Radiotherapy in Non-small Cell Lung Cancer

First, in NSCL, as in all other cancers, postoperative RT can have an impact on survival only in patients without occult metastases at the time of initial treatment and in whom residual tumor has been left by the surgeon. This subset of patients ought to be limited, because if the tumor is small, the surgeon will be able to resect it completely; if it is large and has spread along the lymphatic pathways, resection is difficult and there is a high likelihood of distant dissemination. The natural history of breast cancer is fairly well quantitated, and it was estimated that for breast cancer the subset of patients who can benefit from postoperative radiotherapy is not larger than approximately 20% of patients. This subset is probably smaller in patients with NSLC because postmortem examinations in patients who died shortly after an apparently satisfactory surgical resection of NSLC suggest that residual tumor is present in about one third and that occult metastases are already present in at least half of these. Therefore, the maximum gain in survival which can be expected is about 15%. To demonstrate such a small increase in survival, a controlled clinical trial should include at least 650 patients. However, even this relatively large number is too small, because to identify the subset of patients in whom survival is increased, a proper stratification is mandatory at least by stage (TNM) and histology. Therefore, 1,200 patients is a good compromise both for the statisticians who wish to include a large number of patients to increase the reliability of the data and the clinicians whose intent is to shorten the duration of the trial. These considerations emphasize the fact that controlled trials which include only a few hundred patients cannot lead to significant results and can even be misleading. Trials with a small number of patients can contribute to the advancement of knowledge only if designed to be included in a metaanalysis; otherwise, they merely initiate discussions or controversies but fail to provide valid answers.