Abstract With the wider use of local ablation for hepatocellular carcinoma (HCC), patients who have large tumors close to major hepatic vessels remain a difficult group to treat. In these patients, local tumor progression (LTP) and treatment-related complications are both common. For many years, the evaluation and technical optimization of radiofrequency ablation (RFA) in perivascular HCC have been insufficient, and current guidelines give only general advice rather than clear recommendations. This article is a narrative review with a structured literature search. We searched PubMed, Web of Science, CNKI, and Wanfang from database inception to September 2024 using the keywords “hepatocellular carcinoma, radiofrequency ablation, perivascular tumor, large tumor, heat-sink effect, combined therapy, navigation, artificial intelligence, computational modeling". We mainly included phase I–III clinical studies, randomized controlled trials, observational cohorts, meta-analyses, translational experiments, and modeling studies. The available evidence shows that large perivascular tumors often have uneven heating due to the heat-sink effect. The ablation effect and the risk of vascular or biliary injury are determined by tumor size, vessel diameter and shape, electrode placement, and system power configuration. Perivascular HCC differs clearly from non-perivascular HCC in local control, recurrence pattern, and complication rate. Consequently, RFA alone is generally insufficient and should be considered only within multimodal treatment strategies. At present, the use of multimodal and individualized strategies for large perivascular HCC is still limited, and practical treatment algorithms are lacking. In this review, we discuss how large perivascular HCC should be managed within a multimodal framework, in which non-ablative therapies such as SBRT, transarterial approaches, and systemic treatment constitute the therapeutic backbone, while RFA-based strategies may serve selected adjunctive roles under specific clinical conditions.
Wang et al. (Mon,) studied this question.