The relationship of cancer characteristics and patient outcome with time to lung cancer diagnosis after an abnormal screening CT

Abstract

Purpose: The National Lung Screening Trial (NLST) demonstrated a reduction in lung cancer and all-cause mortality with low-dose CT (LDCT) screening. The aim of our study was to examine the time to diagnosis (TTD) of lung cancer in the LDCT arm of the NLST and assess its relationship with cancer characteristics and survival.

Methods: The subjects (N = 462) with a positive baseline screen and subsequent lung cancer diagnosis within 3 years were evaluated by data and image review to confirm the baseline abnormality. The cases were analysed for the relationship between TTD and imaging features, cancer type, stage and survival for 7 years from baseline screen.

Results: Cancer was judged to be present at baseline in 397/462 cases. The factors that showed significant association (p value trend less than 0.05) with longer TTD included smaller nodule size, pure ground glass nodules (GGNs), smooth/lobulated margins, stages I/II, adenocarcinoma, and decreasing lung cancer mortality. The logistic regression model for lung cancer death showed significant inverse relationships with size less than 20 mm (OR = 0.32), pure GGNs (OR = 0.24), adenocarcinoma (OR = 0.57) and direct relationship with age (OR = 1.4).

Conclusion: TTD after a positive LDCT screen in the NLST showed a strong association with imaging features, stage and mortality.

Key points: • NLST observed variable time to lung cancer diagnosis from positive baseline screen. • Time to diagnosis was associated with imaging features, cancer type and stage. • In univariate but not multivariate analysis, longer TTD correlated with decreased mortality.

Keywords: Computed tomography; Diagnosis; Lung cancer; Mortality; Screening.