Isoniazid exposure in relation to cancer incidence and mortality in a cohort of tuberculosis patients.

41 226 isoniazid exposed and 22 811 non-exposed patients admitted to Canadian institutions for the treatment of tuberculosis from 1952 to 1960 were followed to 1973 by computerised record linkage to the National Death Index and, for 1969 to 1973, to the National Cancer Incidence Reporting System; The mortality for the total cohort compared to that expected from the mortality in the Canadian population, was greater than expected for tuberculosis, other chest diseases and certain cancer sites associated with cigarette smoking (except bladder cancer). There was, however, no difference between the isoniazid-exposed and non-exposed in mortality from all cancer and individual cancer sites, except for oesophageal cancer in males. Cancer incidence from 1969 to 1973 was also similar for the isoniazid- exposed and non-exposed groups. An elevated standardised incidence ratio was seen only for buccal cavity and pharynx in male exposed subjects. It is concluded that this study provides additional negative evidence on the carcinogenicity of isoniazid in man.