Identification of biomarkers for the detection of early stage lung adenocarcinoma by microarray profiling of long noncoding RNAs

Background: Lung adenocarcinoma has one of the poorest outcomes of any cancer worldwide, in part due to the lack of a reliable means of early detection. Long noncoding RNAs (IncRNAs) have been shown to be deregulated in some types of cancer; however, the contributions of lncRNAs to lung adenocarcinoma remain unknown. Methods: We described the expression profile of IncRNAs in human lung adenocarcinoma at an early stage and the corresponding adjacent nontumorous tissues (NT) by microarray. From the microarray analysis, a total of 1170 IncRNAs were significantly differentially expressed in three early stage lung adenocarcinoma tissues compared with NT (fold-change >= 2.0, p <= 0.05). Candidate biomarkers were selected from the significantly differentially expressed IncRNAs based on our established filtering pipeline; subsequently, marker optimization and validation by reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) on a total of 102 pairs of early stage lung adenocarcinoma and NT samples. Results: A panel of 5-IncRNA was identified that could distinguish early stage lung adenocarcinoma from NT samples with high sensitivity and specificity. The area under the receiver operating characteristic curve (AUC) for tumor identification in the training and validation sets were 0.978 and 0.987, respectively. Conclusions: Our results are the first to reveal differentially expressed lncRNAs in early stage lung adenocarcinoma and suggest that IncRNAs may be novel candidate biomarkers for the early detection of this disease. (C) 2015 Elsevier Ireland Ltd. All rights reserved.