Introduction Hepatocellular carcinoma (HCC) remains a leading cause of cancer-related mortality globally, with treatment outcomes closely tied to liver function. This study evaluates the prognostic utility of the albumin-bilirubin (ALBI) score compared to the traditional Child-Pugh (CP) grading, leveraging real-world evidence from a large-scale, multi-center database. Methods The Liver Cancer IN Korea (LINK) research network, a multi-center initiative, retrospectively collected electronic health records from three academic hospitals in South Korea, encompassing HCC patients diagnosed between 2015 and 2020. Inclusion criteria mandated at least one HCC treatment and excluded patients with other primary cancer diagnoses. The study followed patients until death, the last visit, or June 2021, employing standardized data processing and rule-based algorithms for data consistency. The prognostic efficacy of ALBI scores and CP scores was compared through time-dependent receiver operating characteristic (ROC) curves and the inverse probability censoring weighting method. Results From 25,248 newly diagnosed patients, 10,297 were included, with 65.82% having hepatitis B etiology and a mean follow-up of 27.49 months. Patients' classification by modified ALBI (mALBI) grade at diagnosis revealed: grade 1 (48.87%), 2a (20.50%), 2b (24.54%), and 3 (5.17%), with a minimal percentage missing (0.92%). Transarterial therapy (54.07%) and tyrosine kinase inhibitors (84.14% as the first-line systemic therapy) were predominant treatments. The ALBI score demonstrated greater prognostic efficacy than the CP score in long-term outcomes, with time-dependent area under the ROC curve analysis showing a score of 0.71 for ALBI versus 0.67 for CP at 60 months. Furthermore, higher mALBI grades were significantly associated with poorer survival outcomes, as indicated by both univariate and multivariate Cox proportional regression model analyses (p < 0.001). Conclusions The study confirmed the ALBI score's superior prognostic ability over the CP score, especially evident in long-term outcomes, suggesting a shift toward more nuanced liver function assessment tools in real-world clinical practice.

中文翻译：

---

肝细胞癌白蛋白-胆红素评分和治疗结果的预后疗效：一项大规模、多中心真实世界数据库研究。


简介 肝细胞癌 （HCC） 仍然是全球癌症相关死亡的主要原因，治疗结果与肝功能密切相关。本研究利用来自大规模、多中心数据库的真实世界证据，评估了白蛋白-胆红素 （ALBI） 评分与传统 Child-Pugh （CP） 分级相比的预后效用。方法 韩国肝癌 （LINK） 研究网络是一项多中心计划，回顾性收集了韩国三家学术医院的电子健康记录，包括 2015 年至 2020 年间诊断的 HCC 患者。纳入标准要求至少进行一种 HCC 治疗，并排除了其他原发性癌症诊断的患者。该研究跟踪患者直至死亡、最后一次就诊或 2021 年 6 月，采用标准化数据处理和基于规则的算法来实现数据一致性。通过时间依赖性受试者工作特征 （ROC） 曲线和逆概率删失加权法比较 ALBI 评分和 CP 评分的预后疗效。结果 25,248 例新诊断患者中，共纳入 10,297 例，其中 65.82% 为乙型肝炎病因，平均随访时间为 27.49 个月。诊断时按改良 ALBI （mALBI） 分级对患者进行分类显示：1 级 （48.87%） 、 2a 级 （20.50%） 、 2b 级 （24.54%） 和 3 级 （5.17%），缺失百分比最小 （0.92%）。经动脉治疗 （54.07%） 和酪氨酸激酶抑制剂 （84.14% 作为一线全身治疗） 是主要治疗方法。ALBI 评分显示，在长期结局方面，预后效优于 CP 评分，ROC 曲线分析下的时间依赖性面积显示 60 个月时 ALBI 评分为 0.71，而 CP 评分为 0.67。 此外，较高的 mALBI 等级与较差的生存结果显著相关，如单变量和多变量 Cox 比例回归模型分析所示 （p < 0.001）。结论 该研究证实 ALBI 评分优于 CP 评分的预后能力，在长期结果中尤其明显，表明在现实世界的临床实践中转向更细致的肝功能评估工具。