Serological assessment of gastric intestinal metaplasia and atrophy using pepsinogen-I, pepsinogen-II and gastrin-17 levels in a low incidence area of gastric cancer endemic for H. pylori infection.

ABSTRACT

Background: Intestinal metaplasia (IM), a precursor of gastric cancer (GC), may be amenable to non-invasive assessment. Aims:We evaluated the diagnostic utility of serum PG-I, PG-II, PG-I/PG-II ratio and gastrin-17 (G-17) to detect IM and atrophy. Methods: The study was conducted at a tertiary care center located in a low-incidence area of GC, endemic for H. pylori infection. Patients with GC and dyspepsia were evaluated by endoscopy, histology for IM (H&E, PAS and Alcian blue stains), gastritis and H. pylori (H&E and Giemsa stains) infection, which was considered to be present if two of three tests (rapid urease test, IgG antibody and histology) were positive. Serum levels of PG-I, PG-II and G-17 were estimated using ELISA. Results: Of 98 patients with GC and 62 with dyspepsia, 35 (36%) and 9 (14%) had IM, respectively (p=0.004). Patients with IM (n=44) had lower PG-I/PG-II ratio than those without IM (n=116; median 4.4, 0.37-23.6 vs. 6.3, 0.19-38.6, respectively; p=0.005). A cut-off value of PG-I/PG-II ratio of 6.0 had 64% sensitivity and 52% specificity for detecting IM (area under ROC curve 0.64). 26/44 (60%) patients with IM and 52/98 (53%) with GC had PG-I/PG-II ratio <6. Serum G-17 was comparable among patients with and without IM. Conclusions: Though PG-I/PG-II ratio was lower in patients with IM, only 60% had a lower ratio suggesting that this test and G-17 may not be useful to detect IM in a low-incidence area of GC, endemic for H. pylori infection.