- Email: [email protected]
Dietary Factors for Cancer of Specific Sites
1316
ONCOLOGY AND CHEMOTHERAPY
Finally, 2 additional factors believed to be related to cancer have not been addressed directly but they must be mentioned: 1) total caloric intake and energy expenditure, and 2) alcohol consumption. G. P. M. 3 tables, 12 references
Dietary Factors for Cancer of Specific Sites C. METTLIN, Cancer Control and Epidemiology, Roswell Park Memorial Institute, Buffalo, New York
Surg. Clin. N. Amer., 66: 917-929 (Oct.) 1986 One reason for such great recent interest in the role of diet and nutrition in the cause and prevention of cancer possibly is that, unlike many other sources of risk, dietary factors appear to be implicated in a wide range of tumor types. Although tobacco is the single most important carcinogen to which the public commonly is exposed, the over-all cancer risk attributable to diet may be greater because of its suspected role in so many different cancers. Some dietary factors linked by epidemiological data to 12 different cancer sites are reviewed. The cancer sites examined involve the digestive, respiratory, genitourinary and reproductive systems, which encompass cancer sites that may account for as much as 85 per cent of all cancer incidences. Excluded from review are cancers of the connective and nerve tissues, blood and lymphatics. Also excluded is a large body of literature based on laboratory data alone, since much of this research is examined elsewhere in this issue. Even though the issues have been simplified greatly by citing only the most salient data, when a number of different cancers are examined a diverse range of dietary factors is implicated. Little is known about a number of cancers and competing hypotheses are supported by conflicting data for others. Despite these obvious limits, some generalizations regarding common patterns of risk at different cancer sites may be possible. For example, although the nutrients and mechanisms may vary by site, lack of consumpt10n orfrmts and vegetables has been linked repeatedly to cancers of the mouth, larynx, esophagus and lung. A similar association is suggested for cancers of the cervix and bladder. Another common factor is the association of dietary fat with cancers of the breast, ovary, prostate and, possibly, colon. Alcohol is associated with cancers of the mouth, larynx and, possibly, rectum. Coffee has been studied in association with cancers of the pancreas, ovaries and lungs but few of these studies have been replicated and the role of this common dietary exposure remains in question. In most other instances unique cancers are linked to unique dietary exposures, for example fiber and colorectal cancer or nitrates and stomach cancer. Given that common and unique sources of dietary risk may be identified for different cancers, future research may benefit from a 2-pronged research strategy. Basic nutrients must be examined for multiple cancers and specific aspects of diet require in-depth study for specific cancers. From such an approach it may be possible to resolve some of the questions and issues reviewed. G. P. M . 53 references
Food and Cancer: Cause and Effect? G. M. WILLIAMS AND J. H. WEISBURGER, Division of Pathology and Toxicology, Naylor Dana Institute for Disease Prevention, American Health Foundation, Valhalla, New York Surg. Clin. N. Amer., 66: 873- 889 (Oct.) 1986
The clinician knows well from the various approaches required in the treatment of cancer in different organs that cancer represents many distinct diseases. This also is true from the perspective of etiology. Thus, a detailed analysis of the complex factors inherent in the occurrence of each specific type of cancer is essential to delineate the elements causing, enhancing or reducing the risk of that particular disease. It is evident that most of the main human cancers in the world do not stem from intentional or even inadvertent chemical contaminants in the environment. Rather, life-style elements have been established to be involved in the multifactorial causes of most neoplastic diseases. The data relating to the main risk factors responsible for cancer in the upper alimentary tract, stomach, large bowel, pancreas, breast and prostate are reviewed. The application of this knowledge to the primary prevention of these cancers and control of recurrences are considered. The induction of neoplasms is a multi-event process that may be divided into 2 distinct sequences on the basis of the underlying mechanisms: 1) the conversion (initiation) of a normal cell to a neoplastic cell, and 2) the development (promotion) and progression of the neoplastic cell to form a clinically apparent, invasive cancer. Comparative studies in human populations and laboratory investigations have been instrumental in providing sound information on specific risk factors for cancers. The 4 key elements are 1) incidence and mortality data from different geographic areas throughout the world that identify populations as having high or low risk for specific cancers (worthwhile information also stems from studies of select populations within geographic areas but with life-styles different from other inhabitants, reflecting for example religious beliefs, cultural differences or ethnic traditions), 2) changes in disease incidence in migrants between regions of different risk that provide insight into contributions of environmental factors, 3) long-term observations of incidence or mortality as recorded by established, reliable cancer registries in different parts of the world, which provide a picture of trends with which changing life-style patterns can be compared, and 4) laboratory studies for each type of cancer in human subjects through the techniques of biochemical epidemiology or in animal models through other appropriate techniques. Such studies help to define the effect of suspected risk factors and generate information on the underlying mechanisms of action of each agent or groups of agents, particularly with respect to the genotoxic or epigenetic effects discussed previously. Epidemiology has provided data on the incidence of prostatic cancer as a function of several factors, such as race, age, area of residence, endocrine levels and diet. White patients in the Western world have a higher incidence than do the Japanese. In regard to breast cancer first-generation Japanese migrants to the United States retain a low risk but later generations have a higher risk. However, native J apanese have in situ lesions, as do other low risk populations. In general, populations that have a high risk for breast, colon and endometrial cancer also have a high risk for prostatic cancer. An association with diet has been documented for breast and colon cancer as described previously and, therefore, dietary fat also may have a role in the etiology of prostatic cancer. Until recently, a reliable and realistic animal model involving specific chemical carcinogens for prostatic cancer has not been available. Several new approaches currently are under investigation based on mechanistic considerations t hat a tissue stimulated to proliferate during carcinogen administration is more susceptible to carcinogenesis. These models, if validated, should permit exploration of the conditions required for tumor pro-
ONCOLOGY AND CHEMOTHERAPY
Finally, 2 additional factors believed to be related to cancer have not been addressed directly but they must be mentioned: 1) total caloric intake and energy expenditure, and 2) alcohol consumption. G. P. M. 3 tables, 12 references
Dietary Factors for Cancer of Specific Sites C. METTLIN, Cancer Control and Epidemiology, Roswell Park Memorial Institute, Buffalo, New York
Surg. Clin. N. Amer., 66: 917-929 (Oct.) 1986 One reason for such great recent interest in the role of diet and nutrition in the cause and prevention of cancer possibly is that, unlike many other sources of risk, dietary factors appear to be implicated in a wide range of tumor types. Although tobacco is the single most important carcinogen to which the public commonly is exposed, the over-all cancer risk attributable to diet may be greater because of its suspected role in so many different cancers. Some dietary factors linked by epidemiological data to 12 different cancer sites are reviewed. The cancer sites examined involve the digestive, respiratory, genitourinary and reproductive systems, which encompass cancer sites that may account for as much as 85 per cent of all cancer incidences. Excluded from review are cancers of the connective and nerve tissues, blood and lymphatics. Also excluded is a large body of literature based on laboratory data alone, since much of this research is examined elsewhere in this issue. Even though the issues have been simplified greatly by citing only the most salient data, when a number of different cancers are examined a diverse range of dietary factors is implicated. Little is known about a number of cancers and competing hypotheses are supported by conflicting data for others. Despite these obvious limits, some generalizations regarding common patterns of risk at different cancer sites may be possible. For example, although the nutrients and mechanisms may vary by site, lack of consumpt10n orfrmts and vegetables has been linked repeatedly to cancers of the mouth, larynx, esophagus and lung. A similar association is suggested for cancers of the cervix and bladder. Another common factor is the association of dietary fat with cancers of the breast, ovary, prostate and, possibly, colon. Alcohol is associated with cancers of the mouth, larynx and, possibly, rectum. Coffee has been studied in association with cancers of the pancreas, ovaries and lungs but few of these studies have been replicated and the role of this common dietary exposure remains in question. In most other instances unique cancers are linked to unique dietary exposures, for example fiber and colorectal cancer or nitrates and stomach cancer. Given that common and unique sources of dietary risk may be identified for different cancers, future research may benefit from a 2-pronged research strategy. Basic nutrients must be examined for multiple cancers and specific aspects of diet require in-depth study for specific cancers. From such an approach it may be possible to resolve some of the questions and issues reviewed. G. P. M . 53 references
Food and Cancer: Cause and Effect? G. M. WILLIAMS AND J. H. WEISBURGER, Division of Pathology and Toxicology, Naylor Dana Institute for Disease Prevention, American Health Foundation, Valhalla, New York Surg. Clin. N. Amer., 66: 873- 889 (Oct.) 1986
The clinician knows well from the various approaches required in the treatment of cancer in different organs that cancer represents many distinct diseases. This also is true from the perspective of etiology. Thus, a detailed analysis of the complex factors inherent in the occurrence of each specific type of cancer is essential to delineate the elements causing, enhancing or reducing the risk of that particular disease. It is evident that most of the main human cancers in the world do not stem from intentional or even inadvertent chemical contaminants in the environment. Rather, life-style elements have been established to be involved in the multifactorial causes of most neoplastic diseases. The data relating to the main risk factors responsible for cancer in the upper alimentary tract, stomach, large bowel, pancreas, breast and prostate are reviewed. The application of this knowledge to the primary prevention of these cancers and control of recurrences are considered. The induction of neoplasms is a multi-event process that may be divided into 2 distinct sequences on the basis of the underlying mechanisms: 1) the conversion (initiation) of a normal cell to a neoplastic cell, and 2) the development (promotion) and progression of the neoplastic cell to form a clinically apparent, invasive cancer. Comparative studies in human populations and laboratory investigations have been instrumental in providing sound information on specific risk factors for cancers. The 4 key elements are 1) incidence and mortality data from different geographic areas throughout the world that identify populations as having high or low risk for specific cancers (worthwhile information also stems from studies of select populations within geographic areas but with life-styles different from other inhabitants, reflecting for example religious beliefs, cultural differences or ethnic traditions), 2) changes in disease incidence in migrants between regions of different risk that provide insight into contributions of environmental factors, 3) long-term observations of incidence or mortality as recorded by established, reliable cancer registries in different parts of the world, which provide a picture of trends with which changing life-style patterns can be compared, and 4) laboratory studies for each type of cancer in human subjects through the techniques of biochemical epidemiology or in animal models through other appropriate techniques. Such studies help to define the effect of suspected risk factors and generate information on the underlying mechanisms of action of each agent or groups of agents, particularly with respect to the genotoxic or epigenetic effects discussed previously. Epidemiology has provided data on the incidence of prostatic cancer as a function of several factors, such as race, age, area of residence, endocrine levels and diet. White patients in the Western world have a higher incidence than do the Japanese. In regard to breast cancer first-generation Japanese migrants to the United States retain a low risk but later generations have a higher risk. However, native J apanese have in situ lesions, as do other low risk populations. In general, populations that have a high risk for breast, colon and endometrial cancer also have a high risk for prostatic cancer. An association with diet has been documented for breast and colon cancer as described previously and, therefore, dietary fat also may have a role in the etiology of prostatic cancer. Until recently, a reliable and realistic animal model involving specific chemical carcinogens for prostatic cancer has not been available. Several new approaches currently are under investigation based on mechanistic considerations t hat a tissue stimulated to proliferate during carcinogen administration is more susceptible to carcinogenesis. These models, if validated, should permit exploration of the conditions required for tumor pro-