科研之友
微信
新浪微博
Facebook
分享链接
摘要
Background: Early diagnosis of gastric cancer (GC) is crucial for improving patients' outcome, but reliable biomarkers are scarce. This study measured gastric juice pH value, aromatic amino acid (AAA), and total protein concentrations to build a model using these parameters to diagnose GC. Method: We performed a case-control study by comparing the different levels of gastric juice parameters between GC and non-neoplastic gastric disease (NGD) patients. Gastric juice was collected from 200 consecutive patients who underwent gastroscopy at our institution. A total of 120 samples were divided into the derivation subgroup, and others into the validation subgroup by simple random sampling. The demographic and clinicopathological data were also assimilated. Results: Significantly higher levels of male to female ratio, pH value, and AAA content in gastric juice were observed in GC patients than in NGD individuals in the derivation subgroup (all P<0.001). By performing logistic regression analysis with these parameters, we developed a predicting model, defined as gastric juice index (GJI). For the detection of GC, the AUC of GJI was 0.897 (95% CI, 0.838-0.956) in the derivation subgroup and 0.805 (95% CI, 0.704-0.906) in the validation subgroup. Importantly, for the detection of early GC, its AUC was 0.848 (95% CI, 0.771-0.924). At the optimal cutoff value (8.995), its sensitivity, specificity, and accuracy were 75.0%, 78.9%, and 78.2%, respectively. Conclusion: The GJI model established with the above parameters is effective in predicting the existence of GC, and may be used as a diagnostic tool for early GC.
- 出版日期2017
- 单位北京大学; 中国人民解放军第306医院
