Search

Background: Large randomized trials have demonstrated that lung cancer (LC) screening with low-dose computed tomography (LDCT) reduces LC mortality in heavy smokers. We previously showed in the MILD screening trial that the combination of a prespecified circulating microRNA (miRNA) signature classifier (MSC) and LDCT improves the accuracy of LDCT alone. The primary aim of the prospective BioMILD study was to assess the additional value of the blood MSC assay at the time of baseline LDCT with the goal of personalizing LC screening intervals. Patients and methods: The study enrolled 4119 volunteers from January 2013 to March 2016, with a median follow-up of 5.3 years. Baseline LDCT and miRNAs stratified participants into four groups: CT−/MSC− (n = 2664; 64.7%); CT−/MSC+ (n = 800; 19.4%); CT+/MSC− (n = 446; 10.8%); and CT+/MSC+ (n = 209; 5.1%). As per the protocol, those in the CT−/MSC− and CT−/MSC+ groups were allocated to LDCT repeat at 3-year and 1-year intervals; CT+ participants were allocated for 1-year or earlier intervals on the basis of LDCT features independent of MSC results. Results: CT+ participants had a 15.8-fold higher 4-year LC incidence than CT− participants (95% confidence interval 10.34-24.05), and MSC+ participants had a 2.0-fold higher 4-year LC incidence than MSC− participants (95% confidence interval 1.40-2.90); there was no evidence that the MSC effect differed between CT+ and CT− participants. LC incidence at 4 years was 0.8% in CT−/MSC−, 1.1% in CT−/MSC+, 10.8% in CT+/MSC−, and 20.1% in CT+/MSC+ participants. LC mortality rates at 5 years in the four risk groups were 0.5 in CT−/MSC−, 1.5 in CT−/MSC+, 4.2 in CT+/MSC−, and 10.1 in CT+/MSC+. Conclusion: The combined use of LDCT and blood miRNAs at baseline predicts individual LC incidence and mortality, with a major effect of MSC for LDCT-positive individuals. These findings may have important implications in personalizing screening intervals.