Gastric Mucosal Abnormality and Risk of Pancreatic Cancer: A Population-Based Gastric Biopsy Cohort Study in Sweden.

Background: It remains open whether gastric precancerous lesions are associated with an elevated risk of pancreatic cancer. Our aim was to investigate the association between gastric mucosal status and pancreatic cancer risk.

Methods: Patients with gastric biopsies [normal, minor changes, superficial gastritis, and atrophic gastritis/intestinal metaplasia/dysplasia (AG/IM/Dys)] from the Swedish histopathology registers during 1979 to 2011 were included. Cross-linkages with several nationwide registries allowed complete follow-up and identification of pancreatic cancer cases until 2014. Standardized incidence ratios (SIR) and HRs were estimated.

Results: During 3,438,248 person-years of follow-up with 318,653 participants, 3,540 cases of pancreatic cancer were identified. The same pattern of excess risk of pancreatic cancer compared with the general population was observed across all groups: a peak of 12- to 21-fold excess risk in the first year after biopsy [e.g., normal: SIR = 17.4; 95% confidence interval (CI), 15.7-19.3; AG/IM/Dys: SIR = 11.5; 95% CI, 9.9-13.4], which dropped dramatically during the second and third years, followed by 20% to 30% increased risk after the third year (e.g., normal: SIR = 1.2; 95% CI, 1.1-1.4; AG/IM/Dys: SIR = 1.3; 95% CI, 1.1-1.5). However, no significant excess risk was observed with the normal gastric mucosa as reference.

Conclusions: This unique, large pathologic cohort study did not find evidence that abnormal gastric mucosal status is causally associated with a long-term pancreatic cancer risk. However, a highly increased short-term risk was observed for people undergoing gastroscopy with biopsy sampling compared with the general population.

Impact: Further studies for a long-term risk of pancreatic cancer in patients with gastric biopsies are needed, with further adjustments.