Western long-term accuracy of serum pepsinogen-based gastric cancer screening

Even though mass gastric cancer screening is controversial in Western countries, diverse strategies have been proposed in intermediate-risk to high-risk areas.

The aim of this study was to determine the long-term accuracy of the pepsinogen test (PG) for early diagnosis of gastric cancer.

A cohort of inhabitants from the Northern part of Portugal, aged between 40 and 74 years (n=5913), subjected to the PG test (PGI≤70 ng/ml and PGI/PGII≤3), were followed up between November 2006 and December 2015. The diagnosis of gastric cancer was determined through linkage to the population-based registry of cancer [North Region Cancer Registry of Portugal (RORENO)].

Twenty-six gastric cancers were diagnosed (0.4%): nine (4%) among individuals 'positive' for the PG test (n=225) and 15 (0.3%) among those who were 'negative' (n=5688) [hazard ratio=12.7; 95% confidence interval (CI): 5.6-28.6]. Individuals with a 'negative PG test' had a 3-year risk of gastric cancer of 0.1%, representing a sensitivity of 35% (95% CI: 17-56%), globally, and of 58% (95% CI: 28-85%) at 3 years follow-up. The median survival rate in both groups was over 24 months.

The PG test was found to be suboptimal as a screening test and, if used (before upper gastrointestinal endoscopy), it is mandatory to repeat it after 3 years.