Prognostic analysis and nomogram prediction of early recurrence in patients with combined hepatocellular-cholangiocarcinoma after radical hepatectomy
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background: Combined hepatocellular-cholangiocarcinoma (cHCC-CCA) has a high recurrence risk despite radical hepatectomy. This study aimed to determine the prognostic impact of early recurrence (ER) and construct a nomogram to predict the ER of cHCC-CCA after radical hepatectomy. Methods: We retrospectively enrolled 109 consecutive cHCC-CCA patients who underwent radical hepatectomy from May 2012 to February 2024 at Sun Yat-Sen University Cancer Center and experienced recurrence. These patients were grouped based on postoperative recurrence time. Prognoses were analyzedand a nomogram for predicting ER was constructed and compared with liver cancer staging systems using receiver operating characteristic (ROC) curves and decision curve analysis (DCA). Results: The ER group had shorter median overall survival (12.47 vs. 30.60 months, P <0.0001) and median progression-free survival (2.90 vs. 7.00 months, P <0.0001) after recurrence. Postoperative recurrence time was an independent prognostic factor. A nomogram considering AFP, tumor size, microvascular invasion, macrovascular invasion and tumor differentiation was constructed to predict ER. The calibration curve revealed high consistency between the nomogram predictions and actual observations. The nomogram yielded a greater area under the curve (AUC, 0.750) than Barcelona Clinic Liver Cancer staging system (AUC, 0.660) and TNM staging systems for hepatocellular carcinoma (AUC, 0.673) and intrahepatic cholangiocarcinoma (AUC, 0.696) in predicting ER risk. Both the AUC and DCA indicated superior predictive performance of the nomogram. Conclusions: cHCC-CCA patients with ER have a poor prognosis, and our nomogram can adequately predict the risk of ER after radical hepatectomy, which can assist in the planning of individual postoperative surveillance protocols.