Ann Surg Oncol. 2025 Apr 5. doi: 10.1245/s10434-025-17176-1. Online ahead of print.
ABSTRACT
BACKGROUND: Statistical cure, defined as achieving life expectancy comparable with that of disease-free individuals, has not been specifically investigated in hepatitis B virus-associated hepatocellular carcinoma (HBV-HCC), which accounts for more than 50% of the global HCC burden. This study aimed to develop a cure model for HBV-HCC after hepatectomy using matched HBV carriers and the general population as reference groups.
METHODS: From a Chinese multicenter database, HBV-HCC patients who underwent curative-intent hepatectomy were retrospectively reviewed. Independent prognostic factors were identified through Cox regression. A spline-based cure model was applied using two reference populations: matched Chinese HBV carriers (from Shanghai Center for Disease Control and Prevention) and the general population (from the National Bureau of Statistics).
RESULTS: The study analyzed 740 HBV-HCC patients. The following eight independent risk factors were identified: preoperative high viral load (hazard ratio [HR] 1.27), Child-Pugh grade (HR 1.21 and 1.43), multiple tumors (HR 1.70), tumor size greater than 5.0 cm (HR 1.47), macrovascular invasion (HR 3.33), microvascular invasion (HR 1.25), intraoperative blood transfusion (HR 1.21), and postoperative HBV reactivation (HR 1.89). The overall cure probability was 21.2% versus that for HBV carriers and 11.1% versus that for the general population. Risk stratification identified distinct groups relative to HBV carriers. Low risk (64.2%) showed an initial cure rate of 30.3% and achieved a 95% cure probability by 8.6 years, whereas high risk (10.5%) showed negligible cure probability.
CONCLUSIONS: This first HBV-HCC-specific cure model demonstrated that statistical cure is achievable for a subset of patients after hepatectomy. Risk stratification identifies patients with varying cure probabilities, providing valuable guidance for personalized treatment strategies and surveillance protocols.
PMID:40188279 | DOI:10.1245/s10434-025-17176-1
