ABSTRACT

　　Early diagnosis and treatment is known to improve prognosis for nasopharyngeal carcinoma (NPC). The study determined the specific peptide profiles by comparing the serum differences between NPC patients and healthy controls, and provided the basis for the diagnostic model and identification of specific biomarkers of NPC. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) can be used to detect the molecular mass of peptides.Mass spectra of peptides were generated after extracting and purification of 40 NPC samples in the training set, 21 in the single center validation set and 99in the multicenter validation set using weak cationic-exchanger magnetic beads. The spectra were analyzed statistically using FlexAnalysis TM and ClinProt TM bioinformatics software. The four most significant peaks were selected out to train a genetic algorithm model to diagnose NPC. The diagnostic sensitivity and specificity were 100% and 100% in the training set, 90.5% and 88.9% in the single center validation set, 91.9% and 83.3% in the multicenter validation set, and the false positive rate (FPR) and false negative rate(FNR) were obviously lower in theNPC group (FPR,16.7%; FNR, 8.1%) than in theother cancer group (FPR,39%; FNR, 61%), respectively. So, the diagnostic model including four peptides can be suitable for NPC but not for other cancers. FGA peptide fragments identified may serve as tumor-associated biomarkers for NPC. J. Cell. Biochem. 113: 2268–2278, 2012. ? 2012WileyPeriodicals,Inc.