Development of panel of three-dimensional biomarkers to identify gastric carcinoma and precancerous lesions of the stomach - An analytical cross-sectional study

Abstract

Purpose Serological biomarkers have a predictive potential for gastric cancer (GC) and can be classified into three dimensions: stomach-specific biomarkers, GC-related environmental factors, and cancer-associated biomarkers. Inflammation impacts multiple serum markers, and relying on a single marker limits diagnostic accuracy. Combining multiple predictive markers improves GC detection. This study aimed to assess the association and combined diagnostic accuracy of a three-dimensional biomarkers panel in GC patients. Methods In this analytical cross-sectional study, patients were recruited into three groups: GC, precancerous conditions/lesions, and controls. The primary outcome was to assess the association and diagnostic accuracy of three-dimensional biomarkers in identifying GC and precancerous conditions/lesions. The panel constituted stomach-specific markers (Pepsinogen I, Pepsinogen II, Pepsinogen I & II ratio, Trefoil factor 3, Gastrin 17), GC-related environmental factors (Blood Group Antigen Binding Adhesin A, H. pylori IgG), and cancer-associated biomarkers (Carbohydrate Antigen 19.9, Carbohydrate Antigen 125 and Osteopontin). Results A total of 228 patients, 76 in each group, were enrolled. The combination of all three-dimensional biomarkers showed a high discriminatory ability for diagnosing GC with AUC of 0.938, sensitivity 94.7%, specificity 81.6% and precancerous conditions/lesions with AUC 0.951, sensitivity 93.4%, specificity 92.1%. Among the three-dimensional biomarkers, the combination of TFF3, H. pylori and CA125 demonstrated a high sensitivity in identifying GC, while the combination of PGI, H. pylori, and CA125 exhibited the highest sensitivity in identifying precancerous conditions/lesions. Conclusion These results emphasize that combined three-dimensional biomarkers showed good discrimination and could be used as a screening panel for diagnosing GC and precancerous conditions/lesions.