Extent of surgery and survival in early lung carcinoma - Implications for overdiagnosis in stage IA nonsmall cell lung carcinoma

BACKGROUND. With the advent of minimally invasive surgical techniques, a de termination of the efficacy of limited resection in Stage IA nonsmall cell lung carcinoma (NSCLC) must be determined. The critical question is whether the probability of cure is equivalent with limited resection and with stan dard lobectomy. DATA ANALYSIS. The results of three independent reports are analyzed. The o nly randomized trial is that of the Lung Cancer Study Group, in which 247 p atients with peripheral Stage IA lesions were randomized to lobectomy or li mited resection. The results indicate that recurrence rate and lung carcino ma mortality rates are higher among those undergoing limited resection. In another nonrandomized series, which involved 218 consecutive Stage IA patie nts, there was a significantly inferior survival rates among those undergoi ng limited resection compared with lobectomy. The third study is a retrospe ctive analysis involving 244 Stage I patients treated at the Brigham and Wo men's Hospital (BWH). This trial shows that limited resection is inferior t o anatomic resection in both Stage IA and Stage IB NSCLC. Indeed, the impor tance of complete resection is underscored by the finding that Stage IA pat ients undergoing limited resection were significantly less likely to be cur ed that Stage IB patients undergoing lobectomy. Accordingly the BWH trial s upports the conclusion that the extent of surgical resection is more powerf ul determinant of survival than natural history of disease in Stage I NSCLC . Cumulatively, the results of these three trials show that among patients with Stage I NSCLC, limited resection is inferior to lobectomy with regard to the probability of producing cure. CONCLUSIONS. The finding that limited resection is inferior to lobectomy ha s relevance to the hypothesis that chest X-ray screening may lead to the id entification of clinically unimportant lung carcinomas, which have been ter med pseudotumors. This hypothesis, known as overdiagnosis, is the only alte rnative to the conclusion from four existing randomized trials that chest X -ray screening leads to an improvement in lung cancer cure rates. However, if extent of surgical resection is the major determinant of survival in Sta ge IA NSCLC, then these lesions must be clinically important. These finding s support the conclusion that chest X-ray screening does not lead to the de tection of Stage IA pseudotumors of the lung. Accordingly, current public p olicy calling for no routine screening for the early detection of lung canc er must be reconsidered. Cancer 2000;89:2432-7. (C) 2000 American Cancer So ciety.