Duodenal Eosinophils and Mast Cells in Functional Dyspepsia: A Systematic Review and Meta-Analysis of Case-Control Studies

BACKGROUND & AIMS: This study explored the link between duodenal eosinophils and mast cells in patients with functional dyspepsia (FD). METHODS: MEDLINE (PubMed) and Embase electronic databases were searched until June 2021 for case-control studies reporting duodenal eosinophils and mast cells in FD. Pooled standardized mean difference (SMD), odds ratio, and 95% CIs of duodenal eosinophils and mast cells in FD patients and controls were calculated, using a random-effects model. RESULTS: Twenty-two case-control studies with 1108 FD patients and 893 controls were identified. Duodenal eosinophils (SMD, 1.29; 95% CI, 0.85-1.73; P = .0001) and mast cells (SMD, 2.11; 95% CI, 1.14-3.07; P = .0001) were increased in FD patients compared with controls. Substantial heterogeneity was found (I-2 = 93.61, P = .0001; and I-2 = 96.69, P = .0001, respectively) and visual inspection of funnel plots confirmed publication bias. Degranulation of duodenal eosinophils was significantly higher in FD patients compared with controls (odds ratio, 3.78; 95% CI, 6.76-4.48; P = .0001), without statistically significant heterogeneity. We conducted a sensitivity analysis for duodenal eosinophils, by including only high-quality studies, and the results remained unchanged (SMD, 1.73; 95% CI, 1.06-2.40; P = .0001), with substantial heterogeneity. Postinfectious FD patients had increased duodenal eosinophils compared with controls (SMD, 3.91; 95% CI, 1.32-6.51; P = .001) and FD patients without any history of infection (SMD, 1.42; 95% CI, 0.88-1.96; P = .001). Helicobacter pylori-negative FD patients had significantly higher duodenal eosinophils compared with controls (SMD, 3.98; 95% CI, 2.13-5.84; P = .0001), with substantial heterogeneity. No significant difference in duodenal eosinophils was seen according to FD subtypes. CONCLUSIONS: This meta-analysis suggests a link between duodenal microinflammation and FD. However, the quality of evidence is very low, largely owing to the unexplained heterogeneity and serious risk of publication bias in all comparative analyses. Thus, causality remains uncertain and further studies are required.