Gallstones, cholecystectomy, and colorectal cancer.

PURPOSE

To examine the controversial association of gallstones, cholecystectomy and colorectal cancer. Methodologic explanations for the association include ascertainment bias, unequal diagnostic testing, and necropsy selection bias. Necropsy screening, which eliminates unequal diagnostic testing and ascertainment bias and reduces necropsy selection bias, was used to study this controversy.

METHODS

Adult necropsies at the University of Kansas Medical Center from 1950 to 1984 were reviewed. Patients with colorectal cancer, gallstones, or who had cholecystectomy during life were excluded. The remaining patients were those in whom neither colorectal cancer nor gallstones were suspected during life (reducing selection bias). The occurrence of gallstones and colorectal cancer among these individuals was then determined (reducing ascertainment bias and unequal diagnostic testing).

RESULTS

Of 7485 persons receiving necropsy, 239 had colorectal cancer diagnosed during life and an additional 604 had gallstones or cholecystectomy, leaving 6642 patients available for study. Overall, no association between colorectal cancer and gallstones was found. In women, gallstones were associated with colorectal cancer; 6/447 (1.3%) with gallstones had colorectal cancer compared with 11/2259 (0.4%) without gallstones who had colorectal cancer, p = 0.048, odds ratio 2.78 (95% CI 0.84-8.25). A stronger association was found between right-sided colorectal cancer and gallstones (odds ratio 6.79, 95% CI 1.14-46.46).

CONCLUSIONS

These data suggest an association between gallstones and colorectal cancer among women. Gallstones may indicate patients at higher risk for colorectal cancer. Studies associating cholecystectomy with colorectal cancer may be explained--not by ascertainment bias--but, rather, by susceptibility bias. The reason for the cholecystectomy (gallstones) may be the correct association and not the cholecystectomy itself.