Abstract
Method: A total of 103 lung adenocarcinomas from 57 never smokers and 56 ever smokers were studied using gene expression arrays.
Results: Unsupervised hierarchical clustering analysis revealed two groups of patients. In the first group, the tumor gene expression profile resembled that of normal lung. In the second group, there were two distinctive gene clusters, which were not seen in the first group. Gene Ontology terms associated with these two distinctive gene clusters were “cell proliferation” and “invasion”, respectively. The classification was associated with smoking status (p-value 0.01), lepidic component (p-value 0.00005) and EGFR (p-value 0.0002) or KRAS (p-value 0.02) mutation status.
Conclusion: Lung adenocarcinomas have gene expression profiles that correspond to their clinical and histological presentation. Our study confirms the greater aggressiveness of non-lepidic versus lepidic adenocarcinomas and offers new opportunities to uncover genetic alterations that drive clinical evolution.
- © 2013 ERS