Serum TNF-α, IL-8, IgA, and Stool Antigen as Immunological Markers of Helicobacter pylori Infection: A Case-Control Immunoprofiling Study

Abstract

Background: Helicobacter pylori (H. pylori) infection orchestrates a coordinated innate–adaptive immune response involving TNF-α-driven mucosal amplification, IL-8-mediated neutrophilic recruitment, IgA humoral activation, and stool antigen reflecting direct bacterial load. Concurrent integrated immunoprofiling of these four mechanistically distinct axes within an Iraqi clinical cohort has not been previously reported in the literature. Objectives: To characterise within-sample immunological differences between confirmed H. pylori cases and screened healthy controls across four biomarkers; to examine stool-antigen-stratum-dependent biomarker gradients; to evaluate the TNF-α/IL-8 ratio; and to identify the strongest confounder-adjusted independent predictor using Firth penalised logistic regression. Methods: Two-gate case-control immunoprofiling study (University of Samarra, Iraq; March–September 2024): n=60 confirmed H. pylori cases (reference standard: ¹³C-UBT + endoscopic biopsy/histopathology + RUT; stool antigen ELISA was measured as an index test and was not included in case definition) vs. n=30 screened healthy controls. Disease burden strata (mild/moderate/severe) were defined by the updated Sydney System histological grade, not by stool antigen tertiles. Statistical methods: Wilcoxon rank-sum; pROC DeLong AUC; Kruskal-Wallis + Dunn-Bonferroni; Spearman + Benjamini-Hochberg FDR; Firth penalised logistic regression (R package logistf v1.24, Heinze 2002). STARD 2015 reporting / QUADAS-2 bias-reduction framework applied. Results: All four biomarkers showed substantially higher concentrations in cases than controls (rank-biserial r=1.00; FDR-corrected q<0.001 for all). Median concentrations (cases vs. controls): TNF-α 48.7 vs. 8.3 pg/mL; IL-8 186.4 vs. 32.1 pg/mL; IgA 3.84 vs. 1.21 mg/dL; stool antigen 102.5 vs. 14.2 U/mL. Within-sample AUC=1.000 for all biomarkers (upper-bound estimate under two-gate design; bootstrap-optimism-adjusted C=0.998). Stool antigen was the only biomarker with a statistically significant gradient across Sydney-graded severity strata (p<0.001). TNF-α/IL-8 ratio was stable across strata (p=0.612). Firth regression: TNF-α was the strongest independent predictor (adjusted OR=15.8; 95% CI 2.4–104.3; p=0.004). Conclusion: Within this case-control immunoprofiling study, all four biomarkers show complete within-sample separation. Important methodological caveats apply: (1) the two-gate design with healthy controls inflates apparent discriminatory performance; (2) stool antigen was an index test and was excluded from case definition to avoid incorporation bias. These findings constitute Proof-of-Concept immunological data. Multi-centre prospective validation in consecutive symptomatic patient series, incorporating virulence genotyping, a validated clinical severity index, and formal incremental value analysis (NRI/IDI, DCA), is required before any clinical application.