轻中度哮喘支气管舒张试验预测模型构建研究  

Construction of models in predicting the result of bronchodilator reversibility test in mild to moderate asthmatics

作　　者：马梓文 张丽[1,2,3,4] 邓稞 王霁[1,2,4] 李敏 王刚[1,2,4] MA Ziwen;ZHANG Li;DENG Ke;WANG Ji;LI Min;WANG Gang(Department of Respiratory and Critical Care Medicine,Clinical Research Center for Respiratory Disease,West China Hospital,Sichuan University,Chengdu,Sichuan 610041,P.R.China;State Key Laboratory of Respiratory Health and Multimobidity,Chengdu,Sichuan 610041,P.R.China;Pneumology Group,Department of Integrated Traditional Chinese and Western Medicine,West China Hospital,Sichuan University,Chengdu,Sichuan 610041,P.R.China;Laboratory of Pulmonary Immunology and Inflammation,Frontiers Science Center for Disease-related Molecular Network,Sichuan University,Chengdu,Sichuan 610041,P.R.China;Department of Respiratory and Critical Care Medicine,First Affiliated Hospital of Kunming Medical University,Kunming,Yunnan 650032,P.R.China)

机构地区：[1]四川大学华西医院呼吸与危重症医学科呼吸疾病临床医学研究中心,四川成都610041 [2]呼吸和共病全国重点实验室,四川成都610041 [3]四川大学华西医院中西医结合内科,四川成都610041 [4]四川大学教育部疾病分子网络前沿科学研究中心肺免疫炎症研究室,四川成都610041 [5]昆明医科大学第一附属医院呼吸与危重症医学科,云南昆明650032

出　　处：《中国呼吸与危重监护杂志》2023年第3期189-196,共8页Chinese Journal of Respiratory and Critical Care Medicine

基　　金：国家自然科学基金国际合作重点项目(81920108002);国家自然科学基金青年基金项目(82100032);四川省科技厅重点项目(2022YFS0263);成都市科技局国家级国际科技合作项目(2021-GH03-00007-HZ)。

摘　　要：目的探索轻中度气道阻塞哮喘患者支气管舒张试验(bronchodilator reversibility test,BDT)结果的影响因素,并建立一个数学模型以预测患者BDT结果。方法采用横断面设计纳入2014年—2021年澳亚重症哮喘协作网通过支气管激发试验确诊哮喘且第1秒用力呼气容积(forced expiratory volume in the first second,FEV_(1))占预计值百分比(FEV_(1)%pred)≥70%的患者,进行结构问卷调查、BDT、呼出气一氧化氮(fractional concentration of exhaled nitric oxide,FeNO)检测、痰诱导及外周血常规检测。根据BDT结果将患者分为阳性组与阴性组并进行组间比较、相关性分析及多元回归,通过LASSO回归筛选变量构建多元logistic回归模型。结果共纳入334例轻中度哮喘患者。与BDT阴性组(n=240)比较,BDT阳性组(n=94)的肺功能、哮喘控制和生活质量更差、外周血和诱导痰嗜酸性粒细胞及FeNO更高。多元回归分析显示,哮喘控制问卷得分、哮喘相关生活质量问卷得分、FEV_(1)%pred、最大中段呼气流速(maximal mid-expiratory flow,MMEF)占预计值百分比(MMEF%pred)、FEV_(1)与用力肺活量(forced vital capacity,FVC)的比值(FEV_(1)/FVC)、外周血和诱导痰嗜酸性粒细胞及FeNO与BDT阳性结果存在明显的相关性。模型训练集共纳入326例患者,LASSO筛选发现FEV_(1)%pred、MMEF%pred、FEV_(1)/FVC、吸烟包年数、外周血和诱导痰嗜酸性粒细胞计数、FeNO为稳定预测变量,以此构建3个预测模型(P<0.001)的受试者操作特征曲线下面积为0.810~0.834,内部验证示模型区分度(分别为0.810、0.834、0.812)、校准度(分别为0.135、0.133、0.192)均较良好。结论轻中度哮喘BDT结果与哮喘控制、肺功能水平、全身或诱导痰嗜酸性粒细胞炎症和FeNO相关,建立包括肺功能、嗜酸性粒细胞、FeNO等的模型,可以较好预测BDT结果。Objective To investigate factors influencing the results of bronchodilator reversibility tests(BDT)in mild to moderate asthma,and to develop a model predicting the result of BDT in this population.Methods A crosssectional study was designed to recruit patients with forced expiratory volume in the first second(FEV_(1))≥70%predicted from the Australasian Severe Asthma Network during 2014 to 2021,whose asthma diagnosis was confirmed by a positive bronchial challenge test.Structural questionnaires,BDT,fractional exhaled nitric oxide(FeNO),induced sputum and peripheral blood sampling were conducted.Patients were further divided into positive group and negative group according to their BDT result.Then the comparative analysis between two groups,correlation analysis,and multivariate logistical regression were performed.Logistic models for predicting BDT result were developed using variables screened through LASSO regression.Results A total of 334 patients were included.Compared with the BDT negative group(n=240),the BDT positive group(n=94)was found to have worse airway obstruction in lung function,asthma control and quality of life,higher eosinophil counts in both peripheral blood and induced sputum,and higher FeNO.According to the multivariate regression,the positive BDT results significantly correlated with Asthma Control Questionnaire score,Asthma Questionnaire of Life Quality score,FEV_(1)%pred,MMEF%pred,FEV_(1)/FVC,blood and sputum eosinophil counts and FeNO.A total of 326 patients were included in the training set,and FEV_(1)%pred,MMEF%pred,FEV_(1)/FVC,smoking pack years,blood and sputum eosinophil counts and FeNO were then screened out by LASSO regression as stable predictors.The areas under the receiver operating characteristic curve of the 3 prediction models(P<0.001)constructed using the variables above ranged from 0.810 to 0.834.Internal validation was performed,and both the discrimination(0.810,0.834 and 0.812,respectively)and the calibration(0.135,0.133 and 0.192,respectively)of the models were acceptab

关 键 词：哮喘 支气管舒张试验 预测模型 嗜酸性粒细胞炎症 

分 类 号：R562.25[医药卫生—呼吸系统]