急性早幼粒细胞白血病早期死亡风险分级法构建  

Establishment of An Risk Grading Method for Early Death in Patients with Acute Promyelocytic Leukemia

作　　者：杨荟圆 路明 闫丽儒[3] 侯文宜[1] 张迎媚[1] 周晋[1] Yang Huiyuan;Lu Ming;Yan Liru(The First Affiliated Hospital of Harbin Medical University,Heilongjiang 150001,China)

机构地区：[1]哈尔滨医科大学附属第一医院血液内科,150001 [2]黑龙江省妇幼保健院,哈尔滨150030 [3]哈尔滨医科大学附属第四医院干部门诊

出　　处：《医学研究杂志》2020年第5期104-108,68,共6页Journal of Medical Research

基　　金：黑龙江省自然科学基金资助项目(H2017032)。

摘　　要：目的构建急性早幼粒细胞白血病(APL)早期死亡(ED)风险分级法,用独立的患者集验证该分级法的预测效果。方法回顾性分析了笔者医院近8年收治的285例连续的、亚砷酸(ATO)单药诱导治疗的初发APL患者的临床数据。患者随机分为训练集(230例)和验证集(55例)。性别、年龄、天冬氨酸氨基转移酶、白蛋白、肌酐、纤维蛋白原(FIB)和D-二聚体水平、外周血白细胞(WBC)计数和血小板(PLT)计数9个变量纳入研究。将所有的数值变量进行二分类转化。各个变量与ED之间相关性的单因素和多因素分析均采用Logistic回归模型。使用Hosmer-Lemeshow检验对模型拟合度进行检测,采用受试者工作特征(ROC)曲线下面积对模型的预测效能进行评估。结果经过单因素和多因素分析,最终性别、年龄、血浆肌酐、FIB、D-二聚体水平和WBC计数6个变量进入模型。经训练集和外部验证集分别验证,证实该模型具有良好的拟合度和ED预测效能。采用训练集数据,利用该模型构建了ED风险分级法。单因素分析结果显示,无论在训练集还是在独立的外部验证集中,该ED风险分级法都与ED的发生密切相关(P<0.05)。结论针对ATO单药诱导治疗的APL患者,本研究建立了ED风险分级法,该方法可以有效地在患者入院早期即快速鉴别出发生ED风险高的APL患者。Objective To construct a risk grading method for early death(ED)in patients with acute promyelocytic leukemia(APL),then to validate the predictive effect of the grading method in a independent patient cohort.Methods We retrospectively analyzed the clinical data of 285 consecutive patients with newly diagnosed APL who were treated with arsenic trioxide(ATO)alone as induction therapy during the past 8 years in our hospital.These patients were randomly divided into a training cohort(n=230)and a validation cohort(n=55).Nine baseline clinical variables were selected for the analysis,including gender,age,aspartate aminotransferase,albumin,creatinine,fibrinogen,D-dimer,peripheral blood white blood cells(WBC)count and platelet(PLT)count.All numerical variables were converted into binary variables.Univariate and multivariate Logistic regression models were used to analyze the correlation between variable and ED.The Hosmer-Lemeshow test was used to detect the model fitting degree,and the area under the receiver operator characteristic(ROC)curve was used to evaluate the predictive effectiveness of the model.Results By univariate and multivariate analyses,6 variables entered the Logistic regression model,including male,age,creatinine,fibrinogen,D-dimer,and WBC count.The model showed good fitting degree and predictive efficiency in both the training set and external validation set.Based on the Logistic regression model,a risk grading method for ED was constructed using the training data set.The univariate analysis revealed that ED risk grading by this method was significant correlated with ED occurrence in both the training and independent external validation cohorts(P<0.05).Conclusion In this study,a risk grading method for ED was established in APL patients who treated with ATO along as induction therapy.The risk grading method could effectively identify a cohort of patients at high risk of ED at admission.

关 键 词：急性早幼粒白血病 早期死亡 风险分级 三氧化二砷 

分 类 号：R733.7[医药卫生—肿瘤]