预测韩国和白种人慢性乙型肝炎患者肝癌风险的人工智能模

StephenW

预测韩国和白种人慢性乙型肝炎患者肝癌风险的人工智能模型

    金辉英†
    彼得罗·兰佩蒂科 †
    俊烈南†
    尹正焕
    George V. Papatheodoridis ‡
    Jeong-Hoon Lee ‡
    显示所有作者
    显示脚注

发布时间：2021 年 10 月 1 日 DOI：https://doi.org/10.1016/j.jhep.2021.09.025
PlumX 指标

    强调
    图形概要
    关键词
    参考
    文章信息
    相关文章

强调

    •
    使用机器学习算法在接受抗病毒治疗的慢性乙型肝炎患者中开发了一种新的 HCC 预测模型 (PLAN-B)。
    •
    该模型的实用性在独立的韩国和高加索人群中得到验证。
    •
    PLAN-B 包括 10 个基线参数：肝硬化、年龄、血小板计数、ETV/TDF、性别、血清 ALT 和 HBV DNA、白蛋白和胆红素水平以及 HBeAg 状态。
    •
    PLAN-B 模型对 HCC 发展表现出令人满意的预测性能，并且优于其他风险评分。

背景与目标
最近开发了几种模型来预测慢性乙型肝炎 (CHB) 患者发生肝细胞癌 (HCC) 的风险。我们的目标是开发和验证 HCC 风险的人工智能辅助预测模型。
方法
使用梯度增强机 (GBM) 算法，使用来自韩国 4 家医院的 6,051 名接受恩替卡韦或替诺福韦治疗的慢性乙型肝炎患者开发了一个模型。独立建立了两个外部验证队列：韩国人（来自 14 个韩国中心的 5,817 名患者）和高加索人（来自 11 个西方中心的 1,640 名患者）PAGE-B 队列。主要结果是HCC的发展。
结果
在推导队列和 2 个验证队列中，基线时 26.9%–50.2% 的患者存在肝硬化。推导出了一个使用 10 个基线参数的模型，并显示出良好的预测性能（c-index 0.79）。该模型在韩国人（c-index 0.79 vs. 0.64-0.74；所有 p <0.001）和高加索人中都显示出比以前的模型（PAGE-B、修改的 PAGE-B、REACH-B 和 CU-HCC）更好的辨别力验证队列（c-index 0.81 vs. 0.57-0.79；所有 p <0.05，除了修改后的 PAGE-B，p = 0.42）。校准图显示了令人满意的校准功能。将患者分为 4 个风险组时，在 8 年的随访期间，最低风险组（韩国队列的 11.2% 和高加索队列的 8.8%）发生 HCC 的风险低于 0.5%。
结论
这种基于 GBM 的模型为接受恩替卡韦或替诺福韦治疗的韩国和高加索慢性乙肝患者的 HCC 风险提供了最佳预测能力。
总结
已经开发了风险评分来预测慢性乙型肝炎患者患肝细胞癌 (HCC) 的风险。我们使用机器学习算法在 13,508 名接受抗病毒治疗的慢性乙型肝炎患者中开发并验证了一种新的风险预测模型。我们的新模型，基于 10 个常见的基线特征，与之前的风险评分相比，在风险分层方面表现出优异的表现。该模型还确定了一组患 HCC 风险最小的患者，这些患者可以进行不太密集的 HCC 监测。

StephenW

An artificial intelligence model to predict hepatocellular carcinoma risk in Korean and Caucasian patients with chronic hepatitis B

    Hwi Young Kim †
    Pietro Lampertico †
    Joon Yeul Nam †
    Jung-Hwan Yoon
    George V. Papatheodoridis ‡
    Jeong-Hoon Lee ‡
    Show all authors
    Show footnotes

Published:October 01, 2021DOI:https://doi.org/10.1016/j.jhep.2021.09.025
PlumX Metrics

    Highlights
    Graphical abstract
    Keywords
    References
    Article Info
    Related Articles

Highlights

    •
    A new HCC prediction model (PLAN-B) was developed using machine learning algorithms in antiviral-treated patients with chronic hepatitis B.
    •
    The utility of the model was validated in independent Korean and Caucasian cohorts.
    •
    PLAN-B comprises 10 baseline parameters: cirrhosis, age, platelet count, ETV/TDF, sex, serum ALT and HBV DNA, albumin and bilirubin levels, and HBeAg status.
    •
    The PLAN-B model demonstrated satisfactory predictive performance for HCC development and outperformed other risk scores.

Background & Aims
Several models have recently been developed to predict risk of hepatocellular carcinoma (HCC) in patients with chronic hepatitis B (CHB). Our aims were to develop and validate an artificial intelligence-assisted prediction model of HCC risk.
Methods
Using a gradient-boosting machine (GBM) algorithm, a model was developed using 6,051 patients with CHB who received entecavir or tenofovir therapy from 4 hospitals in Korea. Two external validation cohorts were independently established: Korean (5,817 patients from 14 Korean centers) and Caucasian (1,640 from 11 Western centers) PAGE-B cohorts. The primary outcome was HCC development.
Results
In the derivation cohort and the 2 validation cohorts, cirrhosis was present in 26.9%–50.2% of patients at baseline. A model using 10 parameters at baseline was derived and showed good predictive performance (c-index 0.79). This model showed significantly better discrimination than previous models (PAGE-B, modified PAGE-B, REACH-B, and CU-HCC) in both the Korean (c-index 0.79 vs. 0.64–0.74; all p <0.001) and Caucasian validation cohorts (c-index 0.81 vs. 0.57–0.79; all p <0.05 except modified PAGE-B, p = 0.42). A calibration plot showed a satisfactory calibration function. When the patients were grouped into 4 risk groups, the minimal-risk group (11.2% of the Korean cohort and 8.8% of the Caucasian cohort) had a less than 0.5% risk of HCC during 8 years of follow-up.
Conclusions
This GBM-based model provides the best predictive power for HCC risk in Korean and Caucasian patients with CHB treated with entecavir or tenofovir.
Lay summary
Risk scores have been developed to predict the risk of hepatocellular carcinoma (HCC) in patients with chronic hepatitis B. We developed and validated a new risk prediction model using machine learning algorithms in 13,508 antiviral-treated patients with chronic hepatitis B. Our new model, based on 10 common baseline characteristics, demonstrated superior performance in risk stratification compared with previous risk scores. This model also identified a group of patients at minimal risk of developing HCC, who could be indicated for less intensive HCC surveillance.

StephenW