Background and Objectives: Serum biomarker discovery in resectable pancreatic ductal adenocarcinoma (PDAC) remains a critical unmet need, as over 80% of patients present with unresectable disease. Serum proteomics offers a promising approach for identifying circulating biomarkers associated with early-stage disease; however, clinical translation has been limited by inconsistent validation and the absence of clinically relevant comparator populations. Materials and Methods: We performed a discovery-phase study using data-independent acquisition mass spectrometry-based serum proteomics in 35 patients with resectable, non-metastatic PDAC and 34 non-cancer controls without hepato-biliary-pancreatic disease. Following quality filtering (≥80% detection threshold), 407 proteins were retained for analysis. Differential abundance was assessed using Welch’s t-test with Benjamini–Hochberg correction (FDR < 0.01, |FC| ≥ 1.5). Diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis and logistic regression with repeated stratified 5-fold cross-validation (100 repetitions) and bootstrap resampling (1000 iterations). Functional enrichment analysis was performed using g:Profiler. Results: Ninety proteins were significantly altered in PDAC (50 increased, 40 decreased). Inter-alpha-trypsin inhibitor heavy chain H3 (ITIH3) demonstrated the highest individual diagnostic performance (AUC = 0.90), followed by coagulation factor XIII A chain (F13A1; AUC = 0.89) and ferritin light chain (FTL; AUC = 0.86). Functional enrichment revealed significant overrepresentation of secretory granule lumen components (adjusted p = 0.001) and complement/coagulation pathways (adjusted p < 0.001). An enrichment-guided three-protein panel (ITIH3, F13A1, and FTL) achieved an AUC of 0.98 (95% CI: 0.95–1.00), with a cross-validated mean AUC of 0.96, sensitivity of 83% (95% CI: 66.4–93.4%), and specificity of 100% (95% CI: 89.7–100%) within the discovery cohort. Conclusions: This discovery-phase study identifies a biologically coherent serum protein signature enriched for secretory granule lumen components in resectable PDAC. The three-protein panel demonstrates strong internal validation performance; however, these estimates may be optimistic due to feature selection performed prior to cross-validation. External validation in independent cohorts—including chronic pancreatitis controls and parallel CA19-9 assessment—will be essential to determine clinical applicability.
Moldovan et al. (Mon,) studied this question.