Estimation of temporal effects in treatment-induced second cancer

Cancer chemotherapy has been remarkably successful in the treatment of several types of malignancies, but has also been demonstrated to cause leukaemia and perhaps other cancer in long-term survivors. Radiotherapy also carries a carcinogenic risk. A large case-control study of second cancer has been carried out, with the aim of quantifying the risk due to chemotherapy and radiotherapy. One of the most important goals of this study is the estimation of the temporal pattern of risk following chemotherapy. Methods are presented for modelling risk as a function of type of treatment and the interval since treatment. The methods are applications of generally available linear regression programs for epidemiological data, and could be equally well applied to studies of occupationally induced cancer.