Benjamin Smith

M.D., M.S.

Columbia University

Assistant Professor

NLST (Learn more about this study)

NLST-768

Mar 10, 2021

Dysanapsis and lung cancer

The human airway tree serves as the gateway and filter of inhaled noxious particulates such as tobacco smoke. Using computed tomography (CT), we recently demonstrated that variation in airway tree caliber (i.e., dysanapsis) is common in the general population (Smith BM et al. JAMA 2020).
Computational fluid dynamic studies suggest that airway tree caliber is a major determinant of inhaled noxious particulate deposition. Moreover, using biomarkers of tobacco smoke exposure (urine cotinine and blood mononuclear cell DNA hypomethylation) we have shown that CT-assessed dysanapsis is associated with 40-65% higher levels of cigarette smoke uptake among current smokers independent of body size or daily number of cigarettes smoked. While this in vivo association was observational in nature, we have subsequently demonstrated significantly higher cotinine levels in a mouse model of dysanapsis following controlled exposure to cigarette smoke when compared with control mice.
We therefore hypothesize among older adults with cigarette smoke exposure that CT-assessed dysanapsis will be associated with significantly higher lung cancer incidence and mortality. If confirmed, this study will i) provide evidence that native airway tree structure modifies host susceptibility to tobacco smoking-associated lung cancer and ii) a potential avenue for further risk stratification after initial screening CT.
We propose a nested case-control study of NLST participants to determine whether CT-quantified dysanapsis is associated with higher lung cancer incidence and mortality. The proposed sample will include all participants with screen-detected lung cancer (approximately 600 participants) and a 4:1 ratio of age- and gender-matched controls (approximately 2,400 participants).

We propose a nested case-control study of NLST participants to determine whether CT-quantified dysanapsis is associated with higher lung cancer incidence and mortality. The proposed sample will include all participants with screen-detected lung cancer (approximately 600 participants) and a 4:1 ratio of age- and gender-matched controls (approximately 2,400 participants).

Aim 1: Quantify dysanapsis on the baseline low dose screening CT using semi-automated software (VIDA Diagnostics, Coralville Iowa).

Aim 2: Quantify the association between CT-quantified dysanapsis and incidence density of lung cancer.

Aim 3: Quantify the association between CT-quantified dysanapsis and time-to-death (lung cancer death, and all-cause death).