Low-dose computed tomography for lung cancer screening: comparison of performance between annual and biennial screen.

Abstract

To compare the performance metrics of two different strategies of lung cancer screening by low-dose computed tomography (LDCT), namely, annual (LDCT1) or biennial (LDCT2) screen. Recall rate, detection rate, interval cancers, sensitivity, specificity, positive and negative predictive values (PPV and NPV, respectively) were compared between LDCT1 and LDCT2 arms of the MILD trial over the first seven (T0-T6; median follow-up 7.3years) and four rounds (T0-T3; median follow-up 7.3years), respectively. 1152 LDCT1 and 1151 LDCT2 participants underwent a total of 6893 and 4715 LDCT scans, respectively. The overall recall rate was higher in LDCT2 arm (6.97%) than in LDCT1 arm (5.81%) (p = 0.01), which was counterbalanced by the overall lower number of LDCT scans. No difference was observed for the overall detection rate (0.56% in both arms). The two LDCT arms had similar specificity (99.2% in both arms), sensitivity (73.5%, in LDCT2 vs. 68.5% in LDCT1, p = 0.62), PPV (42.4%, in LDCT2, vs. 40.6%, in LDCT1, p = 0.83) and NPV (99.8%, in LDCT2 vs. 99.7%, in LDCT1, p = 0.71). Biennial screen may save about one third of LDCT scans with similar performance indicators as compared to annual screening. • Biennial LDCT screening may be as efficient as the annual screening. • Annual and biennial LDCT screening have similar frequency of interval lung cancers. • Biennial screening may save about one third of LDCT scans.