Diabetes Mellitus and Lung Cancer Screening Outcomes in the National Lung Screening Trial.

Abstract

Rationale: Current eligibility criteria for lung cancer (LC) screening are derived from randomized controlled trials and largely based on age and smoking history. However, the individualized benefits of screening are highly variable and may be affected by the presence of coexisting comorbid disease, including diabetes mellitus (DM). Objectives: This study examines differences in screening outcomes for those with or without DM. Methods: This was a secondary analysis of 53,452 high-risk subjects from the National Lung Screening Trial and compared outcomes after screening with computed tomography (CT) or chest radiography according to DM status. Models of LC mortality were derived after adjustment, and LC rate ratios (per 1,000 person-years), including 95% confidence intervals (95% CIs), were examined according to screening arm and DM status. Results: Compared with those without DM, subjects with DM (n = 5,174; 9.7%) had twofold greater baseline prevalence of cardiovascular comorbidity (P < 0.0001), twofold greater non-LC mortality (P < 0.0001), and greater LC lethality (P = 0.02), with more later-stage lung cancer (P = 0.04). We found comparable stage shift and surgical rates favoring the CT arm in both DM and non-DM subgroups, but LC mortality was higher in the CT arm for subjects with DM (2.2% vs. 2.1%), whereas for subjects without DM, it was lower (1.6% vs. 2.0%). However, the unadjusted P value for the interaction between DM status and screening arm was not significant (P = 0.28). In a competing-risk proportional hazards model for LC mortality adjusted for relevant risk factors, the non-DM group had a significant estimated screening benefit (hazard ratio, 0.82; 95% CI, 0.72, 0.94; P = 0.003), whereas the DM group did not (hazard ratio, 1.03; 95% CI, 0.71, 1.50; P = 0.88). However, the interaction between DM status and screening arm was again not significant (P = 0.27), indicating no overall screening difference according to DM status. Conclusions: Those reporting DM experienced more advanced LC, greater LC lethality, and greater non-LC mortality, whereas the benefits of CT-based screening remain unclear. Limitations from underpowering, lack of DM severity data, and older treatment approaches may have contributed to inconclusive results, and larger studies are warranted to better examine the effects of comorbid DM on current LC screening outcomes.