Construction of dynamic online nomogram for spontaneous rupture of primary liver cancer
10.3760/cma.j.cn113884-20241015-00300
- VernacularTitle:原发性肝癌自发性破裂动态在线列线图的构建
- Author:
Yunfang DONG
1
;
Peng CHEN
;
Ziyan YIN
;
Ji LIANG
;
Wei SHI
;
Feng LIU
;
Manqin HU
Author Information
1. 昆明医科大学第二附属医院肝胆胰外科一病区，昆明 650033
- Publication Type:Journal Article
- Keywords:
Hepatic neoplasms;
Spontaneous rupture;
Risk factors;
Dynamic nomogram
- From:
Chinese Journal of Hepatobiliary Surgery
2025;31(1):23-28
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct and evaluate the nomogram prediction model of spontaneous rupture of primary liver cancer (STRPLC), and make the web-based dynamic online nomogram.Methods:Clinical data of 346 patients with PLC treated in the Second Affiliated Hospital of Kunming Medical University were retrospectively analyzed, including 87 males and 15 females, aged 58.15±10.32 years. Single factor and multiple factor logistic regression analysis were used to screen the influencing factors of STRPLC, and the prediction model was constructed based on the nomogram. Receiver operating characteristic (ROC) curve, calibration curve and clinical decision analysis were used to evaluate the model. The web-based dynamic online nomogram was developed using the DynNom package in R4.3.1 software.Results:Multivariate logistic regression analysis showed that the independent risk factors for spontaneous rupture and hemorrhage of tumor were no history of systematic anti-tumor therapy, alpha-fetoprotein (AFP) level, tumor protrusion on liver surface, tumor length, invasion of major blood vessels, and moderate to large amount of ascites (all P<0.05). The area under the receiver operating characteristic curve (AUC) of the prediction model constructed by this nomogram is 0.913 (95% CI: 0.884-0.943), the best cutoff value is 0.254, with a sensitivity of 0.892, and a specificity of 0.803. The calibration curve shows a good agreement between the predicted probability and the actual probability. The decision curve of the model is above the two invalid lines of " none" and " all" in the horizontal range of 0.07-0.98, and the clinical net benefit of the model is >0. Then user-friendly web-based dynamic online nomogram is constructed. Conclusion:Large tumor size, superficial location, no history of systematic anti-tumor therapy, high AFP level, invasion of major blood vessels, and moderate to large amount of ascites are independent risk factors for STRPLC. The prediction model and dynamic online nanogram constructed by this method can effectively assess the risk of STRPLC.
