Ratio between serum IL‐8 and pepsinogen A/C: a marker for atrophic body gastritis

Abstract

Elevated serum gastrin and a low pepsinogen A/C ratio are well-recognized markers for atrophic body gastritis (ABG). We have shown that the presence of body atrophy is also associated with elevated serum pro-inflammatory cytokines. This study tested the hypothesis that serum cytokines provide additional information to gastrin and pepsinogens in screening for ABG. Two hundred and twenty-six consecutive patients were investigated on referral for upper gastrointestinal endoscopy: 150 were patients with gastro-oesophageal reflux disease, receiving acid inhibitory medication either with proton pump inhibitors (n = 113) or with histamine2-receptor antagonists (n = 37), and 76 were nontreated controls, who had normal endoscopic findings. Gastric mucosal biopsies were sampled for histological examination (Sydney classification). Serum samples were analyzed for gastrin, chromogranin A (CgA), and pepsinogens A and C by RIA, and for the interleukins (IL)-1beta, IL-6, and IL-8 by ELISA. Subjects with ABG had significantly higher serum gastrin (P < 0.01) and serum CgA (P < 0.01) levels and significantly lower pepsinogen A/C ratios (P < 0.001) than those without ABG. Additionally, serum IL-1beta, IL-6 and, especially, IL-8 levels were significantly higher in the subjects with than in those without ABG (P < 0.0001, for all cytokines). To optimize the detection of body atrophy we defined the ABG index: the ratio between the simultaneously measured IL-8 and pepsinogen A/C. The area under the ROC curve for the ABG index was significantly greater than that for serum gastrin and for serum pepsinogen A/C alone (0.91 +/- 0.029 vs. 0.72 +/- 0.042, and vs. 0.83 +/- 0.031, P = 0.018 and P = 0.049). Using the ABG index at a cut-off value of 1.8 pg mL-1, 91% of the cases were classified correctly. The ratio between serum IL-8 and pepsinogen A/C accurately predicts the presence of ABG. We therefore propose the ABG index as a noninvasive screening test for ABG in population-based studies.