2025 年第 3 期 第 20 卷

探索60岁以上服用利伐沙班治疗的患者抗Ⅹa因子活性与血药浓度的相关性

Exploration of the correlation between anti-factor Ⅹa activity and blood concentration in patients over 60 years old treated with rivaroxaban

作者：张瑞琪杜佳丽刘梅林

英文作者：Zhang Ruiqi Du Jiali Liu Meilin

单位：北京大学第一医院老年病内科，北京100034

英文单位：Department of Geriatric Medicine Peking University First Hospital Beijing 100034 China

关键词：直接口服抗凝药；利伐沙班；抗Ⅹa因子活性；血药浓度

英文关键词：Directoralanticoagulants;Rivaroxaban;Anti-factorⅩaactivity;Bloodconcentration

• 摘要：

• 目的 在接受利伐沙班治疗的60岁以上老年患者中，评估抗Ⅹa因子活性（AXA）与利伐沙班血药浓度的相关性，探索建立AXA与血药浓度的线性回归模型。方法 本研究为前瞻性队列研究，纳入2022年3月至2023年3月在北京大学第一医院老年病内科住院，具有抗凝治疗指征，接受利伐沙班治疗的50例患者，年龄≥60岁，临床资料完整。连续服用利伐沙班至少3 d后，抽血检测AXA谷值、峰值及血药浓度谷值、峰值。采用线性回归模型评估AXA与血药浓度的关系。结果 在研究人群中，AXA谷值与血药浓度谷值的散点图可见二者呈线性相关。以服用利伐沙班3 d后AXA谷值为自变量，血药浓度谷值为因变量，构建线性回归模型可得：①ｙ＾=10.378+160.185×AXA谷值。即AXA谷值每升高1 kIU/L，血药浓度谷值增加170.563 μg/L。其中，决定系数为0.730，德宾-沃森检验值为1.845，模型统计检验F=129.538，P＜0.001，残差符合正态分布。AXA峰值与血药浓度峰值的散点图可见二者呈线性相关。以服用利伐沙班3 d后AXA峰值为自变量，血药浓度峰值为因变量，构建线性回归模型可得：②ｙ＾=33.600+119.449×AXA峰值。即AXA峰值每升高1 kIU/L，血药浓度峰值增加153.049 μg/L。其中，决定系数为0.656，德宾-沃森检验值为2.083，模型统计检验F=91.655，P＜0.001，残差符合正态分布。结论 AXA与利伐沙班血药浓度之间存在线性回归模型，在临床实践中，可通过AXA与血药浓度的线性回归模型，计算得出血药浓度预测值，与目标值相比来评估抗凝强度，指导进一步治疗。

• Objective  To evaluate the association between anti-factor Ⅹa activity (AXA) and rivaroxaban blood concentration in elderly patients over 60 years old treated with rivaroxaban, and to establish a linear regression model for the relationship between AXA and blood concentration. Methods This study was a prospective cohort study. Fifty patients aged ≥60 years with complete clinical data and indications for anticoagulant therapy who were hospitalized in the Department of Geriatric Medicine, Peking University First Hospital from March 2022 to March 2023 and received rivaroxaban were enrolled. After continuous administration of rivaroxaban for at least 3 days, blood samples were drawn to detect the trough and peak values of AXA and the trough and peak values of blood drug concentration. Linear regression models were used to evaluate the relationship between AXA and blood concentration. Results In the study population, there was a linear correlation between the trough value of AXA and the trough value of blood concentration. Taking the AXA trough value of rivaroxaban 3 days after administration as the independent variable and the trough value of blood drug concentration as the dependent variable, the linear regression model was constructed, and the following results  were obtained: ①ｙ＾=10.378+160.185×trough value of AXA. That was, for every 1 kIU/L increase in the trough value of AXA, the trough value of blood concentration increased by 170.563 μg/L. The determination coefficient was 0.730, the Durbin-Watson test value was 1.845, the model statistical test F=129.538, P＜0.001, and the residual was in accordance with normal distribution. There was a linear correlation between the peak value of AXA and the peak value of blood concentration. Taking the peak value of AXA after taking rivaroxaban for 3 days as the independent variable and the peak value of blood drug concentration as the dependent variable, the linear regression model was constructed, and the following results  were obtained: ②ｙ＾=33.600+119.449×peak value of AXA. That was, when the peak value of AXA increased by 1 kIU/L, the peak blood concentration increased by 153.049 μg/L. The determination coefficient was 0.656, the Durbin-Watson test value was 2.083, the model statistical test F=91.655, P＜0.001, and the residual was in accordance with normal distribution. Conclusion There is a linear regression model between AXA and blood concentration of rivaroxaban. In clinical practice, the predicted bleeding concentration can be calculated by the linear regression model between AXA and blood concentration, which can be compared with the target value to evaluate the anticoagulant strength and guide further treatment.