The problem of second cancers and the European experience with chemoprevention

Abstract

The problem of second primary tumors in patients curatively treated for a prior lung cancer has been underestimated for many years, due to the limited expected survival, inaccurate follow-up and misclassification of solitary relapses in the lungs. None the less, a few careful clinical trials testing the efficacy of lung cancer screening procedures, or exploring the causes of failure after surgical treatment, have contributed to clarifying the magnitude of this problem [l-3]. In patients cured of a cancer of the lung, the frequency of second primary tumours in all sites is lo-25%, depending on the intensity of follow-up, histologic type and stage of the index tumor, and corresponds to an incidence of 2-4% per year. The majority of second primary tumours occur again in the lung (8-20% overall) and, particularly in patients with a good initial prognosis, may represent a relevant cause of mortality. In fact, second primary tumours have emerged as an important field of research, with potential implications in each of the following areas: (a) preclinical events in lung carcinogenesis, (b) screening and early diagnosis, (c) biological classification, (d) salvage surgery and/or chemoradiotherapy, (e) chemoprevention. Multiple primary tumours may arise in the upper aerodigestive tract due to a general&d exposure of this epitelium to the same etiological factors, namely tobacco smoking and other environmental carcinogens. This theory of ‘field cancerization’ indicates that repeated exposure of the entire epithelial surface to carcinogenic insults may result in the occurrence of multiple, independent premalignant or malignant foci [4]. In support of this concept is the recent demonstration that significant genetic changes may be detectable, with various degrees of severity, in bronchial dysplasia as well as in pathologically normal mucosa of patients with lung cancer [5-71. In a series of 65 cases of resected lung cancer, we have shown that the normal bronchial epithelium at distant sites from the primary tumor was affected by multiple genetic abnormalities, including a rearranged karyotype (3p, 7p, 17) or the overexpression of specific oncogenes (EGFR, HER2/NEU) [8]. The overall frequency of genetic changes in the normal epithelium was 46%, being significantly higher