Hepatobiliary Center, The First Affiliated Hospital of Nanjing Medical University, Key Laboratory of Liver Transplantation, Chinese Academy of Medical Sciences, NHC Key Laboratory of Living Donor Liver Transplantation (Nanjing Medical University), Nanjing, 210029, China.
Department of General Surgery, Nanjing Drum Tower Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Graduate School of Peking Union Medical College, Nanjing, 210008, China.
World J Surg Oncol. 2024 Sep 7;22(1):240. doi: 10.1186/s12957-024-03523-1.
Postoperative recurrence is a vital reason for poor 5-year overall survival in hepatocellular carcinoma (HCC) patients. The ADV score is considered a parameter that can quantify HCC aggressiveness. This study aimed to identify HCC patients at high-risk of recurrence early using the ADV score.
The medical data of consecutive HCC patients undergoing hepatectomy from The First Affiliated Hospital of Nanjing Medical University (TFAHNJMU) and Nanjing Drum Tower Hospital (NJDTH) were retrospectively reviewed. Based on the status of microvascular invasion and the Edmondson-Steiner grade, HCC patients were divided into three groups: low-risk group (group 1: no risk factor exists), medium-risk group (group 2: one risk factor exists), and high-risk group (group 3: coexistence of two risk factors). In the training cohort (TFAHNJMU), the R package nnet was used to establish a multi-categorical unordered logistic regression model based on the ADV score to predict three risk groups. The Welch's T-test was used to compare differences in clinical variables in three predicted risk groups. NJDTH served as an external validation center. At last, the confusion matrix was developed using the R package caret to evaluate the diagnostic performance of the model.
350 and 405 patients from TFAHNJMU and NJDTH were included. HCC patients in different risk groups had significantly different liver function and inflammation levels. Density maps demonstrated that the ADV score could best differentiate between the three risk groups. The probability curve was plotted according to the predicted results of the multi-categorical unordered logistic regression model, and the best cut-off values of the ADV score were as follows: low-risk ≤ 3.4 log, 3.4 log < medium-risk ≤ 5.7 log, and high-risk > 5.7 log. The sensitivities of the ADV score predicting the high-risk group (group 3) were 70.2% (99/141) and 78.8% (63/80) in the training and external validation cohort, respectively.
The ADV score might become a valuable marker for screening patients at high-risk of HCC recurrence with a cut-off value of 5.7 log, which might help surgeons, pathologists, and HCC patients make appropriate clinical decisions.
术后复发是影响肝细胞癌（HCC）患者 5 年总生存率的重要原因。ADV 评分被认为是量化 HCC 侵袭性的参数。本研究旨在使用 ADV 评分早期识别 HCC 患者的高复发风险。
回顾性分析南京医科大学第一附属医院（TFAHNJMU）和南京鼓楼医院（NJDTH）连续行肝切除术的 HCC 患者的临床资料。根据微血管侵犯和 Edmondson-Steiner 分级的状态，将 HCC 患者分为低危组（组 1：无风险因素存在）、中危组（组 2：存在一个风险因素）和高危组（组 3：存在两个风险因素共存）。在训练队列（TFAHNJMU）中，使用 R 包 nnet 基于 ADV 评分建立多类别无序逻辑回归模型来预测三个风险组。采用 Welch's T 检验比较三个预测风险组的临床变量差异。NJDTH 作为外部验证中心。最后，使用 R 包 caret 开发混淆矩阵来评估模型的诊断性能。
TFAHNJMU 和 NJDTH 分别纳入 350 例和 405 例患者。不同风险组的 HCC 患者肝功能和炎症水平存在显著差异。密度图表明 ADV 评分可以最好地区分三组风险。根据多类别无序逻辑回归模型的预测结果绘制概率曲线，ADV 评分的最佳截断值如下：低危组≤3.4 log，3.4 log<中危组≤5.7 log，高危组>5.7 log。ADV 评分预测高危组（组 3）的灵敏度在训练和外部验证队列中分别为 70.2%（99/141）和 78.8%（63/80）。
ADV 评分可能成为筛选 HCC 复发高危患者的有价值标志物，截断值为 5.7 log，有助于外科医生、病理学家和 HCC 患者做出适当的临床决策。
