LandmarksintheHistoryofCancerEpidemiology · characteristics, or they serve as putative surrogates...

Landmarks in the History of Cancer Epidemiology

Peter Greenwald and Barbara K. Dunn

Division of Cancer Prevention, National Cancer Institute, NIH, Bethesda, Maryland

Abstract

The application of epidemiology to cancer prevention isrelatively new, although observations of the potential causesof cancer have been reported for more than 2,000 years.Cancer was generally considered incurable until the late 19thcentury. Only with a refined understanding of the nature ofcancer and strategies for cancer treatment could a systematicapproach to cancer prevention emerge. The 20th century sawthe elucidation of clues to cancer causation from observedassociations with population exposures to tobacco, diet,environmental chemicals, and other exogenous factors. Withrepeated confirmation of such associations, researchersentertained for the first time the possibility that cancer, likemany of the infectious diseases of the time, might beprevented. By the mid-20th century, with antibiotics success-fully addressing the majority of infectious diseases and highblood pressure treatment beginning to affect the prevalenceof heart disease in a favorable direction, the focus of muchof epidemiology shifted to cancer. The early emphasis was onexploring, in greater depth, the environmental, dietary,hormonal, and other exogenous exposures for their potentialassociations with increased cancer risk. The first majorbreakthrough in identifying a modifiable cancer risk factorwas the documentation of an association between tobaccosmoking and lung cancer. During the past four decades,epidemiologic studies have generated population data iden-tifying risk factors for cancers at almost every body site, withmany cancers having multiple risk factors. The developmentof technologies to identify biological molecules has facilitatedthe incorporation of these molecular manifestations ofbiological variation into epidemiologic studies, as markersof exposure as well as putative surrogate markers of canceroutcome. This technological trend has, during the past twodecades, culminated in emphasis on the identification ofgenetic variants and their products as correlates of cancerrisk, in turn, creating opportunities to incorporate thediscipline of molecular/genetic epidemiology into the studyof cancer prevention. Epidemiology will undoubtedly continuecontributing to cancer prevention by using traditionalepidemiologic study designs to address broad candidate areasof interest, with molecular/genetic epidemiology investiga-tions honing in on promising areas to identify specific factorsthat can be modified with the goal of reducing risk. [CancerRes 2009;69(6):2151–62]

Introduction

Cancer epidemiology is the study of patterns and determinantsof cancer risk, outcome, and control in healthy or diseasedpopulations. Unlike the experimental nature of most clinicalresearch, much of epidemiology is inherently observational, thestudy of people with and without disease (1, 2). One ofepidemiology’s initial roles in supporting cancer research was thecollection of descriptive data on cancer incidence and mortalityto identify the health effects of cancer on society. As illustratedin Figs. 1 and 2, site-specific data on the mortality of women andmen in the United States from cancer has been collected andpublished since 1930. Understanding mortality trends for majorcancer types has focused research efforts and public policies inareas that can affect public health.

The search for associations between cancer outcomes andmeasurable exposures has evolved in response to increasinglysophisticated laboratory technologies to analyze ‘‘biomarkers’’ intissue specimens. Molecular epidemiology is the discipline that hasemerged to investigate such molecular factors. Biomarkers reflecteither the body’s response to specific environmental exposures(‘‘surrogates of exposure’’) or innate physiologic or geneticcharacteristics, or they serve as putative surrogates of clinicalcancer outcomes. Modern genetic epidemiology includesapproaches that focus on the molecular material of genes andchromosomes in constitutional DNA, reflecting inherited cancerpredisposition, as well as on tumor type–specific somatic geneticchanges.

Two key features of cancer epidemiology must be kept in mind ifits proper application to serving the general health and well beingis to be implemented. First, caution must be exercised in inferringcausality from measured associations because some factors occurin association with both a real causal factor and with the diseaseitself; and second, the ultimate goal is to improve the health ofpeople, preferably via cancer prevention; observational data alonedoes not accomplish this. Whenever possible, data generated fromepidemiologic studies, however sophisticated, must be integratedinto clinical research, in which experimental manipulation of theimplicated risk factors can be evaluated for its effect on canceroutcomes. In this article, we take a historical approach to surveyingkey areas of epidemiologic investigations into cancer risk, alwayswith an eye to the implications of study findings in terms of publichealth. Our goal is not to be totally comprehensive; rather, we wishto delve into selected epidemiologic research endeavors with acritical eye grounded in a modern scientific perspective. We hopein this way to elicit the relevance of historical study findings interms of their relevance to current and future epidemiologicinvestigation and especially in relation to their effect on publichealth.

Tobacco

The establishment of tobacco as a carcinogen is one of the signalachievements in epidemiology. Observational studies published as

Requests for reprints: Peter Greenwald, Division of Cancer Prevention, NationalCancer Institute, NIH, 6130 Executive Boulevard, Suite 2040, Bethesda, MD 20892-7309.Phone: 301-496-6616; Fax: 301-496-9931; E-mail: [email protected].

I2009 American Association for Cancer Research.doi:10.1158/0008-5472.CAN-09-0416

www.aacrjournals.org 2151 Cancer Res 2009; 69: (6). March 15, 2009

AACR Centennial Series

Research. on March 24, 2020. © 2009 American Association for Cancercancerres.aacrjournals.org Downloaded from

Published OnlineFirst March 10, 2009; DOI: 10.1158/0008-5472.CAN-09-0416





http://cancerres.aacrjournals.org/



early as the 1700s suggested an association of tobacco with nasalcancer among snuff workers (3), and mouth and lip cancers amongpipe smokers (4). The strongest evidence supporting tobaccoexposure as an etiologic factor for various cancers emerged from20th century epidemiologic studies. In 1912, Isaac Adler publishedhis observation that smoking tobacco may have been linked toreported increases in lung cancer cases in local hospitals (5). By the1950s, strong epidemiologic and experimental evidence indicatedthat cigarette tar and other tobacco constituents produced byburning the product and inhaling the smoke were responsible forcancers of the lung (6, 7) and that an unequivocal associationexisted between smoking and lung cancer, especially whenconsidering duration of smoking (6, 8). These seminal studiesfrom the work of Ernst Wynder and Richard Doll (6, 8) eventuallyled to the recognition by the U.S. Surgeon General and other publichealth institutions and medical societies of the dangers of tobaccouse in any form. This, in turn, led to important public healthprograms to discourage people, especially young people, frombeginning smoking and to encourage those who currently smoketo quit.

From 1950 to the present, epidemiology has played a critical rolein the search for the etiologic factors responsible for the dramaticincrease in lung cancer incidence and mortality during the 20thcentury. Animal studies identified cigarette smoke constituents,including polynuclear aromatic hydrocarbons, that have carcino-genic initiator-promoter activity (9). Eventually, mechanisticstudies elucidated the numerous carcinogenic genetic alterationsin relevant biological pathways: xenobiotic metabolism, control ofgenomic instability (including DNA repair mechanisms), cell cyclecheckpoints, apoptosis, telomere length, control of the microenvi-ronment by elements such as metalloproteinases, inflammation,and growth factors (10). Complete answers to the question ofcancer susceptibility would have to await the development of

molecular epidemiology. This field integrates molecular biologyand cancer genetics with classic epidemiologic study designs,allowing the identification of individuals or groups at increased riskor increased vulnerability to exogenous risk factors, i.e., those whomay benefit from targeted cancer prevention or treatmentstrategies (11).

One example of the variability of risk among populationsemerged from the long-held knowledge that association of CYP1enzymes with cancer risk due to tobacco exposure is due to theinvolvement of these enzymes in activating carcinogens such aspolynuclear aromatic hydrocarbons. Specific CYP1A1 polymor-phisms (MspI and Ile462Val polymorphisms) were associated withincreased lung cancer risk in the Japanese population, whereas onlythe CYP1A1-MspI polymorphism was strongly associated with riskin Western populations (10).

As early as the 1980s, the National Cancer Institute has directedantismoking efforts based on population data supplied byepidemiologists. The two largest efforts involved a clinical trial(Community Intervention Trial for Smoking Cessation) and a statelevel intervention program to address policy-driven initiatives topromote smoke-free environments, counter tobacco advertisingand promotion, limit tobacco access and availability, and increasetobacco prices through new excise taxes (American Stop SmokingIntervention Study). The Community Intervention Trial forSmoking Cessation and American Stop Smoking InterventionStudy showed that a multi-tiered approach could decrease theprevalence of smoking and increase the quit rate among currentsmokers (12–14).

Diet/Body Mass Index/Physical Activity

Although the associations between diet and disease have beenknown for centuries (e.g., limes to prevent scurvy in British sailors),

Figure 1. Landmarks in the history of cancer epidemiology (1900–present).

Cancer Research

Cancer Res 2009; 69: (6). March 15, 2009 2152 www.aacrjournals.org




not until the results of rigorous epidemiologic studies werereported in the 1970s and 1980s did compelling evidence for thebreadth of the association emerge (15–17). The application ofepidemiologic methods to diet/cancer studies is relatively newbecause of the difficulty in determining the effects of specificdietary factors on risks of chronic diseases, especially at theindividual level. Initially, diet/cancer epidemiology involved eco-logical studies of populations and suggested broad associations,such as that of red meat intake and increased colon cancer risk(18). Migrant studies provided a unique tool that, in the early daysof diet/cancer epidemiologic research, focused on contrastingdisease outcomes displayed by similar populations in different‘‘ecological’’ settings. An early migrant study comparing the rates ofgastric and colon cancers in Japanese migrating to the UnitedStates (19) indicated lower rates of gastric cancer and higher ratesof colon cancer the longer immigrants lived in the United States.Patterns similar to that for colon cancer were seen for cancers ofthe breast, uterine corpus, and ovary among women and forprostate cancer among men.

By the mid-1990s, evidence from numerous large populationstudies suggested a strong association between a high intake ofvegetables and fruits and reduced risks of cancers of the breast,prostate, colorectum, lung, stomach, and other sites (20). Also inthe 1990s, the National Cancer Institute initiated a community-based partnership based on epidemiologic and interventionstudies—the 5 A Day For Better Health Program (5 A Day)—toencourage increased vegetable and fruit intake among Americans;an evaluation in 2000 indicated a slow but steady increase invegetable and fruit consumption among American consumerssince the program’s inception (21).

Beyond overall diet, the association of cancer risk with specificdietary factors, including individual nutrients, has also beenevaluated. Conflicting results have led to controversies over theroles of such dietary factors. One of the longest-running

controversies involves the association of dietary fat with breastcancer: does an increase in the consumption of dietary fat increasethe risk of breast cancer, and does a low-fat diet confer lower risk?A multitude of studies (case-control, cohort, meta-analyses)spanning two decades attempted to resolve the issue withoutreaching conclusive agreement (22, 23). The dietary fat-breastcancer controversy exemplifies the challenges of basing cancerprevention recommendations solely on observational epidemio-logic studies; the need for clinical trials to test results suggested inobservational studies is critical to clarify the role of lifestyle choicesin cancer risk, especially in the face of conflicting epidemiologicfindings.

Occasionally, epidemiologic findings seem to be inconsistent withclinical trial results, but in fact, the apparent conflict at times seemsto have resulted from the failure to consider individual variability. Inthe case of the association between h-carotene and lung cancer,conflicting evidence in results from epidemiologic studies ( foodscontaining h-carotene reduce lung cancer risk) and prospectiveclinical trials such as the Alpha-Tocopherol, Beta-Carotene (ATBC)Cancer Prevention Study, which suggested that h-carotene insupplements promotes lung cancer risk among smokers, catalyzedintense discussions to discern reasons for the inconsistency (24). Apossible explanation for this apparent conflict emerged frompharmacogenomic studies conducted post-trial in the ATBC cohort.Inmost cases, aGSTM1 null genotype is associated with a higher riskof lung cancer when compared to a GSTM1-present genotype,regardless of treatment or years of smoking. Furthermore, thepositive association between smoking duration and lung canceris greater among GSTM1-null individuals than those with aGSTM1-positive genotype. h-carotene supplementation confers anincreased risk of lung cancer at every duration of smoking with oneexception: in individuals with a GSTM1-null genotype. Although allGSTM1-null individuals who fall into the longest smoking durationcategory have a significantly elevated risk of lung cancer compared

Figure 1 Continued.






to GSTM1-positives, this risk is somewhat lower in those who havehad h-carotene supplementation (25). In addition to identifying asubgroup that might benefit from an intervention, this studyexemplifies how clues from classic epidemiology can be obscured inthe absence of in-depth characterization of individuals likely toexperience a benefit (or harm). Clearly, one size doesn’t fit all. Thesuccess of dietary components as interventionsmust be examined interms of the gene-environment interactions involved in modulatingthe preventive outcomes in epidemiologic studies and in clinicaltrials. The ATBC prospective randomized controlled trial of twopromising nutrients exemplifies the limitations of the priorhypothesis-generating observational data and reinforces the impor-tance of basing dietary recommendations on prospective clinicaltrials, whenever possible.

An important secondary finding in ATBC was the reduction inprostate cancer incidence in men receiving a-tocopherol. Asecondary end point from another randomized controlled trial,which proved negative for its primary end point of testing seleniumfor the prevention of skin cancer, was a reduction in prostatecancers in men receiving selenium (26). Together, these secondaryoutcomes from two prevention trials laid the foundation for theNational Cancer Institute’s second phase 3 trial for the preventionof prostate cancer. The Selenium and Vitamin E Cancer PreventionTrial (SELECT), through its factorial design, tested each nutrientseparately and together in comparison to placebo. Following arecent independent review of study data, it was announced that

neither supplement, taken alone or together, prevented prostatecancer and that the supplements are unlikely ever to produce the25% disease reduction on which the study was designed.Furthermore, a small, statistically nonsignificant increase in casesof prostate cancer was observed in men older than age 50 takingonly vitamin E, and a similar statistically nonsignificant increase incases of adult onset diabetes occurred in men taking only selenium(27). This negative result of SELECT, by not confirming secondaryend points in the hypothesis-generating clinical trials (26, 28), onceagain reinforces the importance of designing prospective clinicaltrials to test putative preventive agents specifically for their efficacyagainst given cancer end points.

In addition to confirming or refuting the benefits of specificbioactive foods during the past decade, epidemiologic and clinicalstudies have yielded evidence linking diet, body mass index (BMI)and obesity, and physical activity to cancer risk. In a recentsystematic review and meta-analysis of 221 data sets with 282,137incident cancer cases from prospective observational studies, BMIwas linked with an increase in the risks of common and lesscommon cancers, with associations differing between the sexes andamong ethnic groups (29). Epidemiologic studies indicate thebenefits of physical activity as an intervention for obesity, a knowncancer risk factor, with the strongest evidence of an inverserelationship being for breast cancer, especially among postmeno-pausal women (30). Intervention strategies that address the effectof each lifestyle factor have been suggested by cancer prevention

U.S.

Figure 2. Cancer death rates for women in the United States (1930–2004).

Cancer Research





research. The second report of the World Cancer Research Fund/American Institute for Cancer Research developed lifestylerecommendations based on research in each lifestyle domain (seeTable 1; ref. 31). The report used systematic literature reviews ofepidemiologic and experimental studies, with assessments of thestrengths and weaknesses of each study, to determine arecommendation based on the best available evidence in 20lifestyle domains. In the same report, the World Cancer ResearchFund/American Institute for Cancer Research issued recommen-dations on specific bioactive foods shown in epidemiologic andclinical studies to reduce cancer risk at specific sites (see Table 2).Each of the recommendations was based on the best availablescientific evidence.

Although relatively new, studies of diet, BMI, and physicalactivity in relation to cancer prevention have yielded importantfindings showing promise for reducing the effect of cancer onsociety by encouraging healthy lifestyle choices. Increased researchin these areas, especially in molecular epidemiologic approaches tobasic nutritional science, should supply sound scientific data toassist policymakers in developing public health strategies.

Reproductive Factors and Cancer Risk

The idea that hormones that control normal growth in targetorgans can, under the wrong circumstances, contribute toneoplastic transformation has been entertained for a long time,and is most evident in hormone-related cancers: breast, endome-trium, ovary, and prostate (32). Estrogen, the best studied hormoneof those implicated in carcinogenesis, is involved in severaldifferent cancer types.

The carcinogenic properties of estrogen are evident in Beatson’s1896 observation that oophorectomy is associated with theregression of breast tumors (41). The development of vaginalclear-cell adenocarcinoma following intrauterine exposure todiethylstilbestrol in daughters whose mothers were treated withthis drug to prevent spontaneous abortion and premature deliverywas a key early epidemiologic observation of the carcinogeniceffect of estrogens (33, 34). The classic epidemiologic approach oftracking the cancers on a case-by-case basis, accompanied bydetailed questioning, led to an initial report of six cases. A

subsequent series of case-control studies ultimately linked this rarecancer to the synthetic estrogen (35).

Later investigations of the association of estrogen with a varietyof chronic diseases offer an interesting twist to our understanding ofthe importance of study design for exposing true relationshipsbetween exposures and disease outcomes. Estrogen has long beenviewed as a carcinogen in the breast, in part from associationsnoted in epidemiologic studies devoted to estimating the correla-tion between endogenous serum estrogen levels and breast cancerrisk (36). Overall, case-control and prospective observational studiesshow that higher levels of endogenous serum estrogens are asso-ciated with increased breast cancer risk (36, 37). The actual estrogenexposure of the breast is presumed to be reflected in severalsurrogate markers of exposure (e.g., early age at menarche, older ageat menopause, nulliparity, older age at first live birth, andpostmenopausal obesity) for which an association with breast

Table 1. 2007 WCRF/AICR dietary recommendations

Recommendations for cancer prevention

1. Be as lean as possible without becoming underweight

2. Be physically active for at least 30 min every day

3. Avoid sugary drinks. Limit consumption of energy-dense foods (particularly processed foods high in added sugar, or low in fiber, or high in fat)4. Eat more of a variety of vegetables, fruits, whole grains and legumes such as beans

5. Limit consumption of red meats (such as beef, pork and lamb) and avoid processed meats

6. If consumed at all, limit alcoholic drinks to two a day for men and one a day for women7. Limit consumption of salty foods and foods processed with salt (sodium)

8. Do not use supplements to protect against cancer

Special population recommendations

9. It is best for mothers to breastfeed exclusively for up to 6 months and then add other liquids and foods

10. After treatment, cancer survivors should follow the recommendations for cancer prevention

NOTE: Taken from the World Cancer Research Fund/American Institute for Cancer Research. Food, Nutrition, Physical Activity, and the Prevention ofCancer: a Global Perspective. Washington, DC: AICR, 2007.

Table 2. WCRF/AICR 2007 recommendations for cancer-preventive bioactive foods in relation to specific tissuesites

Foods Tissue site

Non-starchy vegetables Mouth, pharynx, larynx,

esophagus, stomach

Allium vegetables Stomach

Garlic ColorectumFruits Mouth, pharynx, larynx,

esophagus, lung, stomach

Foods containing folate Pancreas

Foods containing carotenoids Mouth, pharynx, larynx, lungFoods containing h-carotene Esophagus

Foods containing lycopene Prostate

Foods containing vitamin C EsophagusFoods containing selenium Prostate

NOTE: Taken from theWorld Cancer Research Fund/American Institute

for Cancer Research. Food, Nutrition, Physical Activity, and the

Prevention of Cancer: a Global Perspective. Washington, DC: AICR, 2007.






cancer risk has been noted (38). These ‘‘risk factors’’ have beenincorporated into clinically applicable models to assess breastcancer risk level (39). Although observational studies of postmen-opausal use of exogenous estrogen (hormone replacement therapy)have not consistently reported an association, current and longerduration of estrogen use is associated with increased breast cancerrisk (40). This epidemiologically demonstrated association, togetherwith early observations that oophorectomy is associated with breasttumor regression (41), stimulated the development of antiestrogenicagents—selective estrogen receptor modulators and aromataseinhibitors—for breast cancer treatment and prevention (42, 43).

In contrast to breast cancer, other estrogen-responsive chronicdiseases, such as coronary heart disease (CHD), have been shownin observational studies to exhibit an inverse relationship withestrogen levels. Premenopausal women have higher levels ofestrogen and a lower frequency of CHD than age-matched men(44), with the risk of CHD increasing dramatically after menopauseas estrogen levels drop precipitously. Epidemiologic evidencesuggested that higher endogenous estrogen levels or exogenousestrogen intake are associated with lower rates of CHD but higherrisk of breast cancer. Thus, in the Nurses’ Health Study, aprospective observational cohort study of 70,533 postmenopausalwomen, participants with no pre-existing heart disease who weretaking oral conjugated estrogen, had a decreased risk of majorcoronary events (45). Based on its ability to decrease levels ofcholesterol, a presumed surrogate biomarker of CHD, estrogen

activity has been interpreted as ameliorating CHD risk. This benefitfrom estrogen use observed in epidemiologic studies with regard toheart disease, together with the benefit to multiple other organsystems (e.g., bone, skin, brain, and vaginal epithelium), ledprominent clinical researchers to declare that estrogen was a keyantiaging drug and to encourage all menopausal women to take it,despite the absence of a prospective randomized clinical trial ofthis hormone for this purpose (46).

The ‘‘great estrogen conundrum’’ (47) pits the antithetical out-comes involving breast cancer and cardiovascular disease againsteach other. As the level of scientific evidence for the two diseaseoutcomes advanced from observational to experimental, widelyheld views of estrogen’s overall benefits underwent major revision.The observational findings favoring estrogen use for the heartwere not substantiated in prospective trials, such as the Heart andEstrogen/progestin Replacement Study (48), although the general-izability of this finding was limited by the fact that the Heart andEstrogen/progestin Replacement Study cohort consisted of womenwith pre-existing CHD. Publication of the Heart and Estrogen/progestin Replacement Study in 1998 was followed by a moderatedecline in hormone replacement therapy use (Fig. 3; ref. 49). TheWomen’s Health Initiative, a well-designed randomized controlledtrial using hormone replacement therapy as a prospectiveintervention in healthy postmenopausal women (50), clarified theassociation between estrogen and these two chronic diseaseoutcomes. The epidemiologic data regarding increased risk of

Figure 3. Cancer death rates for men in the United States (1930–2004).

Cancer Research





breast cancer were borne out in the Women’s Health Initiativeestrogen plus progestin trial. In contrast, the observational datapertaining to CHD were contradicted (50), indicating that much ofthe apparent benefit seen in the epidemiologic studies was likelydue to biases inherent in their weaker study design (46). Together,these two adverse disease outcomes, shown at a higher level ofevidence in the Women’s Health Initiative than in the earlier, obser-vational studies, had downstream ramifications—massive cessationof hormone replacement use by postmenopausal women, exceedingthat following publication of the Heart and Estrogen/progestinReplacement Study (Fig. 3; 49). The decrease in breast cancerincidence noted 1 year later in 2003 has been attributed to thiswidespread change in medical practice (51), vindicating the imple-mentation of this critically important trial, and emphasizing thenecessity of implementing the highest level of study possible inaddressing epidemiologic questions (46). The results of theWomen’sHealth Initiative emphasize the importance of adhering to the prin-ciples of evidence-based medicine in evaluating inputs, both risk-inducing and risk-reducing, in relation to disease outcomes (Fig. 4).

Prostate cancer, the most common adult cancer in men in theUnited States, is hormone-dependent, with its growth dependingon testosterone (T) and its more active metabolite dihydrotestos-terone (32). This dependency is evidenced in the successful use ofandrogen ablation/blockade for palliation of advanced disease.Overall, epidemiologic studies by investigators such as Ronald Ross

and Brian Henderson support an association between hormonelevels and prostate cancer risk, despite difficulties encountered inestablishing such a link (52). In contrast, a well-conductedprospective case-control study nested in the Physicians’ HealthStudy showed that when levels of hormone and sex hormone-binding globulin were adjusted simultaneously, a strong trend ofincreasing prostate cancer risk was observed with increasingtestosterone levels. Such epidemiologic evidence for testosteroneinvolvement in prostate cancer risk, coupled with the availability ofthe drug finasteride, which inhibits 5a-reductase, the enzyme thatcatalyzes conversion of T to dihydrotestosterone, led to design ofthe Prostate Cancer Prevention Trial (PCPT), first large phase 3trial of its kind (53); this trial showed a reduction in prevalentprostate cancers in men receiving the 5a-reductase inhibitor.

Occupation/Environment

Astute physician observations underlay the recognition ofpotential associations of occupational exposures to natural andman-made agents with cancer. Such ‘‘occupational cancers’’ wereidentified through observational epidemiologic methods as early as1700, when Bernardini Ramazzini, an Italian physician, publishedDe Morbis Artificum Diatriba (Diseases of Workers), a comprehen-sive study of the hazards of exposure to chemicals, dusts, metals,and other agents among workers in 52 occupations (54). This is

Figure 4. The effects of theHeart and Estrogen/progestin Replacement Study and theWomen’s Health Initiative on hormone replacement use by postmenopausal women.






considered the first study of what would become occupationalmedicine. A more focused epidemiologic study was published in1775 by Percival Pott of Saint Bartholomew’s Hospital in London,who described cancer of the scrotum in chimney sweeps, caused byconstant exposure to soot (55).

Links between modern occupational exposures and cancercontinue to be suggested by epidemiologic studies. In some cases,notably the links between asbestos and mesothelioma (56) andbetween vinyl chloride monomer and angiosarcoma of the liver(57), the rarity of the cancer facilitated the recognition of theassociation. Links between occupational exposures and morecommon cancers—such as between aromatic amines and bladdercancer (58) or between asbestos exposure and lung cancer (59)—were less obvious and were only brought to public attentionthrough the efforts of perceptive clinicians and epidemiologists.

Quantitative assessment of exposure is difficult, posing chal-lenges to epidemiologic investigation of occupation/cancer rela-tionships. Whereas individuals usually can describe their smokinghistories and eating habits reasonably well, they are often relativelyignorant about their work exposures. Moreover, because of longlatency periods between exposure and cancer diagnosis andbecause working conditions change over time, data on currentworkplace exposures may not reflect those experienced by pastworkers. Epidemiologists often have had to rely on proxy measuresof exposure in studies of occupational cancer. For example, theepidemiologic studies which showed that physicians who special-ized in radiology prior to 1950 were at increased risk of leukemiabecause of high ionizing radiation exposure relied primarily onknowledge of typical radiation protection practices in differenttime periods, rather than on specifics about the exposure ofparticular individuals (60). Difficulty in obtaining accurate exposuredata remains a problem today, hampering recent efforts such asthe determination of whether occupational exposures to pesticides(61) and electromagnetic fields (62) influence cancer risk.

Occupational cancers represent only f5% of all cancers indeveloped countries (55), and the number of cases is expected todecrease in future years because of improvements in industrialhygiene. Nevertheless, occupational cancers are significant becausepeople will continue to suffer and die from them during the nextfew decades as a result of past exposures, and they are more readilypreventable than cancers caused by tobacco, diet, or other lifestylefactors.

Molecular epidemiology (see below) has increased our under-standing of genetic differences in susceptibility to occupational andenvironmental carcinogens. For example, among men at high riskof bladder cancer because of tobacco smoking and occupationalexposure to aromatic amines, susceptibility to this cancer is influ-enced by polymorphisms in the genes for glutathione-S-transferaseM1 (GSTM1), glutathione-S-transferase T1 (GSTT1), and N-acetyltransferase 2 (NAT2 ; ref. 63). Polymorphisms in GSTM1 and themanganese superoxide dismutase (MnSOD/SOD2) gene influencemesothelioma risk in asbestos-exposed persons (64). Differences inmyeloperoxide (MPO) genotypes modify the risk of lung cancer inasbestos-exposed workers (65). Polymorphisms in GSTT1 and in thegene for cytochrome P450 2E1 (CYP2E1) have been associated withvariations in mutagenic risk in workers exposed to vinyl chloride(66). Among never-smokers, individuals who are homozygous forthe GSTM1-null allele have been found to be at increased risk ofdeveloping lung cancer as a result of exposure to environmentaltobacco smoke (67). Exposure to tobacco smoke through directsmoking was covered in a previous section of this review. Individual

differences in factors other than genotype may also influencesusceptibility to occupational and environmental carcinogens. Forexample, susceptibility to lung cancer associated with asbestosexposure is greatly increased by cigarette smoking (68).

One of the longest-term studies of the environmental effects oncancer is the Life Span Study of atomic bomb survivors. The LifeSpan Study was established to study the long-term carcinogeniceffects of radiation on the survivors of the bombings of Hiroshimaand Nagasaki. The cohort consists of 120,321 registered residents ofHiroshima and Nagasaki who were within 2.5 km (f54,000survivors), between 2.5 and 10 km (f40,000 survivors), or morethan 10 km (26,580 survivors) from the hypocenter at the time ofbombing. Analysis of diagnoses of solid cancer between 1958 and1998 found 17,448 first primary cancers, and f11% wereassociated with atomic bomb radiation exposure (69). Radiation-associated increases in cancer rates were also found to persistthroughout the survivors’ lifetimes. Forty-eight percent of thecancers among survivors receiving doses of at least 1 Gy wereattributable to radiation exposure; f18% of the excess cancersoccurred among individuals in the dose range of 5 to 200 mGy.The Life Span Study provides the best quantitative risk estimates ofthe carcinogenic effects of radiation exposure in humans and isa major source of epidemiologic data used for radiation riskassessment and establishment of radiation protection guidelines.

Infectious Agents

The idea that infectious agents might cause some cancers hasexisted for more than a century, but not until the 1950s didresearchers seriously begin to evaluate such associations (70, 71). Atthe global level, the estimated percentage of cancers associated withinfections is between 10% and 20%, with the implicated organismsranging from viruses and bacteria to parasites, such as Schistosomahaematobium , a risk factor for bladder cancer (72). By the 1980s, thedevelopment of new methodologies, including PCR, facilitateddocumentation of associations between the presence of certainviruses and bacteria in tumor tissue and reproductive and gastriccancers, respectively (71). Using this knowledge, observationalstudies, such as the National Health and Nutrition ExaminationSurvey, included molecular epidemiologic studies that helped definethe breadth of infection in the U.S. population. For example,National Health and Nutrition Examination Survey data showedthat almost one-half of women 14 to 44 years of age were infectedwith the human papillomavirus (HPV), which is responsible forpractically all cases of cervical cancer (and possibly squamous cellskin and other cancers; refs. 72, 73). Population studies alsosuggested an association between hepatitis B and liver cancer,especially in the presence of aflatoxins in the diet (74). In addition, apopulation study indicated a strong link between Helicobacter pyloriinfection and family history of gastric cancer (75).

Some rarer cancers also have been associated with infectiousagents. EBV, a member of the herpesvirus family, infects >95% of allhumans and can cause mononucleosis in adolescents. EBV wasisolated from Burkitt lymphoma, an aggressive B-cell lymphoma;although rare, Burkitt lymphoma accounts for nearly half of allchildhood cancers in equatorial Africa. EBV has also been detectedin nasopharyngeal carcinoma, non-Hodgkin lymphoma in AIDSpatients, and T-cell and Hodgkin lymphomas. EBV is hypothesizedto enhance genomic instability, leading to increased likelihood ofthe c-myc translocation characteristic of Burkitt lymphoma (76).In nasopharyngeal carcinoma, EBV is associated with alterations

Cancer Research





in expression of the Bcl-2 and TP53 genes (77). Another member ofthe herpesvirus family, Kaposi sarcoma-associated herpesvirus, isconsistently detected in all forms of Kaposi sarcoma and in aspecific subset of lymphoproliferative disorders (78).

The human T-cell leukemia virus-1 is the etiologic agent of adultT-cell leukemia-lymphoma. Infection with human T-cell leukemiavirus-1 is endemic in sites in southern Japan, the Caribbean, Centraland South America, and parts of Africa and the Middle East. Thevirus is transmitted by breast-feeding, sexual activity, or exposureto infected blood. The viral gene tax encodes a transcriptionaltransactivator that immortalizes CD4+ lymphocytes, renders themresistant to apoptosis, and also induces angiogenesis (79).

Cancer prevention strategies that developed in response toepidemiologic and clinical findings on infectious agents includelifestyle changes, such as reducing exposure to infectious agents, aswell as immunization by vaccine. Vaccines for cancer preventionhave received increased attention since 1981, when the hepatitis Bvaccine was developed to prevent hepatitis and also prevented livercancer (80). The etiologic association between HPV and cervicalcancer had been established by the research of Harald zur Hausendating back to the 1970s. The importance of this discovery, evi-denced in zur Hausen’s receipt of the 2008 Nobel prize in Physiologyor Medicine, laid the foundation for work such as that of DouglasLowy and John Schiller, who developed an HPV vaccine consisting ofthemajor structural viral protein L1 (virus-like particles; ref. 81). Onesuch HPV vaccine has been tested in clinical trials, is Food and DrugAdministration–approved, and is being distributed among thoseat risk for cervical cancer (82, 83). Presently, a vaccine againstH. pylori (84) is in early-phase testing in small numbers of people.Early results from this controlled, single-blind phase 1 clinical trial in57 volunteers suggest that the vaccine is safe and potentially useful;the authors recommended further clinical study (84).

Genetic Epidemiology

A century ago, the prevailing view was that cancer was aninevitable result of an individual’s genetic destiny. By the 1960s,epidemiologic evidence suggested that up to 90% of cancers werestrongly associated with avoidable factors, i.e., factors other thaninherited predisposition. Recent years have seen a resurgence ofresearch emphasis on genetic contributions. The modern fieldof genetic epidemiology evolved out of the recognition that somecancers cluster within families (85), implicating an inheritedcomponent in human carcinogenesis and suggesting that a similargenetic contribution might occur in nonfamilial, sporadic cancers.Parallel developments in high-throughput genotyping technologiesand accompanying analytic computational methods enabledresearchers to identify the actual physical locations and identitiesof implicated genetic factors, rather than simply establishing theexistence of genetic contributions to carcinogenesis (86).

Traditional Studies Documenting GeneticContributions to CarcinogenesisTwin studies. An early laboratory for studying genetic

contributions to complex human traits, including diseases suchas cancer, was recognized in identical (monozygotic) twins. Giventhe shared environment in twins, differences in phenotypicconcordance between monozygotic and dizygotic twins wereattributed to genetic factors (87). A key study analyzing data from11 cancer sites in 44,788 Scandinavian twin pairs revealed only amoderate increase in overall risk of the same cancer in a twin of a

person with cancer, especially for breast, prostate, colorectal, lung,and stomach cancers (88). The effect of heritability (that portion ofrisk that is due to inherited genes) was statistically significant onlyfor prostate (42% of risk), colorectal (35%), and breast (27%) cancer.The authors concluded that the environment was the ‘‘overwhelm-ing contributor to the causation of cancer’’ in these twinpopulations, precipitating considerable debate (87, 89). Of note,the notion that heritability fails to explain a large proportion ofcancer incidence and mortality is not new, having been suggestedin multiple earlier investigations (90).

The strong association of lung cancer with tobacco presents anunusual challenge to the disentanglement of genetic contributionsfrom environmental contributions. In the Scandinavian twin study(88), lung cancer risk was increased in the twin of an individualwith lung cancer, although the heritability was limited. Asystematic review of case-control, cohort, and twin studiesdesigned to assess the role of family history in lung carcinogenesisled to a similar conclusion of only a minor genetic contribution(91). The assumption that the environmental factor, smoking, issimilarly concordant in monozygotic and dizygotic twins is notsupported by the data, with a stronger concordance formonozygotic twins. Furthermore, men show similar lung cancerconcordance ratios for monozygotic and dizygotic twins, consistentwith a strong shared environmental influence.

Although twin studies offer insight into genetic contributionsto cancer, the rarity of twins limits the effect of twin studies onresearch, necessitating reliance on alternative analytic approaches.Family-based association studies. The aggregation of specific

cancers or constellations of cancers in families is a necessary butinsufficient criterion for inferring genetic contribution to disease(87). Astute clinical observations of ‘‘cancer’’ families led to theidentification of genetic syndromes such as the Li-Fraumeni (92) andLynch (hereditary nonpolyposis colorectal cancer) syndromes (85).

Modern Molecular Genetic Epidemiologic StudiesGenome-wide linkage analysis has successfully superimposed

molecular biological tools on the pedigree approach to map theactual genes responsible for cancers with monogenic, ‘‘Mendelian,’’inheritance patterns (93). Analysis of families with multiple,usually early onset cancer cases by linkage approaches such aspositional cloning has identified genes that are mutated inaffected individuals. BRCA1 and BRCA2 in breast and ovariancancer syndrome, MLH1 and MSH2 in hereditary nonpolyposiscolorectal cancer, and APC in familial adenomatous polyposiswere identified in this manner. Such highly penetrant genevariants are implicated in only a small percentage of cancers.Most cancers are complex, quantitative traits reflecting theinterplay of common variations in multiple genes (epistasis), eachconferring only minor increases in risk, as well as interactionswith environmental factors (93, 94).

Candidate-gene studies are hypothesis-based, involving theselection of genes and polymorphisms based on criteria such asgene function or location in a region of linkage (93). Overall, this‘‘gene-by-gene functional candidate gene approach’’ (94) has hadlimited success; with rare exceptions, including those discussedabove, few polymorphisms show consistent associations withcancer from one study to another (95). An example is a recentbreast cancer candidate gene study that examined variants in 11genes encoding proteins in the estrogen metabolic pathway (96).Although single factor analyses suggested a possible associationof CYP1B1_1294_GG with risk, no significant associations were






observed by multifactor analyses. Candidate gene approachesto lung cancer risk must necessarily separate genetic fromenvironmental (tobacco) contributions (97). Thus, an XRCC1(DNA repair gene) polymorphism shows significant associationwith risk overall but especially in nonsmokers, whereas heavysmokers show inverse risk (97). In addition, a recent study of gene-environment interactions identified polymorphisms in the nicotinicacetylcholine receptor gene cluster on chromosome 15q24 witheffects on smoking quantity, nicotine dependence, and the riskof lung cancer and peripheral artery disease in populations ofEuropean descent (98).

The genome-wide association study (GWAS) approach isagnostic, requiring no prior knowledge of position or function ofa marker (91, 99). Putative cancer-associated loci are localized byexploiting their tight linkage disequilibrium with known singlenucleotide polymorphisms that are densely distributed throughoutthe genome and serve as surrogates for the polymorphism ofinterest (100). To economize on cost and the large sample sizerequired for statistical validation of complex disease association,analysis is restricted to a subset of ‘‘tag single nucleotidepolymorphisms’’ that capture most of the linkage in a region(100). Other cost-conserving measures include multistaged designs(sequential GWAS analyses; refs. 99–102), minor allele frequenciesof z0.1, and odds ratios for effect sizes of z1.3 (94). Still, a samplesize of z1,000 cases is required, necessitating collaboration amongmultiple individual projects (100).

Supporting evidence for a strong genetic component in prostatecancer, GWAS implicated the 8q24 locus in this disease, as well ascolorectal, breast, and ovarian cancers (102, 103). GWAS of menaged V60 years at diagnosis and with a family history revealedseven loci on multiple chromosomes associated with prostatecancer, two of which (8q24 and 17q) had previously been reported(99). A GWAS study of breast cancer in high-risk Ashkenazi Jewishwomen who were negative for a BRCA mutation showed thestrongest disease association with 6q22.33, but also a significantassociation with FGFR2 (chromosome 10q25.3-q26; ref. 104). ABreast Cancer Association Consortium study showed strong,consistent associations of FGFR2 and three other genes (TNRC9,MAP3K1 , and LSP1) with breast cancer (101). In addition, lungcancer susceptibility was mapped to 15q24-25.1 by GWAS (7).

Thus, GWAS has the potential to localize cancer risk to specificchromosomal regions. Its central assumption, however, that amarker in tight linkage disequilibrium with a polymorphism thatdirectly influences cancer risk will always be detectable in asso-ciation with disease at an appropriate sample size, has beenchallenged (105, 106). GWAS is not self-sufficient, and embracingGWAS should not be taken as a rejection of earlier strategies. GWASand candidate-gene investigations should proceed side-by-side.The scientific paradigm of confirmation and validation mustapply to GWAS, with ultimate confirmation coming from alternatetechnologies. The value of GWAS will be to direct investigations topromising sites containing plausibly causative genes for candidate-gene investigations, more refined regional mapping, and detailedfunctional analyses.

Conclusion: Implications for Cancer Prevention

Epidemiologic research has successfully identified numerousputative and some confirmed causal factors in specific cancers.Where the strength of association is indisputable, as betweentobacco and lung cancer, observational knowledge has stimulatedefforts at control of the implicated factor. For this discipline to have ameaningful effect on its sister field of cancer prevention, decisionsregarding the design of epidemiologic studies, including theselection of putative risk factors for evaluation, need to be madeprospectively with an eye to the implications of study findings forsubsequent clinical and public health research evaluating potentialinterventions.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

Received 9/29/2008; revised 12/8/2008; accepted 2/3/2009; published OnlineFirst3/10/09.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

The authors are grateful to Dr. Kenneth Buetow and Darrell Anderson for generouscontributions to the scientific and technical preparation of this article.

References

1. Potter JD. At the interfaces of epidemiology, geneticsand genomics. Nat Rev Genet 2001;2:142–7.2. Franco EL, Correa P, Santella RM, Wu X, Goodman SN,Petersen GM. Role and limitations of epidemiology inestablishing a causal association. Semin Cancer Biol2004;14:413–26.3. Hill J. Cautions against the immoderate use of snuff.Founded on the known qualities of the tobacco plantand the effects it must produce when this way takeninto the body and enforced by instances of persons whohave perished miserably of diseases, occasioned, orrendered incurable by its use. Second ed. London: R.Baldwin and J. Jackson; 1761.4. National Cancer Institute. Changes in cigarette-related disease risks and their implication for preven-tion and control. Smoking And Tobacco Control Series,National Institutes of Health, National Cancer Institute;1997.5. Adler I. Primary malignant growths of the lungs andbronchi; a pathological and clinical study. London:Longmans, Green; 1912.

6. Wynder EL, Graham EA. Tobacco smoking as apossible etiologic factor in bronchiogenic carcinoma; astudy of 684 proved cases. J Am Med Assoc 1950;143:329–36.7. Herbst RS, Heymach JV, Lippman SM. Lung cancer. NEngl J Med 2008;359:1367–80.8. Doll R, Hill AB. Smoking and carcinoma ofthe lung; preliminary report. Br Med J 1950;2:739–48.9. U.S. Public Health Service. The health consequences ofsmoking. A report of the Surgeon General. Bethesda,MD: United States. Public Health Service, Office of theSurgeon General; 1972. Report No.: DHEW PublicationNo. (HSM) 72–7516.10. Wu X, Zhao H, Suk R, Christiani DC. Geneticsusceptibility to tobacco-related cancer. Oncogene2004;23:6500–23.11. Perera FP, Weinstein IB. Molecular epidemiology andcarcinogen-DNA adduct detection: new approaches tostudies of human cancer causation. J Chronic Dis 1982;35:581–600.12. The COMMIT Research Group. Community Inter-vention Trial for Smoking Cessation (COMMIT): I.

Cohort results from a four-year community interven-tion. Am J Public Health 1995;85:183–92.13. The COMMIT Research Group. Community inter-vention trial for smoking cessation (COMMIT): II.Changes in adult cigarette smoking prevalence. Am JPublic Health 1995;85:193–200.14. Stillman FA, Hartman AM, Graubard BI, Gilpin EA,Murray DM, Gibson JT. Evaluation of the American StopSmoking Intervention Study (ASSIST): a report ofoutcomes. J Natl Cancer Inst 2003;95:1681–91.15. U.S. Department of Health and Human Services.Food, nutrition, physical activity, and the prevention ofcancer: a global perspective. Washington DC: U.S.Government. Printing Office; 1988. Report No.: DHHSPub. No. (PHS) 88–50210.16. National Academy of Sciences. Committee on Dietand Health, Food and Nutrition Board, Commission onLife Sciences, National Research Council. Diet andhealth: implications for reducing chronic disease risk.Washington (DC): National Academy Press; 1989.17. Doll R, Peto R. The causes of cancer: quantitativeestimates of avoidable risks of cancer in the UnitedStates today. J Natl Cancer Inst 1981;66:1191–308.

Cancer Research





18. Willett WC. Epidemiologic studies of diet and cancer.Med Oncol Tumor Pharmacother 1990;7:93–7.19. Dunn JE, Jr. Breast cancer among American Japanesein the San Francisco Bay area. Natl Cancer Inst Monogr1977;47:157–60.20. World Cancer Research Fund/American Institute forCancer Research. Food, nutrition, and the prevention ofcancer: a global perspective. Washington (DC): Ameri-can Institute for Cancer Research; 1997.21. Potter J, Finnegan J, Guinard JX, et al. 5 A Day forBetter Health Program evaluation report. Bethesda(MD): National Institutes of Health, National CancerInstitute; 2000. Report No.: (NIH Publication No. 01-4904).22. Greenwald P. Role of dietary fat in the causation ofbreast cancer: point. Cancer Epidemiol Biomarkers Prev1999;8:3–7.23. Hunter DJ. Role of dietary fat in the causation ofbreast cancer: counterpoint. Cancer Epidemiol Bio-markers Prev 1999;8:9–13.24. Greenwald P. h-Carotene and lung cancer: a lessonfor future chemoprevention investigations? J NatlCancer Inst 2003;95:E1.25. Woodson K, Stewart C, Barrett M, et al. Effect ofvitamin intervention on the relationship between GSTM1,smoking, and lung cancer risk among male smokers.Cancer Epidemiol Biomarkers Prev 1999;8:965–70.26. Clark LC, Combs GF, Jr., Turnbull BW, et al. Effects ofselenium supplementation for cancer prevention inpatients with carcinoma of the skin. A randomizedcontrolled trial. Nutritional Prevention of Cancer StudyGroup. JAMA 1996;276:1957–63.27. National Cancer Institute. Review of prostate cancerprevention study shows no benefit for use of seleniumand vitamin E supplements. National Cancer Instituteweb site press release 2008 October 27 [cited 2008 Nov11]; Available from: http://www.cancer.gov/newscenter/pressreleases/SELECTresults2008.28. The Alpha-Tocopherol, Beta-Carotene Cancer Pre-vention Study Group. The effect of vitamin E and hcarotene on the incidence of lung cancer and othercancers in male smokers. N Engl J Med 1994;330:1029–35.29. Renehan AG, Tyson M, Egger M, Heller RF, ZwahlenM. Body-mass index and incidence of cancer: asystematic review and meta-analysis of prospectiveobservational studies. Lancet 2008;371:569–78.30. Monninkhof EM, Elias SG, Vlems FA, et al. Physicalactivity and breast cancer: a systematic review. Epide-miology 2007;18:137–57.31. World Cancer Research Fund, American Institutefor Cancer Research. Food, nutrition, physical activity,and the prevention of cancer: a global perspective.Washington (DC): American Institute for CancerResearch; 2007.32. Henderson BE, Ross RK, Pike MC, Casagrande JT.Endogenous hormones as a major factor in humancancer. Cancer Res 1982;42:3232–9.33. Herbst AL, Ulfelder H, Poskanzer DC. Adenocarcino-ma of the vagina. Association of maternal stilbestroltherapy with tumor appearance in young women. NEngl J Med 1971;284:878–81.34. Greenwald P, Barlow JJ, Nasca PC, Burnett WS.Vaginal cancer after maternal treatment with syntheticestrogens. N Engl J Med 1971;285:390–2.35. Giusti RM, Iwamoto K, Hatch EE. Diethylstilbestrolrevisited: a review of the long-term health effects. AnnIntern Med 1995;122:778–88.36. Dunn BK, Wickerham DL, Ford LG. Prevention ofhormone-related cancers: breast cancer. J Clin Oncol2005;23:357–67.37. Key T, Appleby P, Barnes I, Reeves G. Endogenous sexhormones and breast cancer in postmenopausal wom-en: reanalysis of nine prospective studies. J Natl CancerInst 2002;94:606–16.38. MacMahon B, Cole P, Brown J. Etiology of humanbreast cancer: a review. J Natl Cancer Inst 1973;50:21–42.39. Gail MH, Brinton LA, Byar DP, et al. Projectingindividualized probabilities of developing breast cancerfor white females who are being examined annually. JNatl Cancer Inst 1989;81:1879–86.40. Nelson HD, Humphrey LL, Nygren P, Teutsch SM,Allan JD. Postmenopausal hormone replacement ther-apy: scientific review. JAMA 2002;288:872–81.

41. Beatson GT. On the treatment of inoperable cases ofcarcinoma of the mamma: suggestions for a newmethod of treatment, with illustrative cases. Lancet1896;148:104–7.42. Dunn BK, Ford LG. From adjuvant therapy tobreast cancer prevention: BCPT and STAR. Breast J2001;7:144–57.43. Dunn BK, Ryan a. Phase 3 trials of aromataseinhibitors for breast cancer prevention: following inthe path of the selective estrogen receptor modulators.In: Bradlow HL, Carruba G, editors. Steroid enzymes andcancer. New York (NY): Blackwell Publishing; 2009.44. Stangl V, Baumann G, Stangl K. Coronary atherogenicrisk factors in women. Eur Heart J 2002;23:1738–52.45. Grodstein F, Manson JE, Colditz GA, Willett WC,Speizer FE, Stampfer MJ. A prospective, observationalstudy of postmenopausal hormone therapy and primaryprevention of cardiovascular disease. Ann Intern Med2000;133:933–41.46. Barrett-Connor E. Heart disease in women. FertilSteril 1994;62:127–32S.47. Couzin J. Estrogen research. The great estrogenconundrum. Science 2003;302:1136–8.48. Hulley S, Grady D, Bush T, et al.; Heart and Estrogen/progestin Replacement Study (HERS) Research Group.Randomized trial of estrogen plus progestin forsecondary prevention of coronary heart disease inpostmenopausal women. JAMA 1998;280:605–13.49. Haas JS, Kaplan CP, Gerstenberger EP, Kerlikowske K.Changes in the use of postmenopausal hormone therapyafter the publication of clinical trial results. Ann InternMed 2004;140:184–8.50. Rossouw JE, Anderson GL, Prentice RL, et al. Risksand benefits of estrogen plus progestin in healthypostmenopausal women: principal results from theWomen’s Health Initiative randomized controlled trial.JAMA 2002;288:321–33.51. Ravdin PM, Cronin KA, Howlader N, et al. Thedecrease in breast-cancer incidence in 2003 in theUnited States. N Engl J Med 2007;356:1670–4.52. Ross RK, Pike MC, Coetzee GA, et al. Androgenmetabolism and prostate cancer: establishing amodel of genetic susceptibility. Cancer Res 1998;58:4497–504.53. Thompson IM, Klein EA, Lippman SM, Coltman CA,Djavan B. Prevention of prostate cancer with finasteride:US/European perspective. Eur Urol 2003;44:650–5.54. Ramazzini B. De morbis artificum diatriba [Diseasesof workers]. The Latin text of 1713 revised, withtranslation and notes by CW Wright. Chicago (IL):University of Chicago Press; 1940.55. Seffrin JR, Gerberding JL, Boyle P. The cancer atlas.American Cancer Society; 2007.56. McDonald JC, McDonald AD. The epidemiology ofmesothelioma in historical context. Eur Respir J 1996;9:1932–42.57. Creech JL, Jr., Johnson MN. Angiosarcoma of liver inthe manufacture of polyvinyl chloride. J Occup Med1974;16:150–1.58. Olfert SM, Felknor SA, Delclos GL. An updated reviewof the literature: risk factors for bladder cancer withfocus on occupational exposures. South Med J 2006;99:1256–63.59. Doll R. Mortality from lung cancer in asbestosworkers. Br J Ind Med 1955;12:81–6.60. Yoshinaga S, Mabuchi K, Sigurdson AJ, Doody MM,Ron E. Cancer risks among radiologists and radiologictechnologists: review of epidemiologic studies. Radiolo-gy 2004;233:313–21.61. Merhi M, Raynal H, Cahuzac E, Vinson F, Cravedi JP,Gamet-Payrastre L. Occupational exposure to pesticidesand risk of hematopoietic cancers: meta-analysis ofcase-control studies. Cancer Causes Control 2007;18:1209–26.62. McElroy JA, Egan KM, Titus-Ernstoff L, et al.Occupational exposure to electromagnetic field andbreast cancer risk in a large, population-based, case-control study in the United States. J Occup Environ Med2007;49:266–74.63. Hung RJ, Boffetta P, Brennan P, et al. GST, NAT,SULT1A1, CYP1B1 genetic polymorphisms, interac-tions with environmental exposures and bladdercancer risk in a high-risk population. Int J Cancer2004;110:598–604.

64. Landi S, Gemignani F, Neri M, et al. Polymorphismsof glutathione-S-transferase M1 and manganese super-oxide dismutase are associated with the risk ofmalignant pleural mesothelioma. Int J Cancer 2007;120:2739–43.65. Schabath MB, Spitz MR, Delclos GL, Gunn GB,Whitehead LW, Wu X. Association between asbestosexposure, cigarette smoking, myeloperoxidase (MPO)genotypes, and lung cancer risk. Am J Ind Med 2002;42:29–37.66. Schindler J, Li Y, Marion MJ, Paroly A, Brandt-RaufPW. The effect of genetic polymorphisms in the vinylchloride metabolic pathway on mutagenic risk. J HumGenet 2007;52:448–55.67. Bennett WP, Alavanja MC, Blomeke B, et al.Environmental tobacco smoke, genetic susceptibility,and risk of lung cancer in never-smoking women. J NatlCancer Inst 1999;91:2009–14.68. Case BW. Asbestos, smoking, and lung cancer:interaction and attribution. Occup Environ Med 2006;63:507–8.69. Preston DL, Ron E, Tokuoka S, et al. Solid cancerincidence in atomic bomb survivors: 1958-1998. RadiatRes 2007;168:1–64.70. de Harven E. Viral etiology of human cancer: ahistorical perspective. Haematologica 1999;84:385–9.71. Javier RT, Butel JS. The history of tumor virology.Cancer Res 2008;68:7693–706.72. zur Hausen H. Papillomaviruses—to vaccination andbeyond. Biochemistry (Mosc) 2008;73:498–503.73. Steben M, Duarte-Franco E. Human papillomavirusinfection: epidemiology and pathophysiology. GynecolOncol 2007;107:S2–5.74. Groopman JD, Scholl P, Wang JS. Epidemiology ofhuman aflatoxin exposures and their relationship toliver cancer. Prog Clin Biol Res 1996;395:211–22.75. Brenner H, Bode G, Boeing H. Helicobacter pyloriinfection among offspring of patients with stomachcancer. Gastroenterology 2000;118:31–5.76. Kutok JL, Wang F. Spectrum of Epstein-Barr virus-associated diseases. Annu Rev Pathol 2006;1:375–404.77. Burgos JS. Involvement of the Epstein-Barr virus inthe nasopharyngeal carcinoma pathogenesis. MedOncol 2005;22:113–21.78. Cesarman E, Mesri EA. Kaposi sarcoma-associatedherpesvirus and other viruses in human lymphoma-genesis. Curr Top Microbiol Immunol 2007;312:263–87.79. Ratner L. Pathogenesis and treatment of human T-cell leukemia virus infection. Immunol Res 2005;32:217–23.80. Chow YH, Huang WL, Chi WK, Chu YD, Tao MH.Improvement of hepatitis B virus DNA vaccines byplasmids coexpressing hepatitis B surface antigen andinterleukin-2. J Virol 1997;71:169–78.81. Lowy DR, Schiller JT. Prophylactic human papillo-mavirus vaccines. J Clin Invest 2006;116:1167–73.82. Da Silva DM, Schiller JT, Kast WM. Heterologousboosting increases immunogenicity of chimeric papillo-mavirus virus-like particle vaccines. Vaccine 2003;21:3219–27.83. Kirnbauer R, Chandrachud LM, O’Neil BW, et al.Virus-like particles of bovine papillomavirus type 4 inprophylactic and therapeutic immunization. Virology1996;219:37–44.84. Malfertheiner P, Schultze V, Rosenkranz B, et al.Safety and immunogenicity of an intramuscular Heli-cobacter pylori vaccine in noninfected volunteers: aphase I study. Gastroenterology 2008;135:787–95.85. Lynch HT, Fusaro RM, Lynch JF. Hereditary cancersyndrome diagnosis: molecular genetic clues and cancercontrol. Future Oncol 2007;3:169–81.86. Altshuler D, Daly MJ, Lander ES. Genetic mapping inhuman disease. Science 2008;322:881–8.87. Risch N. The genetic epidemiology of cancer:interpreting family and twin studies and their implica-tions for molecular genetic approaches. Cancer Epide-miol Biomarkers Prev 2001;10:733–41.88. Lichtenstein P, Holm NV, Verkasalo PK, et al.Environmental and heritable factors in the causationof cancer-analyses of cohorts of twins from Sweden,Denmark, and Finland. N Engl J Med 2000;343:78–85.89. Hemminki K. Correspondence re: Risch, N.: Geneticepidemiology of cancer: interpreting family and twin






studies and their implications for molecular geneticapproaches [Cancer Epidemiol Biomarkers Prev2001;10:733–41] Cancer Epidemiol Biomarkers Prev2002;11:423–6.90. Greenwald P. Cancer risk factors for selectingcohorts for large-scale chemoprevention trials. J CellBiochem Suppl 1996;25:29–36.91. Matakidou A, Eisen T, Houlston RS. Systematicreview of the relationship between family history andlung cancer risk. Br J Cancer 2005;93:825–33.92. Li FP, Fraumeni JF, Jr. Soft-tissue sarcomas, breastcancer, and other neoplasms. A familial syndrome? AnnIntern Med 1969;71:747–52.93. Hirschhorn JN, Daly MJ. Genome-wide associationstudies for common diseases and complex traits. NatRev Genet 2005;6:95–108.94. Wang WY, Barratt BJ, Clayton DG, Todd JA. Genome-wide association studies: theoretical and practicalconcerns. Nat Rev Genet 2005;6:109–18.95. Tamimi RM. Single nucleotide polymorphisms andbreast cancer: not yet a success story. Breast Cancer Res2006;8:108.96. Justenhoven C, Hamann U, Schubert F, et al. Breastcancer: a candidate gene approach across the estrogenmetabolic pathway. Breast Cancer Res Treat 2008;108:137–49.97. Reid ME, Santella R, Ambrosone CB. Molecular

epidemiology to better predict lung cancer risk. ClinLung Cancer 2008;9:149–53.98. Thorgeirsson TE, Geller F, Sulem P, et al. Avariant associated with nicotine dependence, lungcancer and peripheral arterial disease. Nature 2008;452:638–42.99. Eeles RA, Kote-Jarai Z, Giles GG, et al. Multiple newlyidentified loci associated with prostate cancer suscep-tibility. Nat Genet 2008;40:316–21.100. Hunter DJ, Thomas G, Hoover RN, Chanock SJ.Scanning the horizon: what is the future of genome-wide association studies in accelerating discoveries incancer etiology and prevention? Cancer Causes Control2007;18:479–84.101. Easton DF, Pooley KA, Dunning AM, et al. Genome-wide association study identifies novel breast cancersusceptibility loci. Nature 2007;447:1087–93.102. Thomas G, Jacobs KB, Yeager M, et al. Multiple lociidentified in a genome-wide association study ofprostate cancer. Nat Genet 2008;40:310–5.103. Ghoussaini M, Song H, Koessler T, et al. Multipleloci with different cancer specificities within the 8q24gene desert. J Natl Cancer Inst 2008;100:962–6.104. Gold B, Kirchhoff T, Stefanov S, et al. Genome-wide association study provides evidence for a breastcancer risk locus at 6q22.33. Proc Natl Acad Sci U S A2008;105:4340–5.

105. Terwilliger JD, Hiekkalinna T. An utter refutation ofthe ‘‘fundamental theorem of the HapMap’’. Eur J HumGenet 2006;14:426–37.106. Lambert BW, Terwilliger JD, Weiss KM. ForSim: a toolfor exploring the genetic architecture of complex traitswith controlled truth. Bioinformatics 2008;24:1821–2.107. Szmuness W, Stevens CE, Zang EA, Harley EJ,Kellner A. A controlled clinical trial of the efficacy of thehepatitis B vaccine (Heptavax B): a final report.Hepatology 1981;1:377–85.108. Groden J, Thliveris A, Samowitz W, et al. Identifi-cation and characterization of the familial adenomatouspolyposis coli gene. Cell 1991;66:589–600.109. Nishisho I, Nakamura Y, Miyoshi Y, et al. Mutationsof chromosome 5q21 genes in FAP and colorectal cancerpatients. Science 1991;253:665–9.110. Miki Y, Swensen J, Shattuck-Eidens D, et al. A strongcandidate for the breast and ovarian cancer suscepti-bility gene BRCA1. Science 1994;266:66–71.111. Wooster R, Bignell G, Lancaster J, et al. Identifica-tion of the breast cancer susceptibility gene BRCA2 .Nature 1995;378:789–92. Erratum in: Nature 1996;379:749.112. Fisher B, Costantino JP, Wickerham DL, et al.Tamoxifen for prevention of breast cancer: report ofthe National Surgical Adjuvant Breast and Bowel ProjectP-1 Study. J Natl Cancer Inst 1998;90:1371–88.

Cancer Research





Correction: AACR Centennial Series Article onLandmarks in the History of Cancer Epidemiology

In the AACR Centennial Series article on landmarks in thehistory of cancer epidemiology in the March 15, 2009 issue ofCancer Research (1), the fourth sentence of the Introduction shouldread ‘‘As illustrated in Figs. 1, 2, and 3, site-specific data on themortality of women and men in the United States from cancer hasbeen collected and published since 1930.’’ Also, the fourth sentenceof the last paragraph on page 2156 should read ‘‘Publication of theHeart and Estrogen/progestin Replacement Study in 1998 wasfollowed by a moderate decline in hormone replacement therapyuse (Fig. 4; ref. 49).’’ Finally, the second sentence on page 2157should read ‘‘Together, these two adverse disease outcomes, shownat a higher level of evidence in the Women’s Health Initiative thanin the earlier, observational studies, had downstream ramifica-tions—massive cessation of hormone replacement use by post-menopausal women, exceeding that following publication of theHeart and Estrogen/progestin Replacement Study (Fig. 4; ref. 49).’’

I2009 American Association for Cancer Research.doi:10.1158/0008-5472.CAN-69-16-COR1

Cancer Res 2009; 69: (16). August 15, 2009 6758 www.aacrjournals.org

Correction

1. Greenwald P, Dunn BK. Landmarks in the history of cancer epidemiology. CancerRes 2009;69:2151–62.

2009;69:2151-2162. Published OnlineFirst March 10, 2009.Cancer Res Peter Greenwald and Barbara K. Dunn Landmarks in the History of Cancer Epidemiology

Updated version

10.1158/0008-5472.CAN-09-0416doi:

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/69/6/2151.full#ref-list-1

This article cites 99 articles, 19 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/69/6/2151.full#related-urls

This article has been cited by 8 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/69/6/2151To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-09-0416

http://cancerres.aacrjournals.org/content/69/6/2151.full#ref-list-1

http://cancerres.aacrjournals.org/content/69/6/2151.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/69/6/2151


LandmarksintheHistoryofCancerEpidemiology · characteristics, or they serve as putative surrogates...

Documents

Transcript of LandmarksintheHistoryofCancerEpidemiology · characteristics, or they serve as putative surrogates...