PDF(Pubmed)
Abstract：
UNASSIGNED: We explored the role of tumor size and number in the prognosis of HCC patients who underwent ablation and created a nomogram based on machine learning to predict the recurrence.
UNASSIGNED: A total of 990 HCC patients who underwent transcatheter arterial chemoembolization (TACE) combined ablation at Beijing Youan Hospital from January 2014 to December 2021 were prospectively enrolled, including 478 patients with single small HCC (S-S), 209 patients with single large (≥30mm) HCC (S-L), 182 patients with multiple small HCC (M-S), and 121 patients with multiple large HCC (M-L). S-S patients were randomized in a 7:3 ratio into the training cohort (N=334) and the validation cohort (N=144). Lasso-Cox regression analysis was carried out to identify independent risk factors, which were used to construct a nomogram. The performance of the nomogram was evaluated by C-index, receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) curves. Patients in the training and validation cohorts were divided into low-risk, intermediate-risk, and high-risk groups based on the risk scores of the nomogram.
UNASSIGNED: The median recurrence-free survival (mRFS) in S-S patients was significantly longer than the S-L, M-S, and S-L patients (P<0.0001). The content of the nomogram includes age, monocyte-to-lymphocyte (MLR), gamma-glutamyl transferase-to-lymphocyte (GLR), International normalized ratio (INR), and Erythrocyte (RBC). The C-index (0.704 and 0.71) and 1-, 3-, and 5-year AUCs (0.726, 0.800, 0.780, and 0.752, 0.761, 0.760) of the training and validation cohorts proved the excellent predictive performance of the nomogram. Calibration curves the DCA curves showed that the nomogram had good consistency and clinical utility. There were apparent variances in RFS between the low-risk, intermediate-risk, and high-risk groups (P<0.0001).
UNASSIGNED: S-S patients who underwent ablation had the best prognosis. The nomogram developed and validated in the study had good predictive ability for S-S patients.
UNASSIGNED: A total of 990 HCC patients who underwent transcatheter arterial chemoembolization (TACE) combined ablation at Beijing Youan Hospital from January 2014 to December 2021 were prospectively enrolled, including 478 patients with single small HCC (S-S), 209 patients with single large (≥30mm) HCC (S-L), 182 patients with multiple small HCC (M-S), and 121 patients with multiple large HCC (M-L). S-S patients were randomized in a 7:3 ratio into the training cohort (N=334) and the validation cohort (N=144). Lasso-Cox regression analysis was carried out to identify independent risk factors, which were used to construct a nomogram. The performance of the nomogram was evaluated by C-index, receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) curves. Patients in the training and validation cohorts were divided into low-risk, intermediate-risk, and high-risk groups based on the risk scores of the nomogram.
UNASSIGNED: The median recurrence-free survival (mRFS) in S-S patients was significantly longer than the S-L, M-S, and S-L patients (P<0.0001). The content of the nomogram includes age, monocyte-to-lymphocyte (MLR), gamma-glutamyl transferase-to-lymphocyte (GLR), International normalized ratio (INR), and Erythrocyte (RBC). The C-index (0.704 and 0.71) and 1-, 3-, and 5-year AUCs (0.726, 0.800, 0.780, and 0.752, 0.761, 0.760) of the training and validation cohorts proved the excellent predictive performance of the nomogram. Calibration curves the DCA curves showed that the nomogram had good consistency and clinical utility. There were apparent variances in RFS between the low-risk, intermediate-risk, and high-risk groups (P<0.0001).
UNASSIGNED: S-S patients who underwent ablation had the best prognosis. The nomogram developed and validated in the study had good predictive ability for S-S patients.
摘要：
■我们探讨了肿瘤大小和数量在接受消融的HCC患者预后中的作用，并基于机器学习创建了一个列线图来预测复发。
■前瞻性纳入2014年1月至2021年12月在北京佑安医院接受肝动脉化疗栓塞（TACE）联合消融的990例HCC患者。包括478例单发小肝癌（S-S）患者，209例单个大（≥30mm）HCC（S-L），182例多发性小肝癌（M-S），121例多发性大肝癌（M-L）患者。将S-S患者以7:3的比例随机分配到训练队列(N=334)和验证队列(N=144)中。采用Lasso-Cox回归分析确定独立危险因素,用于构造列线图。通过C指数评估列线图的性能，接收机工作特性(ROC)曲线，校正曲线,和决策曲线分析(DCA)曲线。培训和验证队列中的患者被分为低风险，中等风险，和基于列线图的风险评分的高风险组。
■S-S患者的中位无复发生存期（mRFS）明显长于S-L，M-S,和S-L患者(P<0.0001)。列线图的内容包括年龄，单核细胞对淋巴细胞（MLR），γ-谷氨酰转移酶转淋巴细胞(GLR)，国际标准化比率(INR)，和红细胞（RBC）。C指数(0.704和0.71)和1-，3-,训练和验证队列的5年AUC（0.726、0.800、0.780和0.752、0.761、0.760）证明了列线图的出色预测性能。校准曲线DCA曲线表明，列线图具有良好的一致性和临床实用性。低风险之间的RFS存在明显差异，中等风险，和高危人群(P<0.0001)。
■接受消融的S-S患者预后最好。在研究中开发和验证的列线图对S-S患者具有良好的预测能力。
■前瞻性纳入2014年1月至2021年12月在北京佑安医院接受肝动脉化疗栓塞（TACE）联合消融的990例HCC患者。包括478例单发小肝癌（S-S）患者，209例单个大（≥30mm）HCC（S-L），182例多发性小肝癌（M-S），121例多发性大肝癌（M-L）患者。将S-S患者以7:3的比例随机分配到训练队列(N=334)和验证队列(N=144)中。采用Lasso-Cox回归分析确定独立危险因素,用于构造列线图。通过C指数评估列线图的性能，接收机工作特性(ROC)曲线，校正曲线,和决策曲线分析(DCA)曲线。培训和验证队列中的患者被分为低风险，中等风险，和基于列线图的风险评分的高风险组。
■S-S患者的中位无复发生存期（mRFS）明显长于S-L，M-S,和S-L患者(P<0.0001)。列线图的内容包括年龄，单核细胞对淋巴细胞（MLR），γ-谷氨酰转移酶转淋巴细胞(GLR)，国际标准化比率(INR)，和红细胞（RBC）。C指数(0.704和0.71)和1-，3-,训练和验证队列的5年AUC（0.726、0.800、0.780和0.752、0.761、0.760）证明了列线图的出色预测性能。校准曲线DCA曲线表明，列线图具有良好的一致性和临床实用性。低风险之间的RFS存在明显差异，中等风险，和高危人群(P<0.0001)。
■接受消融的S-S患者预后最好。在研究中开发和验证的列线图对S-S患者具有良好的预测能力。
