Tobacco and cancer: how to react to the evidence

Tobacco is a remarkable product when considering the number of diseases to which it can reliably be linked as a causal agent or at least as a factor which is partially responsible for the occurrence of disease. If we only look at cancer, tobacco is responsible for the vast majority of cancers of the lung, trachea and bronchus; a proportion of cancers of the bladder and of the renal pelvis; a considerable proportion of cancers of the oral cavity, lip, pharynx, larynx and oesophagus with, for these sites, a large potentiation of risk when associated with the drinking of alcohol; a proportion of cancers of the pancreas and possibly of renal adenocarcinoma. In addition, among women, tobacco use has been linked with an increased risk of cervical cancer. In order to evaluate the present burden of tobacco-related cancers in the world precisely, one needs to be able to consult precise statistics of cancer occurrence, tobacco use and the relationship between the two. The availability of such data varies greatly across countries. Although mortality, and to a lesser extent, morbidity cancer statistics are collected in many parts of the world, the coverage is far from exhaustive and of uniform quality. Commercial or governmental sources of data on tobacco also exist, but the most important missing link is knowledge of the population-specific risk attributable to tobacco use in various countries for all tobacco-related diseases. Despite imperfect knowledge, the link between tobacco use and cancer is clear enough to demand strong action.(ABSTRACT TRUNCATED AT 250 WORDS)