Seminars in Cancer Biology 14 (2004) 397–398

Editorial

Modern criteria to establish human cancer etiology

Despite great advances in our understanding of cancerat the cellular and molecular level there has been littleimprovement in our ability to establish cancer causation.Our greatest successes in fighting cancer derive from theidentification and removal or inactivation of carcinogenicsubstances, and from the identification and removal ofpre-malignant lesions. Therefore, understanding cancercausation and cancer prevention should be emphasized. Inthe recent past, several chemical and infectious agents havebeen linked to tumor development. This has led to the im-plementation of preventive and therapeutic measures thathave had a tremendous impact in decreasing cancer inci-dence (e.g.,�-anylin removal and bladder carcinoma; HBVvaccination and hepatocellular carcinoma). It is clear thatthe ideal intervention is to prevent cancer from developingat all. Infectious agents, chemical substances, and physicalfactors have all been associated with disease and cancer cau-sation. For acute diseases associated with microorganisms,the association and causative relationships are usually read-ily established. Koch’s postulates[1] provided a frameworkfor pinpointing the associated bacterium with the specificillness it caused. In situations where such clarity cannotbe obtained experimentally, epidemiologists have used theHill’s criteria established in 1965[2] to link various diseaseswith extrinsic causative factors. For chronic diseases such ascancer, where there may be a long latent period between theinitiation of the diseases and overt illness, these approacheshave generally been unsatisfactory. Thus, there is a need toclarify and enunciate the significant factors that may enableus to establish causal relationships in a more rapid fashion.Epidemiological studies require that the effect, cancer inthis case, has already occurred, when of course it would bemore desirable to identify potential carcinogenic substancesat an earlier stage before they have caused a large number ofmalignancies and thus become identifiable by epidemiolog-ical studies. Experimental animal studies present a comple-mentary set of strengths and limitations: exposure is clearlydefined but the question of relevance must be addressed.As a result, animal bioassays can demonstrate causality butthe question is whether these causal relationships are rele-vant to humans. In this issue, leading scientists representingvarious disciplines discuss, how the integration of classicalcriteria [1,2] with the technological advances of the past 40years may allow a more rapid and accurate identification of

human carcinogens. Reading the nine articles of this issue,a general consensus emerges on several topics, such as theneed to integrate molecular pathology and epidemiology fora more accurate and rapid identification of human carcino-gens. Epidemiology, molecular pathology (including chem-istry, biochemistry, molecular biology, molecular virology,molecular genetics, epigenetics, genomics, proteomics, andother molecular based approaches) and animal and tissueculture experiments should all be seen as important integrat-ing evidence in the determination of human carcinogenicity.

It is sometimes difficult to determine what a carcinogenis. Is it an agent that causes tumors or an agent that simplyplays a role in tumor development? Classically, we distin-guish initiators or carcinogens, co-carcinogens and tumorpromoters. From a clinical perspective it may be as impor-tant to target initiators, co-carcinogens and promoters, if byremoving any one of them tumor growth can be prevented.Future studies should focus on interactions among and be-tween different biological, chemical and physical agents.Analyses of single agents can at times miss their carcino-genic potential when such agents are carcinogenic onlyin subgroups of individuals because of their genetic back-ground, diet, exposure to other carcinogens or microbialinfection.

To our knowledge, this is the first issue of a scientific jour-nal devoted solely to a discussion of the etiology of humancancer that takes into consideration infectious microbialagents, chemical carcinogens, and other exogenous physicaland environmental factors. This issue could not encompassthe whole multitude of factors involved in cancer etiology,but it is our hope that the conclusions and guidelines thatwe have proposed will provide a framework of acceptablecriteria for establishing causation. Hopefully this issue willassist in a more rapid and accurate identification of hu-man carcinogens and in the implementation of preventivestrategies.

References

[1] Robbins Pathologic Basis of Disease. 6th ed. Koch’s criteria. 1999.p. 330.

[2] Hill B, The environment and disease: association or causation? In:Proceedings of the Royal Society of Medicine. vol. 58; 1965 p. 295–300.

1044-579X/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.doi:10.1016/j.semcancer.2004.06.001

398 Editorial / Seminars in Cancer Biology 14 (2004) 397–398

Michele Carbone∗Department of Pathology, Cardinal Bernardin Cancer

Center Room 205, Loyola University Chicago, 2160 SouthFirst Ave, Maywood, IL 60302 USA

Jack GruberMay Wong

Cancer Etiology Branch, Division of Cancer Biology

National Cancer InstituteBethesda, MD 20892 USA

∗Corresponding author. Tel.:+1 708 327 3250fax: +1 708 327 3238

E-mail address: mcarbon@lumc.edu(M. Carbone)