Systemic inflammatory protein signatures associated with gastric lesion progression and risk of gastric cancer

Background Chronic gastric inflammation contributes to the multistep development of gastric cancer (GC), but the relevance of inflammatory proteins to gastric carcinogenesis and whether inflammatory proteins can serve as indicators for GC risk remains unclear. Methods Inflammatory proteins were profiled in a case-control study of the national Upper Gastrointestinal Cancer Early Detection program (UGCED, n = 200, including 63 GCs) and a nested case-control study of the Mass Intervention Trial in Linqu county, Shandong Province (MITS, n = 300, including 150 GCs). Samples were collected 0–12 years before GC diagnosis (n = 500 plasma and n = 166 tissue), with a subgroup obtained paired samples at diagnosis (n = 33). We examined circulating inflammatory proteins associated with GC risk, with external validation conducted in the UK Biobank Pharma Proteomics Project (UKB-PPP, n = 48,011, including 138 incident GCs). Results Our two-stage study identified 9 inflammatory proteins consistently associated with GC risk. Five upregulated proteins (A2M, AFM, BTD, MAP2K7, and TF) indicated short-term GC risk up to 6 years pre-diagnosis, while four proteins (ADH1B, EZR, FN1 and HDGF) presented long-term risk. A higher overall inflammatory-protein-score (O-IPS) integrating 9 proteins was associated with increased GC risk across UGCED (OR = 2.35, 95% CI: 1.59–3.47, per unit increase), MITS (OR = 1.72, 95% CI: 1.30–2.27), and UKB (HR = 1.33, 95% CI: 1.13–1.56). O-IPS was also higher at GC diagnosis than at pre-diagnostic baseline ( P = 0.0016). Integrating key proteins improved discrimination in MITS (AUC: 0.800 vs 0.513, DeLong’s P = 4.58×10 −5 ) and UKB (0.758 vs 0.704, P = 0.0023). Favourable lifestyle might potentially modify the association between O-IPS and GC risk ( P -for-interaction = 0.066). Conclusions Circulating inflammatory proteins demonstrates informative markers for GC risk across different time horizons and enhance risk prediction, supporting potential application in precision prevention.