OBJECTIVETo evaluate the fast serum pepsinogen level of the healthy adults among local population in areas with high incidence of gastric cancer and to study the suitable cut-off values of serum pepsinogen abnormality for the screen of chronic atrophic gastritis (CAG) and gastric carcinoma (GC) in China.

METHODSSerum PG I and PG II levels were detected with time resolved fluorescence immunoassay (TRFIA). The fast serum PG I and PG I level as well as PG I/PG II ratio of 606 healthy adult residents among local population in Zanhuang county, Hebei province were detected and the normal distribution ranges determined. The relationship between different cut-off values of serum PG I level, PG I/PG II ratio and corresponding pathological changes in gastric mucosae were comparatively analyzed with serum PG detection, endoscopic biopsy and pathological observation in 720 cases of local residents receiving endoscopic examination in the high incidence area of gastric cancer. The efficacy, sensitivity and specificity of different PG I, PG II abnormality cut-off values in the screen p rogram of CAG and GC were statistically analyzed.

RESULTSThe serum PG I, PG II and PG I/PG II ratio levels of healthy adults from a local natural population in the high incidence area of gastric cancer were all skewed from normal distribution. The median level of PG I, PG II and PG I/PG II were 161 microg/L, 14.8 microg/L and 10.5 respectively. Data from comparative studies on serum PG level and pathological changes of gastric mucosae showed that within the serum PG I range from 40 microg/L to 80 microg/L and PG I/PG II ratio range from 3 to 8, sensitivity of the screening program for CAG and GC increased while the specificity decreased along with the increase of cutoff values of serum PG I and PG I/PG II ratio. Results from statistical receiver operator characteristic curve (ROC) analysis suggested that the best cut-off value of PG I and PG I/PG II abnormality for the screening of CAG and GC being PG I < or =60 microg/L,PG I/PG II < or =6 respectively.

CONCLUSIONThe serum PC I, PG II and PG I/PG II ratio levels of healthy adults from a local natural population in the high incidence area of gastric cancer were all skewed from normal distribution. Serum PG I < or =60 microg/L and PG I/PG II ratio < or =6 as abnormal cut-off value for the screen of CAG and GC could result relatively good sensitivity and specificity.