| Abstract of study | For oral cancer, numerous saliva- and plasma-derived protein biomarker candidates have been discovered and/or verified; however, it is unclear about the behavior of these candidates as saliva or plasma biomarkers. In this study, we developed two targeted assays, MRM and SISCAPA-MRM, to quantify 30 potential biomarkers in both plasma and saliva samples collected from 30 healthy controls and 30 oral cancer patients. Single point measurements were used for target quantification while response curves for assay metric determination. In comparison with MRM assay, SISCAPA-MRM effectively improved (>1.5 fold) the detection sensitivity of 11 and 21 targets in measurement of saliva and plasma samples, respectively. The integrated results revealed that the salivary levels of these 30 selected biomarkers weakly correlated (r < 0.2) to their plasma levels. Five candidate biomarkers (MMP1, PADI1, TNC, CSTA and MMP3) exhibited significant alterations and disease-discriminating powers (AUC = 0.914, 0.827, 0.813, 0.77, and 0.753) in saliva sample; nevertheless, no such targets could be found in plasma samples. Our data support the notion that saliva may be more suitable for the protein biomarker-based detection of oral cancer, and the newly developed SISCAPA-MRM assay could be applied to verify multiple oral cancer biomarker candidates in saliva samples. SIGNIFICANCE: In this work we systematically determined the abundance of 30 selected targets in the paired saliva and plasma samples to evaluate the utility of saliva and plasma samples for protein biomarker-based detection of oral cancer. Our study provides significant evidence to support the use of saliva, but not blood samples, offer more opportunity to achieve the success of protein biomarker discovery for oral cancer detection. |