Identification of a 17-protein signature in the serum of lung cancer patients

Early detection of lung cancer may potentially help to improve the outcome of this fatal disease. Currently, no satisfactory laboratory tests are available to screen for this type of cancer. The aim of this study was to improve diagnostic procedures for lung cancer through the discovery of serum biomarkers using SELDI-TOF MS (surface-enhanced laser desorption/ionization time-of-flight mass spectrometry). Mass spectrometric profiling was applied to the serum of a total of 139 lung cancer patients and 158 healthy individuals for developing a prognostic signature. For validation, two separate groups were employed, comprising of 126 and 50 individuals, respectively. Informative regions of mass spectra were identified by forming protein mass clusters and identifying predictive clusters in a logistic regression model. A total of 17 differential predictive protein mass clusters were identified in patients with metastatic lung cancer disease compared to healthy individuals. These clusters provide for a robust risk prediction model. The sensitivity and specificity of this model was estimated to be 87.3 and 81.9%, respectively, for the first validation set, and 96.0 and 66.7%, respectively, for a second validation set of patients with early disease (stages I and II). A differential 11.5/11.7 kDa double-peak could be identified as serum amyloid a (SAA) by peptide mapping. Our data suggest that the SELDI-TOF MS technology in combination with a careful statistical analysis appears to be a useful experimental platform which delivers a rapid insight into the proteome of body fluids and may guide to identify novel biomarkers for lung cancer disease.