Purpose: The field cancerization concept proposes that molecular alterations caused by oncogenic stressors are present throughout the airway epithelium. Cigarette smoking, the major risk factor for the development of lung cancer, induces molecular changes in the airway epithelium prior to the development of disease; these molecular changes may indicate the risk for lung cancer development. Given that gene expression changes reflect genomic and epigenomic alterations preceding the development of pathologic lesions, we hypothesized that the earliest features of molecular cancerization can be detected in the airway epithelium of a subset of healthy smokers based on the expression of genes associated with the progression toward lung cancer. We also asked whether chronic obstructive pulmonary disease (COPD), a smoking-induced lung disease representing an independent risk factor for lung cancer, is associated with further acquisition of the molecular cancerization features in the airway epithelium.

Methods: Large airway epithelium (LAE; 3rd-4th order bronchi) of healthy nonsmokers (LAE-NS, n=21) and healthy smokers (LAE-S, n=31) and small airway epithelium (SAE; 10th-12th order bronchi) of healthy nonsmokers (SAE-NS, n=63), healthy smokers (SAE-S, n=73) and smokers with COPD (SAE-COPD, n=37) were obtained by bronchoscopic brushings and analyzed using Affymetrix HG-U133 Plus 2.0 arrays. LAE/SAE common airway epithelium-expressed genes (“AE genes”) were defined as a set of genes expressed in at least 20% of any of the LAE/SAE groups. Lung cancer samples (adenocarcinoma, adenoCa, n=45; squamous cell carcinoma, SqCa, n=27; large cell carcinoma, LCC, n=19) and adjacent nonmalignant lung tissue (n=65) were from the published data set of Hou et al (PLoS ONE 2010; 5:e10312). “Cancerization” genes were identified by comparison of expression of the AE genes in lung cancer samples vs nonmalignant adjacent tissue (fold-increase >2, p<0.01 with Benjamini-Hochberg correction). A “cancerization index” (CI) was calculated for each LAE/SAE sample as a number of cancerization genes having expression level > mean ± 2 standard deviations of healthy nonsmokers.

Results: Gene expression analysis identified 641 AE genes up-regulated in lung cancer vs nonmalignant adjacent tissue. This “cancerization” gene set was enriched in biological processes related to cell cycle (M phase, p<10−30) and p53 pathway (p<10−5). Progressive up-regulation of these genes in the tumor tissue was confirmed using 50 lung cancer samples of the original cohort having matched adjacent nonmalignant tissue samples (p<0.001). The top 10 cancerization genes were MMP12, SPP1, GREM1, KRT6A, TOP2A, ANLN, MMP1, AKR1B10, DLGAP5, and GPX2 (p values 10−12–10−34). Notably, 20% of the identified cancerization genes overlapped with our previously described signature of human airway basal progenitor cells (Hackett et al, PLoS ONE 2011;6:e18378). In the LAE-S, there was significantly higher overall expression of the cancerization genes vs LAE-NS (p<0.001), whereas in the SAE, only SAE-COPD, but not SAE-S, demonstrated significantly higher overall expression of these genes (p<0.003 vs SAE-NS, p<0.03 vs SAE-S). Among 23 healthy smokers with paired LAE/SAE samples, 14 had high CI in the LAE only, 1 in the SAE only, and 2 in both regions.

Conclusion: Smoking is associated with acquisition of the molecular features of cancerization in the LAE, which likely precedes that in the SAE, where the molecular cancerization has a stronger association with the development of COPD. These molecular cancerization features in the airway epithelium of subjects without cancer are relevant to understanding the early pathogenesis of lung cancer and for the development of strategies to prevent lung cancer in susceptible individuals.

This content is only available via PDF.