Abstract: Background　Most patients with hepatocellular carcinoma (HCC) present with intermediate or advanced stages, for which radiotherapy and chemotherapy serve as the primary palliative interventions. However, owing to substantial tumor heterogeneity at these stages, controversies remain regarding the comparative prognostic benefits of these modalities. Furthermore, there is still a lack of direct comparative evidence from large-scale cohorts. Objective　To investigate the comparative efficacy and prognosis of radiotherapy, chemotherapy and their combination in a large-scale population-based cohort. Methods　Demographic and clinicopathological data of 5 487 patients diagnosed with unresected T3-T4N0M0 HCC between 2006 and 2016 were retrospectively extracted from the Surveillance, Epidemiology, and End Results (SEER) database. Patients were stratified into four cohorts based on regimens: no treatment (n=2 681), radiotherapy (n=359), chemotherapy (n=2 136 patients), and combined therapy (n=311). The endpoints were overall survival (OS) and cancer-specific mortality (CSM). Inverse probability of treatment weighting (IPTW) was employed to mitigate baseline confounding. Factors associated with the selection of treatment modalities were identified via multivariate logistic regression. Survival curves were constructed using the Kaplan-Meier method, and between group differences were compared using the log-rank test. Univariate and multivariate Cox proportional hazards regression analyses were utilized to identify factors associated with OS. Univariate and multivariate Fine-Gray competing risk models were applied to analyze factors affecting CSM. Stratified analyses based on different TNM stages were performed to further evaluate the effect of TNM stage on treatment efficacy. Results　Multivariate Logistic regression analysis revealed that tumor grade II was an independent predictor for receiving radiotherapy (OR=1.71, 95%CI=1.17-2.51). Predictors associated with chemotherapy included age > 70 years (OR=1.56, 95%CI=1.36-1.79), uninsured status (OR=1.89, 95%CI=1.18-3.02), T4N0M0 (OR=1.47, 95%CI=1.19-1.81), tumor grade II (OR=1.51, 95%CI=1.21-1.89) and grade III (OR=1.31, 95%CI=1.10-1.56), and AFP positivity (OR=1.84, 95%CI=1.46-2.32). Factors associated with significantly higher likelihood of receiving combined therapy included age >70 years (OR=2.17,95%CI=1.60-2.95), T4N0M0 stage (OR=1.60, 95%CI=1.01-2.56), tumor grade II (OR=2.80, 95%CI=1.93-4.06), AFP positivity (OR=1.76, 95%CI=1.05-2.93), and fibrosis score F5-6 (OR=2.64, 95%CI=1.68-4.16). Following IPTW adjustment, survival analysis demonstrated significant differences in OS and CSM across the four treatment groups (P<0.001). Patients with T4N0M0 stage exhibited significantly worse OS (P=0.003) and a higher cumulative incidence of CSM (P=0.008) compared with those with T3N0M0. In the IPTW-adjusted multivariate Cox model, radiotherapy (HR=0.31, 95%CI=0.27-0.35), chemotherapy (HR=0.37,95%CI=0.35-0.40), and combined therapy (HR=0.29, 95%CI=0.26-0.33) were identified as independent protective factors against all-cause mortality. Independent risk factors for all-cause mortality included T4N0M0 (HR=1.18, 95%CI=1.06-1.31), tumor grade II (HR=1.22, 95%CI=1.06-1.40) and grade III (HR=1.61, 95%CI=1.38-1.88), AFP positivity (HR=1.55, 95%CI=1.40-1.72), and high fibrosis score (HR=1.27, 95%CI=1.06-1.52). Multivariate Fine-Gray competing risk analysis indicated that radiotherapy (HR=0.50, 95%CI=0.43-0.58), chemotherapy (HR=0.69, 95%CI=0.64-0.74), combined therapy (HR=0.56, 95%CI=0.49-0.64), and married status (HR=0.91, 95%CI=0.85-0.98) were independent protective factors against CSM. Conversely, T4N0M0 (HR=1.18, 95%CI=1.04-1.33), tumor grade III (HR=1.43, 95%CI=1.21-1.70), and AFP positivity (HR=1.43, 95%CI=1.29-1.59) were independent risk factors for CSM. Stratified analysis demonstrated that among patients with T3N0M0, both radiotherapy and combined therapy were superior to chemotherapy in terms of favorable OS and CSM outcomes (all P<0.05). However, in patients with T4N0M0, no statistically significant differences in OS and CSM were observed among radiotherapy, chemotherapy, and combined therapy (all P>0.05). Conclusions　For patients with unresected T3-T4N0M0 stage HCC, radiotherapy alone, chemotherapy alone, and concurrent chemoradiotherapy can all significantly improve OS and reduce CSM. Radiotherapy alone and combined chemoradiotherapy yield more prominent survival benefits, with superior efficacy compared to chemotherapy alone.

Key words: Hepatocellular carcinoma, Radiotherapy, Chemotherapy, Overall survival, Cancer-specific mortality, Surveillance, Epidemiology and end results

摘要： 背景　肝细胞癌（HCC）患者确诊时多数处于中晚期，放疗与化疗为其主要的姑息治疗手段。但由于该分期肿瘤存在显著异质性，不同放、化疗方案的治疗获益及预后影响差异尚存争议，缺乏大样本研究的直接对比证据。目的　基于大规模人群数据，分析放疗与化疗在未手术的中晚期（T3-T4N0M0 期）HCC患者中的治疗反应及预后价值。方法　回顾性提取2006—2016年监测、流行病学和最终结果数据库（SEER）中确诊的5 487例未行手术治疗且接受放疗和/或化疗的T3-T4N0M0期HCC患者的人口学及临床病理资料。根据治疗方案分为未治疗组（2 681例）、放疗组（359例）、化疗组（2 136例）、放化疗联合组（311例）。结局指标为总生存期（OS）和癌症特异性死亡率（CSM）。采用逆概率治疗加权（IPTW）法控制混杂偏倚；应用多因素Logistic回归模型分析影响治疗模式选择的因素；采用Kaplan-Meier法绘制生存曲线并通过Log-rank检验比较组间差异；采用单因素及多因素Cox比例风险回归分析HCC患者OS的影响因素；采用单因素与多因素FineGray竞争风险模型分析HCC患者CSM的影响因素。通过不同TNM分期分层分析，进一步评估不同TNM分期对HCC治疗疗效的影响。结果　多因素Logistic回归分析，结果显示，肿瘤分化程度II级是患者接受放疗的独立影响因素（OR=1.71，95%CI=1.17~2.51）。年龄>70岁（OR=1.56，95%CI=1.36~1.79）、无医疗保险（OR=1.89，95%CI=1.18~3.02）、T4N0M0分期（OR=1.47，95%CI=1.19~1.81）、肿瘤分化程度II级（OR=1.51，95%CI=1.21~1.89）和III级（OR=1.31，95%CI=1.10~1.56）、AFP阳性（OR=1.84，95%CI=1.46-2.32）是接受化疗的独立影响因素。年龄>70岁（OR=2.17，95%CI=1.60~2.95）、T4N0M0分期（OR=1.60，95%CI=1.01~2.56）、肿瘤分化程度II级（OR=2.80，95%CI=1.93~4.06）、AFP阳性（OR=1.76，95%CI=1.05~2.93）、肝纤维化评分F5-6（OR=2.64，95%CI=1.68~4.16）是患者接受放化疗联合治疗的独立影响因素。IPTW矫正后的生存分析结果显示：4组总体OS和CSM比较，差异均有统计学意义（P均<0.001）；T4N0M0期患者的OS低于T3N0M0期患者（P=0.003），且T4N0M0期患者的CSM发生率高于T3N0M0期（P=0.008）。IPTW矫正后多因素Cox比例风险回归分析，结果显示，单纯放疗（HR=0.31，95%CI=0.27~0.35）、单纯化疗（HR=0.37，95%CI=0.35~0.40）、放化疗联合治疗（HR=0.29，95%CI=0.26~0.33）是HCC患者全因死亡风险的独立保护因素。TNM分期为T4N0M0（HR=1.18，95%CI=1.06~1.31）、肿瘤分化为II级（HR=1.22，95%CI=1.06~1.40）、III级（HR=1.61，95%CI=1.38~1.88）、AFP阳性（HR=1.55，95%CI=1.40~1.72）、肝纤维化评分高（HR=1.27，95%CI=1.06~1.52）是HCC患者全因死亡风险的独立危险因素。多因素FineGray竞争风险模型分析结果显示，单纯放疗（HR=0.50，95%CI=0.43~0.58）、单纯化疗（HR=0.69，95%CI=0.64~0.74）、放化疗联合治疗（HR=0.56，95%CI=0.49~0.64）、已婚（HR=0.91，95%CI=0.85~0.98）是患者癌症特异性死亡风险独立保护因素；TNM分期为T4N0M0期（HR=1.18，95%CI=1.04~1.33）、肿瘤分级为III级（HR=1.43，95%CI=1.21~1.70）、AFP阳性（HR=1.43，95%CI=1.29~1.59）是患者癌症特异性死亡风险的独立危险因素。分层分析结果显示，T3N0M0期患者中，单纯放疗及放化疗联合治疗的生存获益及CSM均优于单纯化疗（P均<0.05）；T4N0M0期患者中，单纯放疗、单纯化疗及放化疗联合治疗之间的OS和CSM比较，差异均无统计学意义（P均>0.05）。结论　单纯放疗、单纯化疗及放化疗联合治疗均可显著改善未手术T3-T4N0M0期HCC患者的OS，并降低CSM。单纯放疗与放化疗联合治疗的HCC患者生存获益更为显著，且疗效均优于单纯化疗。

关键词: 肝细胞癌, 放疗, 化疗, 总生存期, 癌症特异性死亡率, 监测、流行病学和最终结果数据库计划