急性早幼粒细胞白血病预后生存模型的建立

Establishment of survival model for predicting prognosis of patients with acute promyelocytic leukemia

目的分析影响急性早幼粒细胞白血病(APL)预后的危险因素,并以多参数指标建立APL患者的预后生存模型。方法选取2013年7月—2017年7月浙江大学医学院附属第一医院218例APL患者作为研究对象,将患者分为早期死亡组(18例)和非早期死亡组(200例),分析影响APL预后的危险因素,并建立预后生存模型。结果年龄>60岁、白细胞(WBC)计数>10×10~9/L、纤维蛋白原(Fib)≤1.4 g/L、血清肌酐(SCr)>105μmol/L、血清乳酸脱氢酶(LDH)>250 U/L、早幼粒细胞白血病-维甲酸受体α(PML-RARa)融合基因SG阳性与APL患者早期死亡密切相关(P<0.05);多因素Logistic回归分析发现,年龄[比值比(OR)=1.049,95%可信区间(CI)为1.003~1.098,P=0.038]和WBC计数(OR=1.030,95%CI为1.007~1.054,P=0.012)为APL患者预后的独立危险因素;预后生存模型根据初诊时APL患者年龄、WBC计数、Fib、LDH、SCr、PML-RARa融合基因亚型6项指标建立,其受试者工作特征(ROC)曲线的曲线下面积(AUC)为0.797,敏感性为71.4%,特异性为92.6%。结论年龄、WBC计数为APL预后的独立危险因素,所建立的预后生存模型对预测患者生存状态有重要意义。

Objective To analyze the risk factors for the prognosis of acute promyelocytic leukemia(APL),and to establish the survival model of APL with multiple parameters.Methods A total of 218 patients with APL treated in our hospital from July 2013 to July 2017 were enrolled.The patients were divided into early death group(18 cases)and non-early death group(200 cases).The risk factors for the prognosis of APL were analyzed,and the survival model was established.Results Patient’s age>60 years old,white blood cell(WBC)count>10×10~9/L,fibrinogen(Fib)≤1.4 g/L,serum creatinine(SCr)>105μmol/L,lactate dehydrogenase(LDH)>250 U/L,positive PML-RARαfusion gene SG were closely related to early death(P<0.05).Multivariate Logistic analysis showed that age[odds ratio(OR)=1.049,95%confidence interval(CI)1.003-1.098,P=0.038]and WBC count(OR=1.030,95%CI 1.007-1.054,P=0.012)were independent risk factors for APL prognosis.The model for predicting the patient’s survival was established by the indicators of age,WBC count,Fib,LDH,SCr,and PMLRARαfusion gene subtype at initial diagnosis,and evaluated by receiver operating characteristic(ROC)curve.The area under curve(AUC)was 0.797,the sensitivity was 71.4%,and the specificity was 92.6%.Conclusions Age and WBC count are independent risk factors for APL prognosis.The established model can be explored in predicting the survival status of patients.