AQP4-IgG阳性视神经脊髓炎谱系病患者免疫抑制治疗后复发危险因素及预测模型构建  

Recurrence Risk Factors and Prediction Model Construction of Patients with AQP4-IgG Positive Neuromyelitis Optica Spectrum Disorder after Immunosuppressive Therapy

作　　者：李小琴[1] 吴燕平 李素萍 LI Xiao-qin;WU Yan-ping;LI Su-ping(Department of Neurology,the First People's Hospital of Yibin,Yibin 644000,China;Department of Neurology,Sichuan Provincial People's Hospital,Chengdu 610000,China)

机构地区：[1]四川省宜宾市第一人民医院神经内科,宜宾644000 [2]四川省医学科学院·四川省人民医院(电子科技大学附属医院)神经内科,成都610000

出　　处：《中国临床神经科学》2024年第5期520-528,共9页Chinese Journal of Clinical Neurosciences

基　　金：宜宾市卫生健康委项目(编号:2020YW057)。

摘　　要：目的探讨水通道蛋白4抗体(AQP4-IgG)阳性的视神经脊髓炎谱系疾病(NMOSD)患者免疫抑制治疗后复发的危险因素,并基于随机森林算法建立AQP4-IgG阳性NMOSD患者免疫抑制治疗后复发的随机森林模型。方法回顾性分析2019年1月至2022年3月收治的行免疫抑制治疗的205例AQP4-IgG阳性NMOSD患者的临床资料,根据首次治疗后1年内是否复发将其分为:复发组64例(31.22%)和非复发组141例(68.78%)。采用多因素Logistic回归分析法筛选AQP4-IgG阳性NMOSD患者复发的危险因素,运用R软件建立预测AQP4-IgG阳性NMOSD患者复发的随机森林模型并验证预测模型效能。结果两组患者的单因素分析显示,年龄、性别、劳累史、前驱感染、脊髓受累≥6个节段例数比较(均P<0.05)。多因素Logistic回归分析显示,年龄≥40岁(OR=3.507,P=0.001)、女性(OR=3.507,P<0.001)、有劳累史(OR=3.502,P=0.001)、前驱感染(OR=2.852,P=0.021)、脊髓受累≥6个节段(OR=2.700,P=0.004)为AQP4-IgG阳性NMOSD患者复发的独立危险因素。预测AQP4-IgG阳性NMOSD患者复发的相对重要预测因子依次排序为:脊髓受累≥6个节段、年龄、劳累史、前驱感染、性别;基于上述因素构建的随机森林模型ROC值稍高于Logistic回归模型(Z=1.878,P=0.060);5折交叉验证的模型正确率73.1%。结论年龄≥40岁、女性、劳累史、前驱感染、脊髓受累≥6个节段均为AQP4-IgG阳性NMOSD患者复发的独立危险因素,基于上述独立危险因素构建的随机森林模型具有较好的风险预测效能。Aim To investigate the risk factors of relapse in patients with aquaporin-4 antibody(AQP4-IgG)positive optic neuromyelitis spectrum disease(NMOSD)after immunosuppressive therapy,and establish a random forest model of relapse in patients with AQP4-IgG positive NMOSD after immunosuppressive therapy based on random forest algorithm.Methods The clinical data of 205patients with AQP4-IgG-positive NMOSD who received immunosuppressive therapy from January 2019to March 2022 were retrospectively analyzed.According to whether they relapsed within 1 year after the first treatment,the patients were divided into a relapsed group(64 cases,31.22%)and a non-relapsed group(141 cases,68.78%).Multivariate Logistic regression analysis was used to screen the risk factors for recurrence in AQP4-Ig G positive NMOSD patients.R software was used to establish a random forest model for predicting recurrence in AQP4-IGG positive NMOSD patients and verify the efficacy of the prediction model.Results Univariate analysis of the two groups showed that age,sex,fatigue history,precursors infection,spinal cord involvement≥6 segments were compared(all P<0.05).Multivariate Logistic regression analysis showed that age≥40 years(OR=3.507;P=0.001),female(OR=3.507;P<0.001),a history of fatigue(OR=3.502;P=0.001),preinfection(OR=2.852;P=0.021),spinal cord involvement≥6 segments(OR=2.700;P=0.004)were independent risk factors for recurrence in AQP4-Ig G positive NMOSD patients.The relative important predictors of recurrence in AQP4-Ig G positive NMOSD patients were ranked as follows:spinal cord involvement≥6 segments,age,fatigue history,preinfection,sex.The ROC value of the random forest model based on the above factors was slightly higher than that of the Logistic regression model(Z=1.878,P=0.060).The accuracy rate of the model was73.1%.Conclusion Age≥40 years old,female,fatigue history,preinfection,spinal cord involvement≥6 segments are all independent risk factors for recurrence in AQP4-Ig G positive NMOSD patients,and the random forest model

关 键 词：水通道蛋白4抗体阳性 视神经脊髓炎谱系疾病 复发 危险因素 预测模型 

分 类 号：R742[医药卫生—神经病学与精神病学]