aMAP评分联合振动控制瞬时弹性成像肝硬度提升乙肝相关肝癌预测
该研究纳入韩国5家三级医院2005年5月至2021年7月启动抗病毒治疗的944例慢性乙型肝炎患者,评估aMAP评分联合肝硬度(LS)预测肝癌的性能。中位随访68.9个月,49例发生肝癌;aMLaf和aMLc模型5年AUROC均为0.82,优于PAGE-B和modified PAGE-B(均P<.05)。香港队列外部验证Harrell's c-index和iAUC均高于0.80。
要点
- aMAP联合肝硬度模型(aMLaf、aMLc)5年AUROC均为0.82,优于PAGE-B和modified PAGE-B。
- aMLaf低风险组随访期间无肝癌发生。
- 香港队列外部验证Harrell's c-index和iAUC均高于0.80。
核心结论
- aMAP评分联合肝硬度测量可显著改善接受抗病毒治疗的慢性乙型肝炎患者的肝癌风险预测。
- 该联合模型可能有助于更精确的风险分层和个体化监测策略。
研究设计与样本
- 分析韩国5家三级医院2005年5月至2021年7月启动抗病毒治疗的944例慢性乙型肝炎患者。
- 评估aMAP单独、aMAP联合肝硬度预测进展性纤维化(aMLaf)及预测肝硬化(aMLc)的预测准确性。
- 使用香港队列进行外部验证。
主要发现
- 中位随访68.9个月,49例发生肝癌,发生肝癌者aMAP显著高于未发生者。
- aMLaf和aMLc模型5年AUROC均为0.82,均P<.05 vs PAGE-B和modified PAGE-B。
- 多变量分析中,aMAP升高(HR 1.16,95%CI 1.09-1.24)、aMLaf(HR 1.08,95%CI 1.03-1.13)、aMLc(HR 1.06,95%CI 1.03-1.09)为肝癌发生的独立预测因素。
- aMLaf低风险组随访期间无肝癌发生。
- 香港队列外部验证显示aMLaf和aMLc的Harrell's c-index和iAUC均高于0.80。
关键数值与推荐
- aMLaf和aMLc模型5年AUROC均为0.82(均P<.05 vs PAGE-B和modified PAGE-B)。
- 多变量分析HR:aMAP 1.16(95%CI 1.09-1.24),aMLaf 1.08(95%CI 1.03-1.13),aMLc 1.06(95%CI 1.03-1.09)。
- 香港队列外部验证Harrell's c-index和iAUC均高于0.80。
临床怎么用
- 对接受抗病毒治疗的慢性乙型肝炎患者,可考虑将aMAP评分与肝硬度测量结合用于肝癌风险分层。
- aMLaf低风险组随访期间无肝癌发生,可能有助于识别低风险人群并优化监测策略。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND & AIMS: The feasibility of age-male-albumin-bilirubin-platelet (aMAP) score combined with liver stiffness (LS) measurement in predicting hepatocellular carcinoma (HCC) risk remains unexplored. This study evaluated the performance of combining the aMAP score with LS measurements for HCC prediction in patients with chronic hepatitis B (CHB). METHODS: We analyzed 944 patients with CHB who initiated antiviral therapy between May 2005 and July 2021 at 5 tertiary hospitals in Korea. The predictive accuracy of aMAP alone, and aMAP combined with LS for advanced fibrosis (aMLaf), and aMAP combined with LS for cirrhosis (aMLc) was assessed. External validation was conducted using Hong Kong cohort. RESULTS: During a median follow-up of 68.9 months, 49 patients developed HCC, and their aMAP was significantly higher than that of the patients who did not develop HCC. The aMLaf and aMLc models demonstrated superior predictive performance compared with conventional models, with areas under the receiver operating characteristic curve (AUROCs) at 5 years of 0.82 and 0.82, respectively (both P < .05 vs PAGE-B and modified PAGE-B). In multivariate analysis, elevated aMAP (hazard ratio [HR], 1.16l 95% confidence interval [CI], 1.09-1.24), aMLaf (HR, 1.08; 95% CI, 1.03-1.13), and aMLc (HR, 1.06; 95% CI, 1.03-1.09) were independent predictors for HCC development. Notably, no HCC cases occurred in the aMLaf low-risk group during follow-up. External validation in the Hong Kong cohort confirmed the robustness of aMLaf and aMLc, with Harrell's c-index and iAUC values above 0.80. CONCLUSIONS: The integration of aMAP score with LS measurements significantly improved HCC risk prediction in patients with CHB receiving antiviral therapy. This combined model may enable more precise risk stratification and personalized surveillance strategies.
引用
- Chon HY, Yim HJ, Heo SJ, et al. Enhanced Prediction of Hepatitis B Virus-Related Hepatocellular Carcinoma Using Age-Male-Albumin-Bilirubin-Platelet (aMAP) and Liver Stiffness Assessed by Vibration-Controlled Transient Elastography. Clin Gastroenterol Hepatol. 2025. PMID: 41135666. doi:10.1016/j.cgh.2025.09.039
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。