论文部分内容阅读
目的:评估血清胃蛋白酶原(PG)和胃泌素17联合检测对无症状胃癌高危人群萎缩性胃炎(AG)的价值。方法:于2016年7月至2017年12月选择居住在上海市中心城区的无症状胃癌高危人群377名，根据内镜检查和病理组织学表现分为无萎缩性胃炎(NAG)组、轻至中度萎缩性胃炎(MAG)组和重度萎缩性胃炎(SAG)组，采用ELISA法测定血清PGⅠ、PGⅡ、胃泌素17水平。采用Kruskal-Wallis n H检验、Mann-Whitney n U检验、卡方检验、n Z检验进行统计学分析，运用二元logistic回归分析血清胃蛋白酶原Ⅰ与胃蛋白酶原Ⅱ比值(PGR)、胃泌素17和两者联合检测(联合预测因子)对SAG的预测价值。n 结果:373例无症状胃癌高危人群中，NAG组146例，MAG组136例，SAG组91例，3组无症状胃癌高危人群的性别、年龄、BMI和n H.n pylori免疫球蛋白G阳性率差异均无统计学意义(n P均>0.05)。SAG组无症状胃癌高危人群的血清PGⅠ、PGR和胃泌素17水平均低于NAG组和MAG组[PGⅠ：92.73 pmol/L(20.39 pmol/L，119.28 pmol/L)比133.69 pmol/L(103.44 pmol/L，183.70 pmol/L)、120.64 pmol/L(23.37 pmol/L，173.91 pmol/L)。PGR：7.40(1.63，9.52)比13.71(12.38，16.73)、10.06(4.49，15.87)。胃泌素17：2.31 pmol/L(0.59 pmol/L，3.24 pmol/L)比4.59 pmol/L(1.96 pmol/L，10.86 pmol/L)、12.51 pmol/L(11.96 pmol/L，15.60 pmol/L)]，差异均有统计学意义(n UPGⅠ＝8 262.500、3 360.000，n P＝0.024、n P<0.01；n UPGR＝4 999.000、2 377.500，n P均<0.01；n U胃泌素17＝3 748.500、1 688.000，n P均<0.01)。二元logistic回归分析结果显示，血清PGR(n β＝-0.423，n OR＝0.655，95%n CI 0.516～0.885，n P<0.01)和胃泌素17(n β＝-0.509，n OR＝0.601，95%n CI 0.415～0.859，n P<0.01)均是避免SAG发生的保护因素，血清PGR和胃泌素17对SAG的联合预测因子n L＝3.621-0.423×PGR-0.509×胃泌素17。血清PGR、胃泌素17和两者联合预测因子预测SAG的ROC AUC值分别为0.714、0.655和0.859，预测SAG的灵敏度分别为91.11%、100.00%和100.00%，特异度分别为83.56%、95.74%和99.29%；联合预测因子预测SAG的ROC AUC值高于PGR和胃泌素17，差异均有统计学意义(n Z＝12.354，n P<0.01；n Z＝6.116，n P＝0.031)。n 结论:监测社区无症状胃癌高危人群的PGR、胃泌素17和两者联合均可用于预测SAG，且两者联合检测的预测效能较PGR或胃泌素17单独检测更高。“,”Objective:To evaluate the value of combined detection of serum pepsinogen (PG) and gastrin 17 (G-17) in detection of atrophic gastritis (AG) in asymptomatic population with high-risk of gastric cancer.Methods:From July 2016 to December 2017, a total of 373 asymptomatic population with high-risk of gastric cancer and lived in central city of Shanghai were selected. According to results of endoscopy and histopathology, the residents were divided into non-atrophic gastritis (NAG) group, mild to moderate atrophic gastritis (MAG) group and severe atrophic gastritis (SAG) group. The serum levels of PGⅠ, PGⅡ and G-17 were measured by enzyme linked immunosorbent assay. Kruskal-Wallis n H test, Mann-Whitney n U test, chi square test and n Z test were used for statistical analysis. Binary logistic regression analysis were used to analyze the value of serum PGR, G-17 and their combination (joint predictor) in prediction of SAG.n Results:Among the 373 asymptomatic population with high-risk of gastric cancer, there were 146 cases in NAG group, 136 cases in MAG group and 91 cases in SAG group. There were no statistically significant differences in gender, age, body mass index and infection rate of n H. n pylori immunoglobulin G among the three groups of asymptomatic population with high-risk of gastric cancer. The serum levels of PGⅠ, PGR and G-17 of SAG group were all lower than those of NAG group and MAG group (PGⅠ: 92.73 pmol/L, 20.39 pmol/L to 119.28 pmol/L vs. 133.69 pmol/L, 103.44 pmol/L to 183.70 pmol/L and 120.64 pmol/L, 23.37 pmol/L to 173.91 pmol/L ; PGR: 7.40, 1.63 to 9.52 vs. 13.71, 12.38 to 16.73 and 10.06, 4.49 to 15.87; G-17: 2.31 pmol/L, 0.59 pmol/L to 3.24 pmol/L vs. 4.59 pmol/L, 1.96 pmol/L to 10.86 pmol/L and 12.51 pmol/L, 11.96 pmol/L to 15.60 pmol/L), and the differences were statistically significant (n UPGⅠ=8 262.500 and 3 360.000, n P=0.024, n P<0.01;n UPGR=4 999.000 and 2 377.500, both n P<0.01;n UG-17=3 748.500 and 1 688.000, both n P<0.01). The results of binary logistic regression analysis indicated that serum PGR (n β=-0.423, odds ratio (n OR)=0.655, 95%n CI 0.516 to 0.885, n P<0.01) and G-17 (n β=-0.509, n OR=0.601, 95%n CI 0.415 to 0.859, n P<0.01) were both protective factors to avoid SAG development. And the combination of predictive factors of serum PGR and G-17 for SAG wasn L=3.621-0.423×PGR-0.509×G-17. The receiver operating characteristic (ROC) area under the curve (AUC) values of serum PGR, G-17 and combined predictors to predict SAG were 0.714, 0.655 and 0.859, respectively. And the sensitivity of predicting SAG was 91.11%, 100.00% and 100.00%, respectively; the specificity was 83.56%, 95.74% and 99.29% respectively. The ROC AUC value of combined predictors to predict SAG was higher than those of PGR or G-17 alone, and the differences were statistically significant (n Z=12.354, n P<0.01;n Z=6.116, n P=0.031).n Conclusions:The PGR, G-17 and their combination can all be used to predict SAG in asymptomatic population with high risk of gastric cancer. Compared to those of PGR or G-17 alone, the predictive effect of the combination of them is high.