Epidemiologic methods in cancer research.

As knowledge of the etiology of a disease increases, the disciplines seeming to have the greatest potential for producing further advances change. In the development of knowledge of the infectious diseases, centuries of clinical and pathologic study devoted to definition preceded a phase of productive epi-demiologic investigation that itself has now given way to an era of laboratory predominance. In a more primitive state of investigation are concepts, for example, "positive health" and some psychiatric illnesses, that hardly seem to have achieved sufficient definition to be measured and have their distribution in the population described. Cancer, on the other hand, seems ripe for epidemiologic investigation. Its many components are reasonably well-defined, reasonably easy to diagnose and even, in many areas, reasonably well-reported. And yet, the state of knowledge of the etiology of cancer in man is to primitive that we do not know whether the major etiologic factors are most likely to be found among genetic, hormonal, infectious, chemical, physical, or psychologic agents. Nor do we know to what degree etiologic factors are likely to be found common to the many clinical and pathologic components of the category. For other diseases, in the past, this is the state of ignorance in which epidemiologic methods have provided some useful direction signals. PAST ACCOMPLISHMEMS AND METHODS What do we know of the etiology of cancer in man? Let us stick to well-established facts. We know that exposure to cigarette smoke accounts for about 90 per cent of the present lung cancer incidence. We know that lung cancer can also be induced by prolonged exposure to some product or products of chromate ore processing and that asbestosis is associated with lung cancer and with mesothelioma of the peritoneum. We know that ionizing radiation can produce leukemia, cancer of the lung, skin and bone, and, at least in the fetus, a variety of other neoplasms. The cancerogenic properties of ionizing radiation are shared by short as well as prolonged * Chairman, Department of Epidemiology.