Screening for lung cancer

Abstract

The large clinical studies of lung cancer screening carried out more than 20 years ago were interpreted as evidence against screening. Those studies have been recently reassessed in the light of methodologic flaws in the randomization of subjects at risk for lung cancer. There is no evidence to support the former conclusion that screening is ineffective and the consequent official recommendation not to screen for lung cancer. The hypothesis of overdiagnosis of lung cancers diagnosed by screening is false. Clinical evidence supports the concept that the current dogma against screening for lung cancer is untrue. Indeed, the 5-year survival rate of patients with NSCLC detected in stage I and radically resected ranges from 60% to 80%. This rate is in sharp contrast to the 10% survival rate of stage I NSCLC not resected. About 90% of lung cancer cases are detected among smokers and former smokers; these well-known at-risk subjects should be offered a screening test with the goal of detecting the disease when it is in stage I. It is expected that the techniques for early detection of lung cancer will be refined and become more sensitive in the near future, so that it will be possible to detect an increasingly large proportion of lung cancers when they are truly in stage I (i.e., nonmetastatic) and curable by radical surgical resection. Low-dose helical CT scan is currently believed to represent a very useful technique for screening for lung cancer, with a higher sensitivity than chest radiograph screening. Chest radiography for lung cancer screening, however, is cheaper and ubiquitously available, and it should still be recommended if CT scan is locally unavailable. As underscored in a recent commentary in The Lancet, the existing public health policy discouraging the screening for lung cancer is in urgent need of reconsideration.