Association Between Screen-Detected Gallstone Disease and Cancer in a Cohort Study.

BACKGROUND & AIMS: Knowledge of temporal associations between screen-detected gallstone disease and specific cancers is limited. The objective of this study was to determine if screen-detected gallstones or cholecystectomy are associated with occurrence of gastrointestinal and nongastrointestinal cancers.

METHODS

We performed a cohort study of 3 randomly selected groups from the general population of Copenhagen. Participants (N = 5928) were examined from 1982 through 1992 and underwent abdominal ultrasound examination to detect gallstone disease, but were not informed of their gallstone status. Participants were followed for the occurrence of cancers through national registers until December 2014. We performed multivariable Cox regression analyses to identify factors associated with development of cancer.

RESULTS

Gallstone disease was identified in 10% of participants (591 of 5928); of these, 6.8% had gallstones and 3.2% had cholecystectomy at baseline. The population was followed for a median of 24.7 years (interquartile range, 18.9-32.4 years) with 1% lost. Pooled gastrointestinal cancers were associated with gallstone disease (11.2% of patients with gallstone disease vs 6.64% without; hazard ratio, 1.50; 95% confidence interval, 1.12-2.01). Right-side colon cancer was also associated with gallstone disease (2.57% of patients with gallstone disease vs 0.96% without; hazard ratio, 2.04; 95% confidence interval, 1.10-3.78). Pancreatic, esophageal, gastric, pooled colorectal, left-side colon, sigmoid colon, and rectal cancers were not associated with gallstone disease. Breast cancer had a weak association with gallstone disease depending on other factors (10.6% of patients with gallstone disease vs 7.41% without; hazard ratio, 1.44; 95% confidence interval, 0.99-2.11). Pooled nongastrointestinal and prostate cancers were not associated with gallstone disease.

CONCLUSIONS

Screen-detected gallstone disease in the general population is associated with pooled gastrointestinal and right-side colon cancers. These associations are not due to detection bias or cholecystectomy. Further studies are needed to determine the mechanism of this association.