Epidemiologic research examines and assesses relationships between biologic/genetic factors, environment, tobacco use, and cancer in humans. INTRODUCTION Recent advances in molecular biology have catalyzed molecular epidemiology. This approach, which tests biological hypotheses about disease in the laboratory through studies in human populations, is being applied by epidemiologists to understand the molecular keys to cancer. In particular, improved technologies to describe the role of genes have increased our ability to define disease risk in terms of genetic traits and provided new insights about how particular genes interact with environmental exposures. The relative roles and magnitude of the effects of inherited susceptibility and exposure to carcinogens can be very different for familial compared to common cancers. Genetic epidemiology seeks to identify the genes that cause cancers in families, often cancers that are found in syndromes that are uncommon in the general population. Molecular epidemiology seeks to identify the genes that contribute to cancer in the general population. While the basic role of tobacco in causing cancer is firmly established, many important questions remain. For example, why do some individuals, smokers and nonsmokers alike, get cancer while others are spared? We need to learn more about cancer in persons who have successfully quit smoking. While the risk of cancer decreases substantially when smokers quit, it never drops to the level of nonsmokers. Unraveling these questions is becoming more important as more people quit. Former smokers, who remain at increased risk for lung cancer for years following cessation because of persistent carcinogen effects at the cellular and genetic levels, are candidates for prevention through chemopreventive agents. Are certain segments of the population, such as women and African-Americans, more vulnerable to tobacco-related cancer and nicotine addiction? New epidemiological studies will allow us to take genetic data obtained from laboratory probes and identify genetic and other biological factors that may increase vulnerability to tobacco-related cancer in these groups. This knowledge could contribute to more rational cancer prevention and treatment strategies for vulnerable individuals. Gender Issues Women started smoking later in this century than men, so the lung cancer epidemic was initially a male phenomenon. However, female lung cancer death rates have increased over 550 percent since 1950, and lung cancer is now the leading cause of cancer death in women in the United States. Today, lung cancer accounts for one in every four female cancer deaths, making it a very significant women's health issue. Following diagnosis, survival is poor (relative 5-year survival rate for women with lung cancer is only 16 percent, and most die within 2 years of diagnosis). Thus, it is generally agreed that attention needs to be focused on smoking prevention and cessation efforts to prevent the disease from developing in the first place. The vast majority of lung cancer cases in women, as in men, are known to be caused by smoking, but important research questions relevant to tobacco and cancer remain: Do similar factors influence girls and boys to initiate smoking or to decide to remain nonsmokers? Are there gender differences in susceptibility to nicotine? Does the same amount of smoking affect women and men differently in terms of lung cancer risk? Are there gender differences in the effect of smoking on risk of specific histologic types of lung cancer? Would cessation programs tailored to concerns expressed more often by women than men who smoke (such as fear of weight gain) increase the proportion of female smokers who quit? The answers to these and other questions will help us to better understand tobacco carcinogenesis and to develop effective tobacco control programs. RECOMMENDATIONS Research should be conducted to understand genetic and environmental interactions in susceptibility to tobacco-related cancers in order to identify subgroups at risk. Continued study of the relationship between genetic and environmental factors is needed to increase understanding of the risk of tobacco-related cancers for specific populations. Some data suggest that there are differences in the risk of tobacco-related cancers between men and women. This might be due to differences in smoking practices or to differences in innate biological mechanisms. Men's and women's lung cancer mutation characteristics also differ. For example, the p53 tumor suppressor gene, which serves many important cellular functions, is more commonly mutated in men, but the type of mutations observed in women is more strongly associated with exposure to the components of tobacco smoke. Women have higher levels than men of estrogen and related hormones, and lung cancers in women more commonly have estrogen and progesterone receptors that are stimulated by these hormones. These receptors may have a special proliferation trigger that increases women's risks for certain types of lung cancer. Despite these observed gender differences in lung cancer biology, it has been difficult to compare the lung cancer risks of men and women smokers, and study results have been inconsistent. Epidemiological studies are needed to establish definitively if there is a difference in tobacco-related cancer risks for men and women and the genetic and other biological determinants of such a difference. Similar studies are needed to identify whether genetic and other biological factors are responsible for different lung cancer risks across races and ethnic groups. Why do some people get cancer from smoking while others with long-term histories of heavy smoking never suffer smoking-related cancers? Answers to this question may arise from studies to characterize the independent and interacting effects of the large number of variations in genes governing the activation and detoxification of tobacco carcinogens as well as continued study of the interplay between genetic and environmental factors. Studies of how diet and potential chemopreventive strategies influence the risk of tobacco-related cancers in people with genetic susceptibilities to cancer from tobacco exposure also are needed. Approaches to reducing the cancer risks of people who have successfully stopped smoking are of special importance. Studies also are needed to determine how smoking alters various types of tissues in former smokers to help explain why they remain more likely to develop cancer than people who have never smoked. Large-scale epidemiologic studies, both case-control and cohort studies, may provide the most definitive answers to these and other questions. However, smaller-scale pilot studies also are needed to characterize and validate individual biological markers of genetic susceptibility to tobacco-induced cancer. Such work will provide critical information on how the genetic makeup of individuals or groups is modified by environmental factors, such as tobacco exposure, thereby leading to the development of new genetic tests that can then be validated in larger studies. THE IMPACT OF EPIDEMIOLOGIC RESEARCH A better understanding of the independent and interacting effects of inherited susceptibilities and tobacco-exposure variables could elucidate risk profiles and the biological mechanisms involved in the development of cancer. Ultimately, this knowledge could lead to the development of tailored approaches to prevention and treatment of tobacco-related cancers in high-risk populations. Finally, studies of youth and young adults, especially cohort studies, may lead to a better understanding of the process of nicotine addiction, as well as genetic and environmental factors that predispose individuals to such addiction.