Aim
To investigate the potential of serum peptides as a diagnostic tool for hepatocellular carcinoma (HCC) with bone metastasis.
Conclusion
Our study suggested that serum peptides may serve as a diagnosis tool for HCC bone metastasis.
Methods
Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF-MS) was used to characterize the serum peptide profile of HCC patients with bone metastasis. Serum samples from 138 HCC patients (66 cases with and 72 cases without bone metastasis) were randomly assigned into a training set (n = 76) and a test set (n = 62). Differential serum peptides were examined using ClinProt magnetic bead-based purification followed by MALDI-TOF-MS. The sequences of differentially expressed serum peptides were identified using liquid chromatography-mass spectrometry. A diagnostic model was established using a learning algorithm of radial basis function neural network verified by a single blind trial. Receiver operating characteristic (ROC) analysis was performed to evaluate the diagnostic power of the established model.
Results
Ten peptide peaks were significantly different between HCC patients with or without bone metastasis (P < 0.001). Sequences of seven peptides with mass to charge ratios (m/z) of 1780.7, 1866.5, 2131.6, 2880.4, 1532.4, 2489.8, and 2234.3 were successfully identified. These seven peptides were derived from alpha-fetoprotein, prothrombin, serglycin, isoform 2 of inter-alpha-trypsin inhibitor heavy chain H4, isoform 1 of autophagy-related protein 16-2, and transthyretin and fibrinogen beta chains, respectively. The recognition rate and predictive power of a diagnostic model established on the basis of six significant peptides (m/z for these six peptides were 1535.4, 1780.7, 1866.5, 2131.6, 2880.4, and 2901.9) were 89.47% and 82.89%, respectively. The sensitivity and specificity of this model based upon a single blind trial were 85.29% and 85.71%, respectively. ROC analysis found that the AUC (area under the ROC curve) value was 0.911.