Development of novel serum peptide biomarkers for screening pancreatic cancer

Background: Pancreatic cancer (PC) has a poor prognosis. To overcome this poor prognosis, early detection is mandatory and the development of highly sensitive biomarkers is required. This study aimed to explore a novel and valuable biomarker based on serum peptidomic analysis for screening patients with PC, especially in the early stages. Methods: Serum samples were collected from 106 patients with PC between May 2020 and May 2021. Peptidomic profiles were analyzed using BLOTCHIP^®-mass spectrometry (BLOTCHIP^®-MS) and then compared with those of a propensity score–matched healthy cohort of the same size. PC-specific peptides with significant differences between the two groups were selected and quantitatively measured by selective reaction monitoring (SRM). A discriminant formula to calculate a risk index (RI) was developed based on logistic regression analysis. The RI was validated in newly collected serum samples from 131 patients with PC and 131 healthy controls. Results: Six PC-specific peptides were selected and an RI formula was established. When the RI cutoff value was set at 0.604, the sensitivity, specificity, positive predictive value, negative predictive value, and area under the receiver operating characteristic curve were 84.0%, 93.4%, 92.7%, 85.3%, and 0.935, respectively. They were similarly high in the validation cohort, at 89.3%, 81.7%, 83.0%, 88.4%, and 0.935, respectively. Furthermore, the sensitivity was relatively high at 76.5% even in the early stages (stage 0/1A). Conclusions: This novel RI based on serum peptide analysis is useful for screening patients with PC, even in the early stages.