Development and validation of a preoperative model in differentiating intrahepatic cholangiocarcinoma from hepatocellular carcinoma / 中华肝胆外科杂志

ABSTRACT

Objective:To develop and validate a preoperative prediction model based on laboratory parameters in differentiating intrahepatic cholangiocarcinoma (ICC) from hepatocellular carcinoma (HCC).Methods:Data from 833 patients with primary liver cancers treated at the First Affiliated Hospital of Nanjing Medical University between January 2016 and May 2020 were retrospectively analyzed. There were 652 males and 181 females, aged (59.0±11.1) years, with 649 patients diagnosed to have HCC and 184 patients ICC. Based on the different admission time, they were divided into a development cohort ( n=577) and a validation cohort ( n=256). Logistic regression analysis was used to identify independent differential factors which were then included in the nomogram. The discrimination and calibration ability of the nomogram were evaluated by using concordance indexes (C-index) and calibration curves. Results:Female ( OR=4.989, 95% CI 2.547-9.772), hepatitis B surface antigen positivity ( OR=0.144, 95% CI 0.074-0.279), α-fetoprotein (21-399 ng/ml, OR=0.142, 95% CI 0.072-0.283; ≥400 ng/ml, OR=0.023, 95% CI 0.006-0.095), carcinoembryonic antigen (>4.7 ng/ml, OR=2.667, 95% CI 1.352-5.261) and carbohydrate antigen 199 >39 ng/ml ( OR=11.019, 95% CI 5.739-21.159) were independent differential factors for ICC. A nomogram was established by incorporating these 5 factors. The C-indexes were 0.942 (95% CI 0.919-0.965) and 0.949 (95% CI 0.914-0.985) in the development and validation cohorts, respectively. Calibration curves showed good agreement between the predicted risk by the nomogram and real outcomes. Conclusion:In this study, a preoperative nomogram for differential diagnosis between ICC and HCC was established. The model could aid clinicians in clinical treatment decision making.