Abstract
Objectives. Lung cancer is the leading cause of cancer deaths in Japan with more than 55,000 lives annually. Surgical resection gives the best hope for a cure for non-small cell lung cancer (NSCLC) cases, which comprise 80-85% of lung cancers. However, their long-term survival rate remains unsatisfactory and no more than 50% of the cases that have successfully undergone potentially curative resection can survive for 5 years after operation. Although various genetic and epigenetic changes of cancer-related genes have been identified and examined in the search for clinically relevant prognosticators, no single variable evaluated so far has been proven to predict accurately a lung cancer patient's outcome. Recent rapid progress in proteomic technology has made it possible to analyze protein expression profiles using tiny amount of biological samples to search for molecular markers for cancer classification. Thus, proteomic-based approaches complement the genome initiatives and are increasingly being used to address biomedical questions. In this study, we hypothesized that a protein profile obtained from human lung tissues can accurately distinguish lung cancer tissues from those of normal lung, moreover good prognosis patients from poor prognosis ones. Methods. We applied a proteomic approach using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) to more than 150 surgically resected frozen tissues of human NSCLC and normal lung (NL) tissue specimens. We used sophisticated bioinformatic analyses to select peak profile related to clinical parameters. Results. We were able to obtain over unique 2600 peaks from histologically selected regions of single frozen sections from more than 150 resected human NSCLC and NL tissue specimens. Applying sophisticated bioinformatic methods, we selected the proteomic patterns consisting of 40 peaks that could distinguish NSCLC from NL tissue specimens. Conclusion. We showed that proteomic patterns obtained directly from picograms of fresh human lung tumor tissues with mass spectrometry can be used to classify and predict survival in surgically resected NSCLC patients. These results demonstrate that the proteomic analyses may lead to the identification of novel biomarkers or therapeutic targets and a better understanding of lung cancer biology.
