Lung cancer screening moving forward

What have we learned so far from the results of all observational and randomized trials on low dose computed tomography (LDCT) lung cancer screening? Four years ago, the results of National Lung Screening Trial (NSLT) showed that low-dose computed tomography reduces cancer related mortality by more than 20% (1). This trial financed by National Cancer Institute in USA was the most expensive cancer screening trial at a cost of over 250 million USD. Eight years ago, results of following the outcomes of CT screen-detected non-small cell lung cancer (NSCLC) within I-ELCAP group showed 80% and 88% estimated 10-year survival in the group of all detected and operated stage I cancer, respectively. Longer term follow-up of that cohort suggested that this benefit was sustained (2). We know as well that nodule detection rate is high. Rates vary on CT between 5–54% depending on variety of factors such as local rates of exposure to pulmonary fungal infection as well as the experience and imaging technique of the screening team. Lung cancer detection rate vary between 0.4–2.7% depending on inclusion criteria but in the most series being within the range of 1–1.5%. Lung cancers are found in early, operative stages in 65–85% of cases. Further, it appears likely that LDCT will be the mainstay of early lung cancer detections for years to come. On the other hand, it would be good to improve the diagnostic work-up efficiency for lung cancer screening as it could improve the cost efficiency of the screening process while also reducing the potential for “harm”. Further research could also suggest more tailored approaches to the surgical management of screen-detected lung cancer that could reduce the potential for overtreatment. Pyenson and co-workers have convincingly demonstrated that the cost of lung cancer screening could be significantly reduced if optimal smoking cessation approaches are routinely administered to screen-eligible cohorts. In light of the cost to complete a fully powered, conventional randomized screening trial, a number of small, underpowered trials were reported. Without formal provision for data pooling with carefully harmonized trial design parameters, small cohort screening studies are of uncertain value. An interesting perspective on this topic is provided by the chapter authored by Henschke et al. in this issue. The Authors suggest the need for ongoing longitudinal observational studies to analyzing quality, risk assessment and other dynamic screening management issues such as nodule measurement.