OBJECTIVETo establish a serum protein fingerprint model for prediction of liver metastasis from colorectal cancer by SELDI-TOF-MS analysis, and to determine the differentiatial proteins associated with the metastatic liver cancers.

METHODSData were collected from the Department of General Surgery in Zhongshan Hospital. A group of patients with colorectal cancer (CRC) without liver metastasis (n = 36) and another group with liver metastasis (n = 36) were included in this study. Serum samples were collected from peripheral venous blood before operation. Special serum protein or peptide fingerprint was determined by surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF-MS). The obtained data were analyzed by Biomarker Wizard software to screen the serum protein markers discriminating colorectal cancer patients with and without liver metastasis. A serum protein fingerprint model was established. This model was blindly verified in of CRC patients with and 44 cases without liver metastasis.

RESULTSComparing the characteristic proteins in those two groups of patients, 10 specific protein peaks were identified with statistical significance (P < 0.05). According to m/z growing from small to large, they were: 2398, 2814, 4084, 4289, 4465, 6422, 6619, 11 482, 11 649 and 13 714. The predictive model had a sensitivity of 91.7% and a specificity of 97.2%. The validation showed a sensitivity of 75.0% and a specificity of 81.8%.

CONCLUSIONA predictive model based on differentiatial serum protein fingerprint with high sensitivity and specificity has been successfully established. It should be a very useful tool in detection and diagnosis of liver metastasis in colorectal cancer patients.