Background & Aims
Current surveillance strategies for autoimmune gastritis (AIG) lack consensus, and the prognostic role of serum biomarkers remains unclear. This meta-analysis aimed to evaluate the correlation between serum biomarkers and the risks of gastric lesions in AIG patients.
Methods
Studies comparing serum biomarkers in AIG patients with versus without neuroendocrine tumor (NET), gastric polyp (GP), or gastric cancer (GC) were identified by searching PubMed, EMBASE, and Cochrane databases (up to March 2025). Basic studies, meta-analyses, reviews, case reports, or studies not published in English were excluded. Two investigators independently extracted data and quality-assessed using the Newcastle–Ottawa and AHRQ scales. Statistical analysis used Review Manager 5.4 and Stata MP18. Heterogeneity and publication bias were evaluated. Sensitivity analysis was used to manage heterogeneity and assess the stability of the result.
Results
Thirteen studies (10 for NET, 4 for GP, and 1 for GC) were included. Elevated serum gastrin correlated significantly with NET (MD: 519.05, 95% CI: 227.96∼810.13, P = 0.0005, I2=85%) and GP (MD: 70.54, 95% CI: 43.59∼97.49, P < 0.00001, I2=0%; weighted mean: 240.02 pg/ml). With a high heterogeneity in the NET group, we arranged sensitivity analysis and removed two studies that caused heterogeneity, the result consistently showed a statistical difference (MD: 196.62, 95% CI: 76.61∼316.64, P = 0.001, I2=22%; weighted mean: 811.53 pg/ml). Publication bias was found neither in NET (Begg’s test: P = 0.1078, Egger’s test: P = 0.3553) nor in GP (Begg’s test: P = 1.0000, Egger’s test: P = 0.4771) groups. Chromogranin A (CgA), vitamin B12, and parietal cell antibody (PCA) showed no consistent associations. For GC, limited data suggested milder gastrin elevation and higher Helicobacter pylori co-infection rates.
Discussion
Serum gastrin is correlated with NET and GP in AIG patients. The weighted mean serum gastrin level was 811.53 pg/ml in AIG patients with NET, and 240.02 pg/ml in those with GP. Non-enrollment of randomized controlled trials (RCTs), inconsistency in laboratory methods for biomarker detection and diagnostic criteria, and different units of biomarkers may cause limitations of the present study.
扫码关注我们
求助内容:
应助结果提醒方式:
