[Development of an early diagnostic model for intrahepatic cholangiocarcinoma based on serum peptidomics]

Objective: To develop an early diagnostic model for intrahepatic cholangiocarcinoma (ICC) based on serum peptidomic profiling. Methods: This is a retrospective cohort study. Serum samples and clinical data of 421 participants treated at four hospitals in China between April 2018 and November 2023 were retrospectively collected. The cohort included 260 males (61.8%) and 161 females (38.2%), with an age (M(IQR)) of 65(15) years (range: 24 to 93 years). According to the diagnostic results, participants were divided into three groups: 166 patients in the intrahepatic cholangiocarcinoma (ICC) group (39.4%), 118 healthy controls (28.0%), and 137 patients with benign liver diseases (32.6%). The 421 samples were allocated to a training set and a testing set at a ratio of 2︰1 using stratified random sampling for model construction and internal validation. To evaluate the application value of the constructed serum peptide ICC diagnostic model in real-world clinical settings, the latter part of this study was conducted as a prospective, two-center real-world study. Subjects who visited the Sir Run Run Shaw Hospital affiliated with Zhejiang University School of Medicine and the First Affiliated Hospital of Wenzhou Medical University from January 2024 to December 2025 were enrolled as an external validation set according to inclusion and exclusion criteria. Serum peptides were detected using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, and peptide spectra of serum samples were analyzed by label-free quantitative methods. Differential peptides were screened using the Mann-Whitney U test, followed by feature screening combined with a genetic algorithm to construct a logistic regression-based diagnostic model in the training set. The model stability was assessed through ten-fold cross-validation, and validation was performed in an external real-world cohort. Diagnostic performance was evaluated using the receiver operating characteristic curve and its area under the curve (AUC), with sensitivity, specificity, and positive predictive value calculated. The clinical application value of the model was assessed using decision curve analysis. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed on the proteins corresponding to the differential peptides. Results: A total of 107 serum peptides showing significant differences between the ICC group and the healthy control and benign liver disease groups were identified. Eight core peptides with the strongest discriminatory ability for ICC were selected using a genetic algorithm, and a logistic regression-based diagnostic model was constructed accordingly. The model demonstrated good discriminative ability and stability in the training set, testing set, and ten-fold cross-validation. In the external real-world validation cohort, the diagnostic performance was favorable, with a positive predictive value of 90.6%, sensitivity of 89.2%, specificity of 91.4%, and an AUC of 0.971. Decision curve analysis showed that the model provided significant net clinical benefit across a wide range of threshold probabilities. GO and KEGG enrichment analyses indicated that proteins corresponding to the differential peptides were mainly involved in the complement-coagulation cascade and platelet activation pathways, suggesting that abnormalities in these networks may be associated with the occurrence and progression of ICC. Conclusion: The serum peptidomics-based diagnostic model provides high accuracy for early ICC detection and offers molecular insights that may inform future mechanistic studies and therapeutic investigations.