, there are certain traits in common with tumors in

different individuals and from different tissues We have identified the

major genetic and external risk factors for most cancers, the principal genes that are altered somatically in tumors, and many of the mecha-nisms While external factors, like mutagens and infectious agents, initi-ate cancer, in the later stages of tumor development we find the tumor cell tapping into diverse tools and pathways that are the basis of life itself

Definition of Cancer

• “Cancer” refers to a large, heterogeneous group of diseases with common underlying pathology characterized by uncontrolled cel-

lular growth and division

• Cancer is inherently a genetic disease at the cellular level.

• Genetic and epigenetic changes accumulate in localized (somatic)

tissue and dysregulate genetic control over basic cellular functions.Cancer involves all of the organ systems and cell types of the body (although some tissues are more susceptible than others) and is there-fore inherently more complicated than other major disease such as car-diovascular or neurological disease

Cancer is a genetic disease in the sense that the tumor cell has suffered from multiple lesions in its DNA, nearly always involving multiple gene mutations, chromosome rearrangements, and extensive alteration of the epigenome, through changes in DNA methylation and histone modification (Figure 2–2) Depending on the tumor type, up to 10% of cancers are caused by inherited germline mutations that are moderately to highly penetrant Common germline variants raising or lowering cancer risk 0.1– 4- fold have also been described for nearly every tumor type

Cancers begin as localized growths or premalignant lesions In a number of cancer types, we can see these growths in the form of colon polyps, cervical dysplasia, enlarged moles, and ductal carcinoma in situ, among others Many of these early lesions will not progress or can regress or disappear spontaneously Early colon cancer lesions are among the best studied of these growths Nearly all colon polyps have been observed to have inactivating mutations in the APC tumor sup-pressor gene (Kinzler and Vogelstein, 1996) The APC protein func-tions as an antagonist of WNT signaling, and was first identified in familial adenomatous polyposis (FAP), an autosomal dominant dis-ease characterized by a large number of colon polyps and a high risk for colon cancer

A very critical transition is the evolution or progression of the premalignant lesion into a local, malignant tumor This stage involves the accumulation of yet more genetic insults Few local malignant lesions can be eliminated by the body The final transition

is to an invasive and/ or metastatic tumor that is capable of ing adjacent tissues or spreading across the body The vast majority

invad-of cancer morbidity and mortality is due to cancers in this third, metastatic stage

Inherited Cancer Syndromes

Family history has been noted as a risk factor since the early days

of medicine, but the major inherited cancer syndromes began to be described in the 1900s For example, the von Hippel Lindau syndrome was independently described by von Hippel and Landau in 1904 and

1927 and involves tumors of the kidney and pheochromocytomas inherited in an autosomal dominant fashion

Trang 31

10 PART I: BASIC CONCEPTS

The Role of Tumor Suppressor Genes

A major conceptual breakthrough came in 1975 when Al Knudson

pub-lished an analysis of retinoblastoma and correctly predicted that

indi-viduals inherited a defective copy of a gene (that we now call a tumor

suppressor gene) and that the tumor cells suffer a somatic loss or loss

of function of the normal copy (Knudson, 1971) Therefore, individuals

affected by these syndromes often display multiple tumors such as

bilat-eral breast cancer or retinoblastoma, multiple kidney or colon tumors

(Knudson, 2002) Although the inheritance pattern is dominant, at the

cellular level, both copies of the gene must be inactivated or disabled

The same tumor suppressor gene can also suffer mutation or loss of

both alleles in a person without an inherited defect (spontaneous

muta-tion) Therefore, the genes identified from inherited cancer syndromes

are almost always genes that are mutated in sporadic cancers (Table 2–1) Most of the genes causing these inherited syndromes have turned out to play critical roles in the cell cycle or growth control, and therefore their loss imparts a lack of control of the cell- division process and/ or loss of recognition of damage and checkpoints to cell cycle progression Calling them tumor suppressor genes explains the role of these genes in cancer formation, but not their role in normal growth and development.Genes involved in DNA repair constitute another class of genes that cause inherited cancer Again, they are mutated in the germline, but the tumor has a further mutation or loss of the normal allele (the sec-ond hit) These tumors contain lesions in critical components of DNA repair and therefore can accumulate lesions that can lead to cancer

A critical conclusion made by Knudson was that the same gene that causes an inherited cancer syndrome can suffer somatic mutation to both copies and contribute to sporadic cancer He demonstrated math-ematically that this can explain the tendency for inherited retinoblasto-mas to occur bilaterally and at an earlier age of onset, whereas tumors

in patients without a family history are more likely to be unilateral and appear at a later age This model has held true for many of these genes,

and APC (colon cancer), RB1 (retinoblastoma), VHL (renal tumors),

very frequently somatically mutated (Hahn et al., 1996; Kinzler et al., 1991; Latif et al., 1993; Murphree and Benedict, 1984; Varley, 2003) Several of these genes have been termed “gatekeepers,” as they are often the first gene mutated in the tumor and/ or are inactivated in close

to 100% of specific tumor types (Kinzler and Vogelstein, 1997) This

is consistent with the critical role that the gatekeeper genes play in the formation of the premalignant lesion, the necessary precursor to

the local malignancy For example, patients with germline APC

muta-tions develop hundreds of colon polyps, and virtually all sporadic

colon polyps display somatic mutation of the APC gene (Kinzler and

Vogelstein, 1996)

There are exceptions to this paradigm, mostly in the DNA repair

genes Germline mutations in the BRCA1 and BRCA2 genes confer

high penetrance for breast cancer, moderate penetrance for ian cancer, and risk for prostate, pancreas, and other tumors In the tumors of such germline- mutated patients, there is virtually always inactivation of the normal allele, consistent with the two- hit model

ovar-Normal Cell

Germline Mutant Cell

Genetic Susceptible Cell

Pre-Malignant Lesion

Locally Invasive Lesion

Therapy Resistant Tumor

MetastaticTumor

GG

X

Point Insertion/

Deletion (indel)

Gene Translocation Deletion/Gain Gene

(CNA)

Figure 2–2 Genetic lesions in cancer cells (a) Some individuals have

germline mutations (rare, highly penetrant mutations) or susceptibility

alleles (b) Point mutations result in the replacement of a single

nucleo-tide, a single-nucleotide variant (SNV) These can cause the replacement

of a single amino acid in a protein (non-synonymous variant), create a new

stop codon (nonsense mutation) or alter a splice site or regulatory element

(c) Insertions or deletions (indels) of one or more nucleotide in the coding

region of a gene can result in a shift in the reading frame and an inactive

protein (d) Chromosome translocations result in the formation of a fusion

protein containing portions of two previously distinct genes (e) An entire

gene can be deleted or amplified Copy number alterations (CNA)

Trang 32

Biology of Neoplasia 11

(Couch et al., 2014; Miki et al., 1994; Moran et al., 2012; Struewing

et al., 1997; Wooster et al., 1995) However, BRCA1 and BRCA2 are

not gatekeeper genes for breast and ovarian cancer, as they are rarely

somatically mutated in breast or ovarian tumors in non- hereditary

cases (Cancer Genome Atlas Research Network, 2011, 2012)

At the molecular level, both inherited and somatic mutations of

tumor suppressor genes serve to inactivate the protein They are

fre-quently frameshift or nonsense mutations generating a truncated

pro-tein, and are less likely to be point mutations causing the change of a

single critical amino acid

Somatic Mutations and Oncogenes

mutated in cancer These are genes that are activated by mutation,

cre-ating a molecule with a dominant function in the cell (Croce, 2008;

Dean, 1997) These genes are almost never found to be mutated in the

germline, and are often lethal when introduced in the activated form

in the germline of a mouse Most oncogenes were discovered as the

cellular homologues of viral oncogenes (Stehelin et al., 1976; Watson

et al., 1982), or by cellular assays in which DNA from a tumor cell was

introduced into immortal cells in culture and a morphological change

was noted, or the cells became tumorigenic in the mouse (Der et al.,

1982; Pulciani et al., 1982)

Most oncogenes have a direct role in cell growth and can be growth

factors (PDGFA, KITL), growth factor receptors (MET, EGFR,

PDGFR), intracellular growth signaling molecules (RAS, PIK3CA,

RAF, MAPK1), or transcription factors mediating growth signals

(MYC, FOS) (Croce, 2008; Dean, 1997; Dean et al., 1985; Weinberg,

1996) The mutations in these genes tend to be either specific

muta-tions that activate the protein (G12E in RAS, H1047Y in PIK3CA)

or gene amplifications creating an overabundance of the protein The

mutations activating oncogenes are among the most common somatic

events in multiple cancer types

Oncogenes can also be activated by chromosomal translocations

that generate fusion proteins In the example of the Philadelphia

chro-mosome, the t(9;21) event joins the BCR protein on chromosome 9 to

the ABL oncogene on chromosome 21 (de Klein et al., 1982; Rowley,

1973) The hybrid protein contains a portion of BCR at the N nus and the kinase portion of ABL at the C terminus This leads to the production of an activated ABL tyrosine kinase constitutively signal-ing and stimulating cell division (Clark et al., 1988) Certain cancer cell types are known to be dependent on this oncogene signal for sustained cell growth and are known as “oncogene addicted.” In the case of the Philadelphia translocation, ABL tyrosine kinase inhibitors can inhibit CML cells and, barring the appearance of drug resistance, can perma-nently keep the tumor cells suppressed (Druker et al., 2001a, 2001b)

termi-Oncogenic Driver Genes

The advent of large- scale exome and whole genome sequencing of tumors has led to an explosion of information on the extent of somatic mutations in tumors (Vogelstein et al., 2013) There is a high variation

in the number of somatic mutations per tumor, with pediatric tumors and most hematopoietic tumors showing low mutation load, and solid tumors higher mutation loads (Kandoth et al., 2013) Critical to this work

is the elucidation of the functional mutations in a given tumor and the separation of these mutations (driver mutations) from random mutations accumulating in the tumor that may have little or no effect on the tumor-igenesis process (passenger mutations) Vogelstein et al developed an operational definition to define oncogenes and tumor suppressor genes based on the frequency of certain types of mutations (Lawrence et al., 2013; Vogelstein et al., 2013) But mathematical calculations have also been performed to determine if a gene is significantly mutated in a given tumor type based on the size of the gene (Lawrence et al., 2013) For most of the well- known oncogenes and tumor suppressor genes, there is experimental evidence in cell culture or model systems to demonstrate that they drive cell growth or tumor formation at the cellular level But genome sequencing of tumors has identified many frequently mutated genes that have not been experimentally validated

Chromatin Remodeling Genes

The genes most frequently mutated in cancers that were not discovered

as oncogenes or tumor suppressors are the class of genes involved in the modification of histone molecules found in chromatin, or proteins

Table 2–1 Major Inherited Cancer Genes

Breast, ovarian, pancreatic, prostate Breast- ovarian cancer syndrome BRCA1, BRCA2

Colorectal Hereditary non- polyposis colorectal

cancer/ Lynch syndrome

MLH1, MSH2, MHS6, PMS2

Melanoma Familial atypical multiple mole- melanoma

Pancreatic Hereditary pancreatitis/ familial pancreatitis PRSS1, SPINK1

Leukemias, breast, brain and soft tissues cancer

Pancreatic cancers, pituitary adenomas, benign skin and fat tumors Multiple endocrine neoplasia 1 MEN1Skin and brain Nevoid basal cell carcinoma syndrome PTCH1

Thyroid, pheochromocytoma Multiple endocrine neoplasia 2 RET, NTRK1

Pancreatic, liver, lung, breast, ovarian, uterine, and testicular

Peutz- Jeghers syndrome STK11 / LKB1

Tumors of the spinal cord, cerebellum, retina, adrenals, and kidneys Von Hippel- Lindau syndrome VHL

Source: Adapted from: http://dceg.cancer.gov/research/what-we-study/gene-host/hereditary-cancer-syndrome and the AACR Cancer Progress Report, 2015.

Trang 33

Table 2–2 Chromatin Remodeling Genes

Gene Symbol RefSeq ID Chromosome Location

directly involved in the position and remodeling of nucleosomes on

DNA The role of nearly all of these genes as frequently mutated genes

was discovered during exome or whole genome sequencing of large

numbers of specific tumors (Kishimoto et al., 2005; Muraoka et al.,

1996; Santos- Rosa and Caldas, 2005) For example, the exome

sequenc-ing of bladder tumors led to the identification of KDM6A, a previously

obscure enzyme functioning as a lysine methylase of histones, as one of

the most frequently mutated genes in this tumor type (Gui et al., 2011;

Guo et al., 2013) In fact, four of the five mutated genes in bladder

can-cer come from this class And for nearly every solid tumor, one or more

chromatin remodeling genes are commonly inactivated (Table 2–2)

Epigenetic and Regulatory Changes in Cancer

At the molecular level, most of these chromatin remodeling genes

have the same spectrum of mutations as the tumor suppressor genes,

frameshift, and nonsense mutations; and both alleles are inactivated

in the tumor cell However, they are not found as germline mutations

causing cancer, but they often cause developmental disorders (Santos-

Rosa and Caldas, 2005)

Therefore, tumor genome sequencing has uncovered a major new

class of genes (epigenetic modifiers) that is commonly mutated in

can-cer In fact, in nearly every tumor type, there is at least one gene from

this class that is frequently mutated While in a few cases, genes from

this class have been shown to accelerate cell growth, invasion, and

tumor growth in model systems, in the majority of these genes, we

do not have a molecular mechanism explaining their function in the

cancer process One exciting idea is that the tumor cell may need to

undergo a wholesale change in gene expression and regulation and that

the loss of critical chromatin regulatory genes allows the tumor cell to

be more plastic and to turn on programs of gene expression that enable

the more complex phenotypes of invasion and metastasis

Except for rare embryonic mutations and certain pediatric

tumors, cancers develop over decades and involve the accumulation

Table 2–2 Continued

Trang 34

of genetic and epigenetic alterations in interplay with the immune

system

HALLMARKS OF CANCER CELLS

From the careful dissection of the properties of many tumor types,

Weinberg and Hanahan, in two review articles over a decade apart,

described the most common features, or hallmarks, of the tumor cell

(Hanahan and Weinberg, 2000, 2011) (Figures 2–3a and 2–3b) These

are common characteristics of most tumor types At the individual

cell level, the most important hallmarks are self- renewal, uncontrolled

growth, absence of growth regulation, loss of DNA repair, and loss of

programmed cell death in the face of these lesions These hallmarks

are especially important at the initial stages of carcinogenesis in the

formation of the premalignant and later malignant lesion

At later stages, a tumor acts as a multicellular entity that interacts

with multiple normal cells of the immune system, stimulates the

pro-duction of new blood vessels (angiogenesis), and develops the ability

to invade other tissues and metastasize

Cancer Stem Cell Model

The most important hallmark of cancer cells is the continual growth

of the tumor, or self- renewal In the body’s normal stem cells, self-

renewal is a mechanism to generate cells identical to the stem cell and

maintain a population of pluripotent, multipotent, or oligopotent cells

One model of cancer places a major emphasis on the ability of cancer

cells to undergo self- renewing replication like that which occurs in

stem cells (Al- Hajj et al., 2004; Dean et al., 2005) In certain tumor

types and tumor models, cancer stem cells or tumor initiating cells

have been isolated (Al- Hajj et al., 2003)

One pathway to tumor development starts with a chronically

rep-licating normal stem cell These cells naturally possess several of the

“hallmarks of cancer.” Sustained or chronic tissue damage and/ or

inflammation can lead to the activation of stem cells and their

con-tinuous replication Such a population of replicating stem cells can

become a target for somatic mutations that cause sustained

prolifera-tive signaling and/ or evasion of growth suppression

This leads to a premalignant cancer stem cell, a target for further

mutation and uncontrolled proliferation and loss of growth control

Early mutations can also promote defective DNA repair, and the loss

of checkpoints in the cell cycle preventing cells with damaged DNA

or aneuploid chromosomes from replicating Genome instability and

defective DNA repair promote further genome instability and mutation and tumor evolution

However, most morbidity and mortality from cancer is the result of tumor invasion and/ or metastasis, and tumor progression involves the activation of invasion and metastasis, the induction of angiogenesis

to provide a blood supply to the growing tumor, the evasion of the immune system, and altered cell energetics

Epigenetic Reprogramming

In virtually all tumor types, a newly described class of frequently mutated genes consists of enzymes involved in histone modification and components of chromatin remodeling complexes One model of cancer involves a stage of cellular reprogramming of expression state (epigenetic state), altering entire programs of gene expression The ability to turn on or off whole programs of expression allows the tumor cell to adapt to virtually all encountered barriers

• Invasion: The cancer cell can turn on (or the cancer stem cells

already possess) the ability to respond to homing signals, migrate through tissue into the bloodstream, and travel to other parts of the body Portions of the tumor may establish a niche whereby other cells in the tumor can proliferate and grow

• Immune evasion: Tumor cells can express immune evasion signals

(such as PD- L1) and inhibit immune cells that recognize the tumor

as foreign

• Cell metabolism: Many tumor cells alter their energetics and

metab-olism For example, initiating programs of gene expression that allow the use of glucose, or anaerobic glycolysis

Defective DNA Repair/

Chromosome Integrity

Replication of Damaged Cell

Tumor Suppressor Loss

DNA Repair Gene Mutation

Figure 2–3a Hallmarks of cancer cells: early stages Tumors often show

properties common across cancer types (a) Self-renewal is a type of cell

division that generates daughter cells identical to the original cell Stem

cells also undergo this type of division These cells have been termed

can-cer stem cells, or cancan-cer-initiating cells (b) Whereas most normal cells

will stop dividing as they come in contact with other cells (contact

inhibi-tion), many tumor cells continue to grow (c) Reduced cell death Cell death

is a normal process for many cells in the body, but many tumor cells are

defective in this process and continue to survive in the presence of damage

(d) Tumor cells can be defective in DNA repair their telomeres

Invasion and Metastasis

Loss of Immune Control

Angiogenesis

Altered energetics/ Metabolism

EpigeneticReprogramming

Figure 2–3b Hallmarks of cancer cells: late stages (e) The immune system can recognize abnormal or infected cells and destroy them Cancer cells can evade the immune system One mechanism is the secretion of inhibitory factors such as checkpoints and cytokines (f) A growing tumor requires a blood supply, and tumors can stimulate the production of new blood ves-sels (g) Many tumors switch to glycolosis as an energy source (Warburg effect) (h) Invasion into surrounding tissues and other parts of the body (metastasis) are common tumor properties (i) Tumor cells mutate one or more genes that modify the histone proteins that form chromatin (chromatin remodeling genes) and permit epigenetic reprogramming of the genome

Trang 35

Tumor Differentiation and Evolution

At first glance, the tumor seems like one of the mythical beasts that

Ulysses encountered on his journey, a Hydra or a Cyclops But in fact

the cancer cell is a human cell and therefore utilizes tools that our

normal cells employ and is subject to most of the same constraints

and restrictions A human being goes from a single fertilized egg to an

adult with thousands of cell types and tissues with diverse function,

without changing the DNA of the cells (Figure 2–4) This development

and differentiation are all accomplished by the coordinated activation

and repression of distinct programs of gene expression, the secretion

of proteins and other factors that influence neighboring cells, and the

contact and interaction of cells with each other The tumor likewise

evolves from a single cell, forms a network of self- renewing and

com-mitted cells, secretes factors that stimulate other cells in the tumor,

and interacts with other tumor cells as well as normal cells

The transition from a normal fetus to an adult human involves the

continued growth and differentiation of cells to form multiple organs

and specialized cell types Cells migrate in the growing human,

estab-lish niches for stem cells, develop blood supplies, and in certain cases

alter metabolism and energy utilization In a parallel fashion, the cells

of a growing and developing tumor alter gene expression profiles in

subpopulations of cells, migrate to different locations, adapt to new

environments, survive over years or decades, and adapt to therapeutic

strategies to eliminate the tumor cell

A key finding in the connection between the developing embryo and

the tumor was the discovery of oncogenes and tumor suppressors that

are also key embryonic differentiation factors Gorlin syndrome, or

nevoid basal cell carcinoma syndrome (NBCCS), is an inherited

disor-der characterized by hundreds to thousands of basal cell carcinoma, jaw

cysts, facial and bone abnormalities, and occasionally medulloblastomas (Anderson et al., 1967; Gorlin and Goltz, 1960) Genetic linkage studies localized the gene for NBCCS to chromosome 9, and the gene responsi-

ble was identified as Patched (PTCH1) (Gailani et al., 1992; Hahn et al.,

1996; Johnson et al., 1996) This gene was first discovered in drosophila

as part of a group of genes required for normal embryonic development

of the fly, and mutants cause the patchy formation of bristles Patched is

a cell surface protein and the receptor for a class of secreted, modified peptides, Hedgehogs (SHH) (Dean, 1996) Patched repressed the activ-ity of a G- protein- coupled receptor Smoothened (SMO), and the binding

of SHH to PTCH1 releases SMO from repression and leads eventually

to changes in gene regulation in the nucleus through the GLI family of transcription factors (van den Heuvel and Ingham, 1996)

All of the major components of this signaling pathway are conserved

in insects, vertebrates, and other organisms The correct functioning of these proteins is required for the development of the growing embryo and the polarity and symmetry of cells and structures during develop-ment We now recognize all the components of this pathway as either

tumor suppressor genes (PTCH1) or oncogenes (SHH, SMO, GLI) In nearly all cases of NBCCS, we find germline mutations in the PTCH1

gene, and in nearly all sporadic basal cell carcinomas and a portion of

medulloblastomas PTCH1 is mutated This work provided a clear

con-nection to the functions of development and differentiation of onic stem cells and the formation of tumors (Bale and Yu, 2001)

embry-Evolution and the Tumor

Our species and all other vertebrates are the product of a billion years

of evolution Through this process we evolved unique characteristics,

Blood Cells

Normal Stem Cell

StemCell

Embryo

Fetus

TumorCell

Pre-Malignant

Invasive Lesion

Neurons

Muscle Cells Kidney Cells

Angiogenic Cells

Tumor StemCell

of a large number of cell types This can allow spread of the tumor, evasion of cellular processes such as immune attack, and/or survival from therapeutic approaches (b) The tumor also undergoes evolutionary selection The tumor cell progresses from a single-cell organism to a small multi-cellular organ-ism Just as evolution allowed diverse classes and species of organisms to evolve from largely the same sets of genes, the tumor can undergo selection for altered, duplicated, or deleted genes To adapt to new conditions and survive therapy, the tumor adapts resistance mechanisms Many of the same principles

of natural selection apply to tumors as apply to populations of species

Trang 36

adapted to changing environments, individuals were selectively

removed, and surviving individuals went on to reproduce and survive

A tumor also undergoes this same selective process, only in the

context of our body and over the time scale of years, decades, or our

lifespan Many cells in the tumor die, and necrotic areas of larger

tumors are often seen Radiation, chemotherapy, and other treatments

can kill off large portions of the cancer, and in some cases all cells,

resulting in cures However, those cells that survive treatment and

retain the property of self- renewal can continue to grow, resulting in

remission Many of the principles and concepts that were developed

and used to understand and explain evolution can also be applied to

the evolution of cancer DNA sequencing of tumor cells at different

sites of the body and/ or stages of the disease is resulting in deeper

understanding of these processes, sometimes at the resolution of

sin-gle cells (Xu et al., 2012)

Key Points

• Cancer cells within a tumor display tremendous diversity

• New mutations, epigenetic changes, and gene rearrangements occur

continually during the cancer process

• Cancer cells face barriers to survival that include immune evasion,

therapy resistance, and development of invasion/ metastasis

Tumor Microenvironment

The cancer cell is a human cell, with many properties in common with

non- malignant human cells In the body of the individual, the cancer

cell interacts in a complex manner with the host cells Therefore, there

exists a microenvironment surrounding the tumor that is composed of

host cells interacting and responding to the tumor and is subject to

modification by the tumor cells The major classes of cells in the tumor

microenvironment (TME) include the following:

greater than 2 millimeters requires the development of new blood

vessels (angiogenesis) However, the interior of many tumors is

hypoxic and may contribute to tumor cell properties including

ther-apy resistance

Stromal cells: Fibroblasts can be recruited into the TME and their

func-tions altered to support tumor growth (Hanahan and Coussens, 2012;

Pickup et al., 2013) Myeloid- derived suppressor cells (MDSCs) are

cells derived from the myeloid lineage’s T cell repressive properties

(Talmadge and Gabrilovich, 2013) Tumor- infiltrating lymphocytes

are T cells with reactivity against the tumor cells In certain cases,

these cells have been expanded and used therapeutically (Hinrichs and

Rosenberg, 2014)

Cancer cells need to evade the immune system’s constant probing

for virally infected or otherwise foreign cells This can be

accom-plished by not expressing foreign antigens or suppressing the immune

response Many of the new checkpoint inhibitor cancer strategists

using PD- L1 or PD- 1 inhibitors seek to reverse the immune

suppres-sion of tumors (Sunshine and Taube, 2015)

FIELDS OF EPIDEMIOLOGY AND

THE CANCER PROCESS

Our understanding of the process of cancer development not only has

been shaped by, but also influences all areas of epidemiology, and

increasingly genetic and genomic studies of cancer are applied to

sub-divisions of the field

Genetic Epidemiology

Classical genetic epidemiology involved the identification of rare

can-cer families, the study of genetic conditions that contain cancan-cer as a

phenotype, and the determination of heritability More recent genetic

epidemiology has involved the study of genome- wide association

studies (GWAS), identification of common but low relative risk alleles,

and more broadly the genetic architecture of genetic risk in tions in relation to cancer (Chung and Chanock, 2011) Currently there are almost 500 loci identified in the genome that modify risk of one

popula-or mpopula-ore cancers However, in only a few cases do we understand the molecular basis of these effects

Occupational/ Environmental Epidemiology

The role of environment (broadly defined, including lifestyle factors) and occupation in the development of cancer has been critical Acute

or long- term environmental or occupational exposures can contribute

to chronic inflammation or tissue damage (smoking, asbestos sure, benzene, etc.) Agents may cause tissue damage/ inflammation (UV exposure, radiation, asbestos, infectious agents), or may induce mutations (radiation, mutagenic chemicals, mutagenic viruses), or commonly both In some cases (tobacco, UV exposure, viral infec-tions), we can now detect specific mutational signatures in cancer genomes

expo-Radiation Epidemiology

Radiation exposure can occur through the environment, occupation, accidents, or medical treatment, and can often be accurately measured and modeled in many organisms While in acute or concentrated doses radiation can cause tissue damage, the majority of the carcinogenesis

of radiation is due to DNA damage and mutation

Hormonal Epidemiology

Many major cancers involve reproductive organs or organs sive or largely to one gender (cervix, prostate, breast, ovary) These tumor types involve major hormonal effects For example, many breast tumors are dependent on estrogens and/ or progesterone, and this knowledge has aided both prevention and therapy

exclu-However, there is also a large gender discordance for many other types: male bias in bladder cancer, any pediatric cancers, and liver cancer; female bias in chronic lymphocytic leukemia and thyroid can-cer In most cases we do not understand these effects As hormone production and exposure change dramatically over time for males and females, this is a critical area of cancer epidemiology

Infectious Disease Epidemiology

Many of the world’s most common cancers and the cancers that play the greatest disparities due to income involve infectious agents

dis-(cervical and HPV, liver and HBV and HCV, and gastric and H. pylori)

(Klein, 2002; Schiffman and Castle, 2005; Torre et al., 2015; Uemura

et al., 2001; zur Hausen and de Villiers, 1994) The list of direct causes

of cancer by infectious agents and pathogens is impressive And the role of other chronic infections remains to be fully explored

Nutritional Epidemiology

The nutritional state of the individual, diet, body mass index (BMI), and related conditions such as diabetes and obesity are increasingly recognized as major risk factors for cancers Studies show that virtu-ally all cancer types increase with increasing BMI Furthermore, the increase in BMI and obesity in world populations has led to projec-tions that this will be a more significant risk factor than tobacco in the future (Calle et al., 2003)

The mechanisms by which increased body mass contributes to cer are poorly studied However, some of the hormones regulated by energy consumption, such as insulin and insulin- like growth factors, are tied to cellular pathways for growth and division Many tumor suppressor and oncogene networks connect to metabolic pathways, and some enzymes of the tricarboxylic acid (TCA) cycle are tumor suppressor genes In addition, fat deposition and obesity can cause an inflammatory state and may therefore be analogous to other inflamma-tory states such as viral and bacterial infections

Trang 37

can-16 PART I: BASIC CONCEPTS

AGING AND CANCER

Just as cancer is a genetic disease, it is also a disease of aging Age is

the most important risk factor for most adult cancers, and many of the

processes that we understand are operative in cancer development can

be thought of in light of aging

DNA Damage

With time, the cells of our body accumulate DNA damage Much

of this occurs through natural processes, such as the deamination of

5- methyl- cytosine found in many DNA regions; 5- methyl- cytosine

spontaneously deaminates to uracil, and the body has an enzyme/ DNA

repair system to recognize this damage and repair it However, this

process is error prone, and DNA lesions accumulate over time, leading

in cases to the accumulation of sufficient mutations in a single cell to

start the cancer process

One of the new areas of cancer genome study is the ability to

exam-ine the whole genomes of cancer cells and discern the signatures of

mutational processes (Alexandrov et al., 2013) By tabulating all

muta-tions in the tumor genome, and noting the nature of the base changes

and the local context of the mutation, signatures can be identified

The most common cancer signature is indeed one in which a CpG is

mutated (Shiraishi et al., 2015)

Chronic Infection

Especially for cervical, gastric, and liver cancer, chronic infections

by HPV, H. pylori, and hepatitis viruses (HBV, HCV) are the

prin-cipal cause None of these infections causes cancer shortly after

infection; all require a long latency period in which the infectious

agent stimulated tissue damage, inflammation, local repair, and

activation of tissue stem cells before premalignant and malignant

lesions arise

It is recently appreciated that activation of innate anti- viral infection

mechanisms, such as the ABOBEC family of anti- viral proteins, over

long periods of time can lead to somatic mutations in tumor cells The

second most common signature is one consistent with APOBEC

activa-tion (Lawrence et al., 2013; Shiraishi et al., 2015) It is indeed commonly

found in cervical cancer and head and neck cancers associated with HPV

infection, but also with acute lymphocytic leukemia (ALL), bladder,

breast, kidney, pancreas, and some forms of lung cancer Whether there

are unappreciated viral infections influencing these cancers, or alternate

pathways to activate APOBECs, remains to be determined

Chronic Inflammation

A large number of cancers are caused by chronic inflammation caused

by radiation, irritants, toxins, or immune inflammatory processes For

example, a person with ulcerative colitis or Crohn’s disease has an

elevated risk to develop colorectal cancer (Jess et al., 2012) The role

of asbestos in lung cancer, UV radiation and sunburn in skin cancer,

and irritants in tobacco products in lung and oral cancers are well

doc-umented As with viral infections, these cancers develop over many

decades, and the aspects of aging on the development of cancer due to

inflammation are poorly understood

Cell Senescence and Telomere Regulation

Telomeres are the protective elements at the chromosome ends that

aid in preventing unscheduled DNA repair and degradation They are

composed of repetitive DNA sequences (TTAGGG repeats) bound to

an array of proteins with specialized functions (Donate and Blasco,

2011) The protection afforded to the chromosome depends on the

length of telomeric repeats and the integrity of the telomere- binding

proteins The length and integrity of telomeres are regulated by specific

epigenetic modifications, indicating that a higher order control of

telo-mere function exists After each round of cell division is completed,

the telomeres are shortened because the conventional DNA ases are unable to replicate the ends of chromosomes This is known as the “end replication problem” (Donate and Blasco, 2011) This is why

polymer-a cellulpolymer-ar enzyme cpolymer-alled telomerpolymer-ase mpolymer-ay be recruited in certpolymer-ain situpolymer-a-tions to add TTAGGG repeats to the chromosome ends The enzyme telomerase consists of a subunit Tert that possesses reverse transcrip-tase activity, an RNA element called Terc that is the template on which DNA is synthesized, and a protein called dyskerin (DKc1) that has the ability to bind and stabilize Terc Upregulated telomerase expression

situa-is a feature of pluripotent stem cells and also of early stage embryonic development; however, telomerase activity can also be present in adult stem cell compartments (Blasco, 2005)

Telomere length and Tert are critical factors determining the zation of stem cells However, it was also found that shelterin, a major protein complex bound to mammalian telomeres, may play a role in the initiation of neoplasia (Donate and Blasco, 2011) Mice conditionally deleted for the shelterin proteins TRF1, TPP1, and Rap1 were gener-ated, and this demonstrated that telomere dysfunction is sufficient to produce premature tissue degeneration, acquisition of chromosomal aberrations, and the initiation of neoplastic lesions (Blasco, 2005) Based on further experiments, a stemcell- based model for the role

mobili-of telomeres related to cancer and aging was proposed (Donate and Blasco, 2011) It was proposed that in young or adult organisms, stem cells have the ability to repopulate certain tissues as needed During this process, the stem cells undergo telomere shortening, which is only partially counterbalanced by the action of telomerase However,

in older organisms, the stem cell telomeres are actually too short Critically short telomeres are recognized by the body as DNA damage This recognition of potential DNA damage activates a p53- mediated DNA damage signaling response that can impair stem cell mobiliza-tion, leading to suboptimal tissue regeneration, which may ultimately lead to organ failure This phenomenon of decreased stem cell mobi-lization reduces the probability that abnormal cells will accumulate in tissues and thus provides a mechanism for cancer protection On the other hand, if stem cells express aberrantly high levels of the enzyme telomerase (by the acquisition of tumorigenic, telomerase- reactivating mutations), stem cell mobilization is more efficient and tissue fitness

is maintained for a longer time This would increase life span, but would also increase the probability of initiating a tumor (Donate and Blasco, 2011)

The Aging Immune System and Cancer

It has been found that the incidence of most common cancers increases with age and may possibly be caused by a decline in immune func-tion This phenomenon is called “immune senescence” (Foster et al., 2011) The causal determinant for this phenomenon, although highly controversial, may be deleterious changes to T- cell subsets that over time may impair immunity In one murine study, tumors were able to limit the function of CD8 + T cells, including specific responses to tumor antigens specific for prostate tumors (Anderson et al., 2007) Another study showed that microvesicles isolated from human tumor cells were able to induce the generation and expansion of Treg (T regu-latory cells) that are capable of suppressing T- cell responses (Szajnik

et al., 2010) In relation to this, the immunoregulatory effects of the tumor (and its environment) on T cells, along with the decreasing abil-ity of the body to replace nạve T- cell populations, may allow cancers

to progress because T cells seem to have a fixed immune repertoire associated with age- related immunosenescence The intricate mecha-nisms that lead to declining cellular immunity are very unclear, but the progressive decline in thymic output of T cells may be the primary event leading to immune senescence in old age With age, the produc-tion of new nạve T cells reduces, resulting in the peripheral T cells having to undergo repeated rounds of cell division in order to main-tain the status quo so that the overall size of the T- cell compartment remains relatively stable This prolonged survival of the peripheral

T cells results in defects in all types of activation states of T cells, from the nạve to the terminally differentiated This reduction in functional immunity can be permissive to tumor formation and might promote

Trang 38

tumor formation by contributing to low- level inflammation In order

to circumvent this pro- inflammatory environment, specific cytokine

blockers may be beneficial to both neoplasia and other inflammatory

disease states, but have the drawback of compromising the immune

system (Zidi et al., 2010) Thus, more research needs to be carried out

to determine the specific mechanisms of action of cytokine blockers to

allow for the development of targeted therapies

Another strategy would be to delay replicative senescence in T cells,

utilizing strategies that sustain telomerase activity (Effros, 2011)

While we have learned a lot about the role of age- related declining

immunity as it relates to cancer incidence, the precise cellular

mecha-nisms through which this occurs still remain elusive As research

pro-gresses, from animal models to human clinical trials, new therapeutic

strategies to enhance immune function may become powerful tools to

reduce cancer incidence (Foster et al., 2011)

CONCLUSIONS

Many of the genetic and biological processes that occur in cancer cells are

now well documented Through the study of animal models, cell culture

systems and human tumors we have detailed catalogs of the following:

• The genes in which germline mutations cause highly penetrant

can-cer syndromes;

• Genetic loci from GWAS studies that influence cancer risk;

• Genes frequently altered somatically in cancer

There is considerable work still in progress to identify rare alleles

involved in cancer risk, to understand the function of cancer GWAS

loci, and to understand complex copy number changes and gene

rear-rangements in cancer cells

Most of the major environmental and occupational risk factors

associ-ated with cancer have been identified and can now be at least partially

understood at the molecular level from current data The exciting new

area of mutation signatures promises to further reveal the effect of

spe-cific exposures on somatic alterations in tumors However, even the well-

known cancer- causing infections, such as human papillomavirus (HPV),

hepatitis B virus (HBV), and hepatitis C virus (HCV), are only partially

understood, and there is much left to be learned regarding agents such

as asbestos, tobacco smoke, and chemical carcinogens The long latency

of cancer and the complexity of the human immune system and tumor

microenvironment leave enormous areas for future investigation

Epigenetic modification has emerged as a major factor in the

devel-opment of cancer, largely in the later stages of the disease Nearly

all cancers have mutations in one or more histone- modifying enzyme

affecting chromatin structure, and/ or enzyme- regulating DNA

methyl-ation (Figure 2–5) An intriguing outcome of these findings is that the

deregulation of the epigenome, and/ or “return to an embryonic state,”

may be a newly appreciated hallmark of cancer This model would

suggest that cancer cells can acquire a plastic state that would allow the generation of cells in the tumor capable of diverse functions and evolution A therapeutic attraction to this model is that this may be a reversible process that could be manipulated

Single cell genomic analyses of tumors are revealing an enormous level of complexity and heterogeneity These studies can be used to follow the evolution of cancer cells over both time and location in the cancer patient

References

Al- Hajj M, Becker MW, Wicha M, Weissman I, and Clarke MF 2004

Therapeutic implications of cancer stem cells Curr Opin Genet Dev,

14(1), 43– 47 PMID: 15108804

Al- Hajj M, Wicha MS, Benito- Hernandez A, Morrison SJ, and Clarke MF

2003 Prospective identification of tumorigenic breast cancer cells Proc

Natl Acad Sci U S A, 100(7), 3983– 3988 PMCID: PMC153034

Alexandrov LB, Nik- Zainal S, Wedge DC, et al 2013 Signatures of

muta-tional processes in human cancer Nature, 500(7463), 415– 421 PMCID: PMC3776390

Anderson DE, Taylor WB, Falls HF, and Davidson RT 1967 The nevoid

basal cell carcinoma syndrome Am J Hum Genet, 19(1), 12– 22 PMCID: PMC1706134

Anderson MJ, Shafer- Weaver K, Greenberg NM, and Hurwitz AA 2007 Tolerization of tumor- specific T cells despite efficient initial priming in a

primary murine model of prostate cancer J Immunol, 178(3), 1268– 1276

PMID: 17237372

Bale AE, and Yu KP 2001 The hedgehog pathway and basal cell carcinomas

Hum Mol Genet, 10(7), 757– 762 PMID: 11257109

Blasco MA 2005 Telomeres and human disease: ageing, cancer and beyond

Nat Rev Genet, 6(8), 611– 622 PMID: 16136653

Calle EE, Rodriguez C, Walker- Thurmond K, and Thun MJ 2003 Overweight, obesity, and mortality from cancer in a prospectively studied cohort of

U.S. adults N Engl J Med, 348(17), 1625– 1638 PMID: 12711737.

Cancer Genome Atlas Research Network 2011 Integrated genomic analyses of

ovarian carcinoma Nature, 474(7353), 609– 615 PMCID: PMC3163504.

Cancer Genome Atlas Network 2012 Comprehensive molecular portraits of

human breast tumours Nature, 490(7418), 61– 70 PMCID: PMC3465532.

Chung CC, and Chanock SJ 2011 Current status of genome- wide association

studies in cancer Hum Genet, 130(1), 59– 78 PMID: 21678065.

Clark SS, McLaughlin J, Timmons M, et al 1988 Expression of a distinctive BCR- ABL oncogene in Ph1- positive acute lymphocytic leukemia (ALL)

Science, 239(4841 Pt 1), 775– 777 PMID: 3422516

Couch FJ, Nathanson KL, and Offit K 2014 Two decades after

BRCA: set-ting paradigms in personalized cancer care and prevention Science,

kaemia Nature, 300(5894), 765– 767 PMID: 6960256.

Dean M 1996 Polarity, proliferation and the hedgehog pathway Nat Genet,

14(3), 245– 247 PMID: 8896548

Dean M 1997 Towards a unified model of tumor suppression: lessons learned

from the human patched gene Biochim Biophys Acta, 1332(2), M43–

M52 PMID: 9141460

Dean M, Fojo T, and Bates S 2005 Tumor stem cells and drug resistance Nat

Rev Cancer, 5(4), 275– 284 PMID: 15803154

Dean M, Park M, Le Beau MM, et al 1985 The human met oncogene is

related to the tyrosine kinase oncogenes Nature, 318(6044), 385– 388

Donate LE, and Blasco MA 2011 Telomeres in cancer and ageing Philos

Trans R Soc Lond B Biol Sci, 366(1561), 76– 84 PMCID: PMC3001316.Druker BJ, Sawyers CL, Kantarjian H, et al 2001a Activity of a specific inhibi-tor of the BCR- ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chro-

mosome N Engl J Med, 344(14), 1038– 1042 PMID: 11287973.

Druker BJ, Talpaz M, Resta DJ et al 2001b Efficacy and safety of a specific inhibitor of the BCR- ABL tyrosine kinase in chronic myeloid leukemia

N Engl J Med, 344(14), 1031– 1037 PMID: 11287972

Effros RB 2011 Telomere/ telomerase dynamics within the human immune

system: effect of chronic infection and stress Exp Gerontol, 46(2– 3),

(a)

(b)

(c)

(d)

Figure 2–5 Chromatin alteration and remodeling (a) DNA is a charged

molecule (b) DNA is covered by histones and other proteins to form

nucleosomes and chromatin (c) Histones undergo extensive modification

(d) Other protein complexes (chromatin remodelers) can alter the location

of nucleosomes, make DNA open to transcription factors turning genes on,

or, covering regulatory regions, turning genes off

Trang 39

Foster AD, Sivarapatna A, and Gress RE 2011 The aging immune

sys-tem and its relationship with cancer Aging Health, 7(5), 707– 718

PMCID: PMC3222953

Gailani MR, Bale SJ, Leffell DJ, et al 1992 Developmental defects in Gorlin

syndrome related to a putative tumor suppressor gene on chromosome 9

Cell, 69(1), 111– 117 PMID: 1348213

Gorlin RJ, and Goltz RW 1960 Multiple nevoid basal- cell epithelioma,

jaw cysts and bifid rib A syndrome N Engl J Med, 262, 908– 912

PMID: 13851319

Gui Y, Guo G, Huang Y, et al 2011 Frequent mutations of chromatin

remodel-ing genes in transitional cell carcinoma of the bladder Nat Genet, 43(9),

875– 878 PMID: 21822268

Guo G, Sun X, Chen C et al 2013 Whole- genome and whole- exome

sequenc-ing of bladder cancer identifies frequent alterations in genes involved

in sister chromatid cohesion and segregation Nat Genet, 45(12), 1459–

1463 PMID: 24121792

Hahn H, Wicking c, Zaphiropoulous PG, et al 1996 Mutations of the human

homolog of Drosophila patched in the nevoid basal cell carcinoma

syn-drome Cell, 85(6), 841– 851 PMID: 8681379.

Hanahan D, and Coussens LM 2012 Accessories to the crime: functions of

cells recruited to the tumor microenvironment Cancer Cell, 21(3), 309–

Harding C, Pompei F, Lee EE, and Wilson R 2008 Cancer suppression at old

age Cancer Res, 68(11), 4465– 4478 PMID: 18519710.

Hinrichs CS, and Rosenberg SA 2014 Exploiting the curative potential

of adoptive T- cell therapy for cancer Immunol Rev, 257(1), 56– 71

PMCID: PMC3920180

Jess T, Rungoe C, and Peyrin- Biroulet L 2012 Risk of colorectal cancer in

patients with ulcerative colitis: a meta- analysis of population- based cohort

studies Clin Gastroenterol Hepatol, 10(6), 639– 645 PMID: 22289873.

Johnson RL, Rothman AL, Xie J, et al 1996 Human homolog of patched, a

candidate gene for the basal cell nevus syndrome Science, 272(5268),

1668– 1671 PMID: 8658145

Kandoth C, McLellan MD, Vandin F, et al 2013 Mutational landscape and

significance across 12 major cancer types Nature, 502(7471), 333– 339

PMCID: PMC3927368

Kinzler KW, Nilbert MC, Su LK, et al 1991 Identification of FAP locus genes

from chromosome 5q21 Science, 253(5020), 661– 665 PMID: 1651562.

Kinzler KW, and Vogelstein B 1996 Lessons from hereditary colorectal

can-cer Cell, 87(2), 159– 170 PMID: 8861899.

Kinzler KW, and Vogelstein B 1997 Cancer- susceptibility genes: gatekeepers

and caretakers Nature, 386(6627), 761– 763 PMID: 9126728.

Kishimoto M, Kohno T, Okudela K et al 2005 Mutations and deletions of the

CBP gene in human lung cancer Clin Cancer Res, 11(2 Pt 1), 512– 519

PMID: 15701835

Klein G 2002 Perspectives in studies of human tumor viruses Front Biosci, 7,

d268– 274 PMID: 11779700

Knudson AG, Jr 1971 Mutation and cancer: statistical study of retinoblastoma

Proc Natl Acad Sci U S A, 68(4), 820– 823 PMCID: PMC389051

Knudson AG 2002 Cancer genetics Am J Med Genet, 111(1), 96– 102

PMID: 12124744

Latif F, Tory K, Gnarra J, et al 1993 Identification of the von Hippel- Lindau disease

tumor suppressor gene Science, 260(5112), 1317– 1320 PMID: 8493574.

Lawrence MS, Stojanov P, Polak P, et al 2013 Mutational heterogeneity in

can-cer and the search for new cancan-cer- associated genes Nature, 499(7457),

214– 218 PMCID: PMC3919509

Miki Y, Swensen J, Shattuck- Eidens D, et al 1994 A strong candidate for

the breast and ovarian cancer susceptibility gene BRCA1 Science,

266(5182), 66– 71 PMID: 7545954

Moran A, O’Hara C, Khan S, et al 2012 Risk of cancer other than breast or

ovarian in individuals with BRCA1 and BRCA2 mutations Fam Cancer,

11(2), 235– 242 PMID: 22187320

Muraoka M, Konishi M, Kikuchi- Yanoshita R, et al 1996 p300 gene

altera-tions in colorectal and gastric carcinomas Oncogene, 12(7), 1565– 1569

PMID: 8622873

Murphree AL, and Benedict WF 1984 Retinoblastoma: clues to human

onco-genesis Science, 223(4640), 1028– 1033 PMID: 6320372.

Pedersen JK, Engholm G, Skytthe A, Christensen K 2016 Cancer and aging:

epidemiology and methodological challenges Acta Oncol, 55(Suppl 1),

7–12 PMID: 26825001

Pickup M, Novitskiy S, and Moses HL 2013 The roles of TGFbeta in

the tumour microenvironment Nat Rev Cancer, 13(11), 788– 799 PMCID: PMC4025940

Pulciani S, Santos E, Lauver AV, et al 1982 Oncogenes in human tumor cell lines: molecular cloning of a transforming gene from human blad-

der carcinoma cells Proc Natl Acad Sci U S A, 79(9), 2845– 2849 PMCID: PMC346303

Rowley JD 1973 Letter: A new consistent chromosomal abnormality in chronic myelogenous leukaemia identified by quinacrine fluorescence

and Giemsa staining Nature, 243(5405), 290– 293 PMID: 4126434.

Santos- Rosa H, and Caldas C 2005 Chromatin modifier enzymes, the histone

code and cancer Eur J Cancer, 41(16), 2381– 2402 PMID: 16226460.

Schiffman M, and Castle PE 2005 The promise of global cervical- cancer

pre-vention N Engl J Med, 353(20), 2101– 2104 PMID: 16291978.

Shiraishi Y, Tremmel G, Miyano S, and Stephens M 2015 A simple model- based approach to inferring and visualizing cancer mutation signatures

PLoS Genet, 11(12), e1005657 PMCID: PMC4667891

Stehelin D, Varmus HE, Bishop JM, and Vogt PK 1976 DNA related to the transforming gene(s) of avian sarcoma viruses is present in normal avian

DNA Nature, 260(5547), 170– 173 PMID: 176594.

Struewing JP, Hartge P, Wacholder S, et al 1997 The risk of cancer associated with specific mutations of BRCA1 and BRCA2 among

Ashkenazi Jews [see comments] N Engl J Med, 336(20), 1401– 1408

PMID: 9145676

Sunshine J, and Taube JM 2015 PD- 1/ PD- L1 inhibitors Curr Opin Pharmacol,

23, 32– 38 PMCID: PMC4516625

Szajnik M, Czystowska M, Szczepanski MJ, Mandapathil M, and Whiteside

TL 2010 Tumor- derived microvesicles induce, expand and up- regulate

biological activities of human regulatory T cells (Treg) PLoS One, 5(7),

e11469 PMCID: PMC2908536

Talmadge JE, and Gabrilovich DI 2013 History of myeloid- derived suppressor

cells Nat Rev Cancer, 13(10), 739– 752 PMCID: PMC4358792 Torre LA, Bray F, Siegel RL, et al 2015 Global cancer statistics, 2012 CA

Cancer J Clin, 65(2), 87– 108 PMID: 25651787

Uemura N, Okamoto S, Yamamoto S, et al 2001 Helicobacter pylori infection

and the development of gastric cancer N Engl J Med, 345(11), 784– 789

PMID: 11556297

van den Heuvel M, and Ingham PW 1996 Smoothened encodes a receptor- like

serpentine protein required for hedgehog signalling Nature, 381(6591),

547– 551 PMID: 8700230

Varley JM 2003 Germline TP53 mutations and Li- Fraumeni syndrome Hum

Mutat, 21(3), 313– 320 PMID: 12619118

Vogelstein B, Papadopoulos N, Velculescu VE, et al 2013 Cancer genome

landscapes Science, 339(6127), 1546– 1558 PMCID: PMC3749880.

Watson R, Oskarsson M, and Vande Woude, GF 1982 Human DNA sequence homologous to the transforming gene (mos) of Moloney

murine sarcoma virus Proc Natl Acad Sci U S A, 79(13), 4078– 4082

PMCID: PMC346580

Weinberg RA 1996 How cancer arises Sci Am, 275(3), 62– 70 PMID: 8701295.

Wooster R, Bignell G, Lancaster J, et al 1995 Identification of the breast cancer

susceptibility gene BRCA2 Nature, 378(6559), 789– 792 PMID: 8524414.

Xu X, Hou Y, Yin X, et al 2012 Single- cell exome sequencing reveals single-

nucleotide mutation characteristics of a kidney tumor Cell, 148(5), 886–

895 PMID: 22385958

Zidi I, Mestiri S, Bartegi A, and Amor NB 2010 TNF- alpha and its inhibitors

in cancer Med Oncol, 27(2), 185– 198 PMID: 19277912.

zur Hausen H, and de Villiers EM 1994 Human papillomaviruses Annu Rev

Microbiol, 48, 427– 447 PMID: 7826013

Trang 40

MARK E SHERMAN, MELISSA A TROESTER, KATHERINE A HOADLEY, AND WILLIAM F ANDERSON

OVERVIEW

Defining etiological, biological, and clinical heterogeneity are

impor-tant goals in epidemiologic research To achieve these aims,

epide-miologists conduct transdisciplinary studies to relate both newly

identified and established cancer risk factors to specific cancer

sub-types defined by histopathological and molecular features Thus, the

traditional model of identifying cancer risk factors by organ site alone

has been supplemented, and largely supplanted

Relating specific cancer risk factors to groups of tumors that share

a common molecular pathogenesis offers opportunities to strengthen

risk associations and provides a strong basis for translational research

on early detection or prevention based on underlying biology The

promise of this strategy has been revealed through efforts to offer

carriers of deleterious mutations in high- penetrance genes, such as

cancers, and the administration of prophylactic human

papillomavi-rus (HPV) vaccines to dramatically reduce risks of cervical cancer

and HPV- induced cancers arising at other anogenital sites and in the

head and neck

This chapter provides a primer on approaches that use pathology

and molecular biology to subclassify cancers in epidemiologic studies

Given the breadth of this topic and its rapid evolution with changing

technologies, the goal is to emphasize important principles and

pro-vide illustrative examples, rather than to comprehensively review this

subject Readers are encouraged to consult specific chapters

through-out the text that demonstrate the application of the principles presented

to cancers of specific organs or exposures

INTRODUCTION

Cancer is a biologically and clinically heterogeneous class of

dis-eases Much of epidemiological research is concerned with

under-standing the causal and mechanistic sources of this diversity

because identification of etiologically “refined” cancer subtypes

may increase the specificity of risk associations and predictions,

enabling more effective screening, prevention, and population

surveillance

Cancers arising from different organs are associated with distinct

clinical presentations and behaviors, and therefore primary

ana-tomic site of origin is a critical determinant of clinical management

However, expanding knowledge of the epidemiology, pathology, and

molecular profiles of cancer, and the intersection of these factors, has

led to development of novel tumor classifications, which categorize

tumors within and across organ sites and disciplines This new

per-spective has had a major influence on therapeutic research, and has

led to development of “basket trials,” in which participant eligibility

is based on molecular characteristics of tumors, rather than primary

sites In treatment trials, such as the MATCH Trial (McNeil, 2015),

individuals with cancers that demonstrate particular candidate driver

mutations are selected to receive targeted therapies, irrespective of site

of origin Basket trials are innovative, but their promise is tempered

by data suggesting that primary site remains an indicator of response,

even among cancers that share a common mutation in a putative driver

gene, such as BRAF (Hyman et al., 2015).

The “basket trial” concept has received less attention in the context

of etiological and prevention research than in the realm of treatment trials, but recognition that cancers of different sites may share com-mon etiologic exposures and molecular mechanisms of carcinogene-sis suggest that this concept may have similar potential For cancers that are causally linked to an infectious agent, such as carcinogenic human papillomavirus (HPV) genotypes (Jemal et al., 2013), highly effective prophylactic HPV vaccines may lower risks of HPV- related cancers of the cervix, vagina, vulva, anus, penis, and head and neck (Herrero et al., 2015) For cancers driven by common molecular mechanisms, such as specific DNA repair defects (e.g., Lynch syn-

drome and BRCA1/ 2 mutations), hormonal imbalances or metabolic

disturbances (e.g., breast and gynecologic cancers), the possibility of identifying individuals at risk for a set of cancers that are related to common etiological and/ or mechanistic factors offers the prospect of lowering incidence rates of several cancers with a single intervention Achieving this aim will require epidemiologists to extend their efforts from accurate assessment of exposures to detailed characterization of endpoints

Evolving methods of tumor classification integrate data across tiple scales, encompassing populations, patient- level factors, primary anatomical site, histopathological typing, and molecular characteriza-tion Research at each of these scales, as well as synthesis of informa-tion across them, is needed to gain the comprehensive understanding

mul-of cancer etiology and pathogenesis that will be required to develop improved strategies for cancer prevention and lower cancer mortality.The aims of this chapter are to introduce the breadth of approaches for tumor classification, comment on their strengths and weaknesses, and illustrate their applications in epidemiological research Given the immensity of this subject, and its rapid evolution, this chapter empha-sizes histopathologic and molecular classifications of cancer and their relationships with complementary approaches (Hoadley et al., 2014) Specific examples are provided to demonstrate principles Detailed discussions of individual organ sites are presented throughout this volume

CHARACTERISTICS AND DEFINITIONS

OF NEOPLASMS (NEW GROWTHS), CANCERS, AND CANCER PRECURSORS

The synonymous terms “neoplasm” and “tumor” refer to new growths that do not respond physiologically to growth signals and disrupt nor-mal anatomy Tumors are further categorized as benign or malignant Benign neoplasms (and tumor- like lesions, such as hamartomas) are circumscribed, remain confined to the primary site of origin, and are usually curable with surgical removal In contrast, lesions desig-nated with the synonymous terms “malignant neoplasm” or “cancer” are defined by the potential to invade and destroy benign tissues and metastasize (spread) to other sites

Metastatic lesions cause death by damaging normal tissues, leading

to organ dysfunction, and by producing deleterious systemic effects, such as coagulation abnormalities or metabolic disturbances that dis-rupt homeostasis Even small metastases may prove lethal, if deposits impair critical physiological functions, such as occurs with metasta-ses to cardiorespiratory centers in the brainstem In some instances,

Định dạng
Số trang	613
Dung lượng	18,45 MB