Research Article
Kazuhiko Uchiyama, Yuji Nai
Abstract
Colorectal cancer (CRC) is a major cause of cancer-related mortality worldwide. Early CRC diagnosis is critical, since patients diagnosed at an early stage have an increased five-year survival rate after surgical resection. Serum biomarkers for CRC detection have been described and peptidomic analysis is a promising approach because of its diagnostic potential. A total of 72 CRC patients and 63 healthy controls were investigated. We used a comprehensive peptide analysis technique, BLOTCHIP®-MS analysis, a combination of electrophoresis and mass spectrometry, for high sensitivity detection of trace amounts of serum peptides. The prediction model comprised five peptides: m/z 1616.66 (fibrinogen alpha chain), m/z 2390.26 (alpha-1-antitrypsin), m/z 2858.42 (AHSG S-cysteinylated form), m/z 3622.78 (VASP), and m/z 3949.98 (F. XIIIa). The three CRC groups, stages II to IV, II + IIIa, and IIIb + IV, were discriminated from controls. High diagnostic performance was suggested by AUC (0.924), sensitivity (83%), specificity (92%), and median probability ratio (6.80) to CRC stage II to IV. We describe a prediction model for CRC diagnosis using candidate biomarker peptides discovered by a one-step direct transfer technology (BLOTCHIP®-MS analysis). The high throughput technology has high reproducibility and is applicable for peptide quantification and differential analysis for biomarker discovery.