Intestinal-type gastric adenocarcinomas usually are preceded by chronic atr
ophic gastritis, Studies of gastric cancer prevention often rely on identif
ication of this condition. In a clinical trial, we sought to determine the
best serological screening method for chronic atrophic gastritis and compar
ed our findings to the published literature. Test characteristics of potent
ial screening tests (antibodies to Helicobacter pylori or CagA, elevated ga
strin, low pepsinogen, increased age) alone or in combination were examined
among consecutive subjects enrolled in a study of H. pylori and preneoplas
tic gastric lesions in Chiapas, Mexico; 70% had chronic atrophic gastritis,
English-language articles concerning screening for chronic atrophic gastri
tis were also reviewed. Sensitivity for chronic atrophic gastritis was high
est for antibodies to H. pylori (92%) or CagA, or gastrin levels >25 ng/l (
both 83%), Specificity, however, was low for these tests (18, 41, and 22%,
respectively). Pepsinogen levels were highly specific but insensitive marke
rs of chronic atrophic gastritis (for pepsinogen I<25 <mu>g/l, sensitivity
was 6% and specificity was 100%; for pepsinogen I:pepsinogen II ratio <2.5,
sensitivity was 14% and specificity was 96%), Combinations of markers did
not improve test characteristics. Screening test characteristics from the l
iterature varied widely and did not consistently identify a good screening
strategy. In this study, CagA antibodies alone had the best combination of
test characteristics for chronic atrophic gastritis screening. However, no
screening test was both highly sensitive and highly specific for chronic at
rophic gastritis.