Final results of early non-invasive gastric cancer detection study using the serum pepsinogen test method in Croatian patients.

Abstract

e15551 Background: Gastric cancer (GC) is the 8th most common cancer cause of cancer deaths in Croatia and 2nd most common cause of cancer deaths worldwide. GC of the intestinal type is usually preceded by a chronic atrophic gastritis which is a precancerous change in the stomach. Methods: A reliable diagnostic tool for the early detection of the GC is essential. Screening programs have led to an improvement of overall 5-y survival rate for GC in Japan and Pepsinogen test method was suggested to reduce mortality from gastric cancer. The gold standard for the diagnosis of GC is the pathohistological study of biopsies obtained during an upper GI endoscopy, an invasive method that is too complicated for use in population screening and patients with comorbidity. We have conducted a prospective single center clinical study over a period of > 2 years to evaluate sensitivity and specificity of reagents, and to determine if these reagents can be part of routine. Inclusion criteria for the study were: signed informed consent, life expectancy > 12, and exclusion criteria were: previous treatments for any malignancy, current usage of IPP or NSAIDs medication, poor ECOG performance status ≥3, known history of H. pylori eradication treatment or gastric surgery. We previously reported preliminary results of PG test method in Croatian population. Here, we present mature data after median follow-up of 26 months. Statistical analyses were performed by using a Mann-Whitney U test, multiple logistic regression and the receiver operating characteristics (ROC) to evaluate the predictive power of biomarkers. Results: Blood samples have been collected from patients with suspicious to GC attending for endoscopy or surgery. We used cut off points to evaluate gastric cancer risk: PGI ≤ 70 and PGI/II ratio ≤ 3.0. Of the 116 patients, 25 patients had GC and 91 had non-malignant pathology on tissue biopsy (like atrophic gastritis). Based on an optimal cut-off value calculated by ROC curve analysis had accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of 86.2%, 60.0%, 93.4%, 65.2%and 89.47%, respectively, for the diagnosis of GC. AUC was 0.767 (95% CI 64.0-89.0). Conclusions: The single use of pepsinogen tests is not sufficient for stomach cancer detection; however, it provides a valuable test for selecting a population that needs further diagnosing. Meanwhile, its high specificity could also help to avoid unnecessary endoscopy, especially in older population or patients with heavy burden of comorbidity. Clinical trial information: 2016-0019-34.