The risk and predictors of gallbladder cancer in patients with gallbladder polyps: A retrospective cohort study with an insight into confounding by indication

Background and aim:<bold> </bold>We aimed to determine the risk and predictors of gallbladder cancer in all individuals with gallbladder polyps (GP) including those who did not have cholecystectomy. Methods:<bold> </bold>The STROCSS guideline was followed to conduct a retrospective cohort study. All individuals with GP between 2010 and 2019 were followed up to determine the risk and predictors of gallbladder cancer. The primary outcomes were gallbladder cancer and gallbladder dysplasia, and the secondary outcomes included polyp growth rate and polyp disappearance rate. Binary logistic regression analysis and receiver operating characteristic curve analysis were conducted to evaluate the outcomes. Results:<bold> </bold>Analysis of 438 patients showed risk of gallbladder cancer was 0.7% in all polyps (0% in polyps < 10 mm; 5.9% in polyps >= 10 mm). The risk of gallbladder dysplasia or cancer was 1.1% in all polyps (0% in polyps < 10 mm; 10% in polyps >= 10 mm). The polyp size (P = 0.0001) was predictor of cancer; however, patient's age (P = 0.1085), number of polyps (P = 0.9983), symptomatic polyps (P = 0.3267), and change in size (P = 0.9012) were not. Size of 21 mm was cut-off for risk of cancer (area under the curve [AUC]: 0.995, P < 0.001) and 11.8 mm for risk of dysplasia or cancer (AUC: 0.986, P < 0.001). The mean polyp growth rate was 0.3 mm/year and polyp disappearance rate was 16%. Conclusions:<bold> </bold>The GP size remains the only predictor of malignant changes regardless of patient's age, patient's symptoms and number of polyps. The polyp growth rate is unremarkable, and a significant proportion disappears during follow-up. We changed our follow-up protocol with reduced number of scans and early discharge policy.