Preoperative prediction model for non-neoplastic and benign neoplastic polyps of the gallbladder

Abstract

Background

Gallbladder adenoma represents a precancerous lesion of gallbladder cancer. However, distinguishing it from cholesteryl polyps of the gallbladder before surgery is challenging. Thus, we aimed to comprehensively explore various risk factors contributing to the formation of gallbladder adenoma to facilitate an informed diagnosis and treatment by clinicians.

Methods

We conducted a retrospective analysis of patients who had undergone cholecystectomy at the Affiliated Hospital of Qingdao University between January 2015 and December 2022. Following postoperative pathological examination, patients were categorized into cholesterol polyp and adenoma groups. We analyzed their baseline characteristics, ultrasound imaging variables, and biochemical data using logistic, lasso, and stepwise regression. Subsequently, we constructed a preoperative prediction model based on the independent risk factors.

Results

Regression analysis of 520 gallbladder polyps and 288 gallbladder adenomas in the model group revealed that age, gallbladder wall thickness, polyp size, echogenicity, pedunculation, and adenosine deaminase (ADA) levels were independent predictors of gallbladder adenoma, all with P < 0.05. Using these indicators, we established a regression equation: Logistic (P) = −5.615 + 0.018 * age - 4.64 * gallbladder wall thickness + 1.811 * polyp size + 2.855 * polyp echo + 0.97* pedunculation + 0.092 * ADA. The resulting area under the curve (AUC) value was 0.894 (95 % CI: 0.872–0.917, P < 0.01), with a sensitivity of 75.47 %, specificity of 90.20 %, and overall accuracy of 84.41 % for adenoma detection.

Conclusion

Age, polyp size, gallbladder wall thickness, polyp echogenicity, pedunculation, and ADA levels emerge as independent risk factors for gallbladder adenoma.