系统聚类分析探讨老年人呼吸道疾病的临床表型

Investigation of distinct clinical phenotypes of airways disease in the elderly based on hierarchical cluster analysis

目的采用系统聚类分析对老年呼吸道疾病患者的临床特征进行识别及分组，形成表型。方法前瞻性纳入67例有喘息症状的老年患者，收集人口统计学资料、呼吸道症状、吸烟年数、急性加重情况、特应性症状、呼气峰流速日志等。患者于病情稳定期进行肺功能检查，并检测血清总IgE水平、外周血嗜酸性粒细胞计数，对患者进行系统聚类分析。结果经系统聚类分析可将患者分为4类：（1）不吸烟、肺功能正常、血清总IgE升高、具有机体特应性的哮喘患者；（2）不吸烟、肺功能正常、仅有喘息症状的患者；（3）吸烟、重度气流受限、生活质量差的慢性阻塞性肺疾病（慢阻肺）患者；（4）吸烟、有气流受限、血清总IgE明显升高的哮喘一慢阻肺重叠综合征患者。4种表型患者在用药前第1秒用力呼气容积（FEV1）／用力肺活量（FVC）、FEV1／预计值、FEV1改善率和最大呼气中段流速（MMEF）／预计值、一氧化碳弥散系数（DLCO）／残气量（VA）／预计值、残气量（RV）／预计值、血清总IgE水平、累计吸烟量、圣乔治呼吸系统问卷（SGRQ）评分差异有统计学意义（P〈0．05或P〈0．01）。结论聚类分析可识别老年呼吸道疾病患者不同的临床表型，与单纯哮喘或慢阻肺患者相比，哮喘慢阻肺重叠综合征患者肺功能下降明显，急性加重次数多，生活质量最差，需要引起重视。

Objective To explore the clinical phenotype of airways disease in elderly patients using hierarchical cluster analysis. Methods A total of 67 elderly patients with respiratory symptoms were enrolled in a prospective study. Demographic and clinical data, such as respiratory symptoms, cumulative tobacco cigarette consumption, acute exacerbation, atopic symptoms and peak flow diary were collected. Pulmonary function tests, blood tests （total serum IgE level and blood eosinophil level） were performed in each patient during the stable stage. Then patients with different clinical phenotype were identified by hierarchical cluster analysis. Results Four clusters were identified with the following characteristics by hierarchical cluster analysis： cluster 1, atopic patients with no smoking, normal lung funetion, but increased total serum IgE levels and asthma symptom; cluster 2, patients with no smoking and normal pulmonary function with wheezing but without chronic cough; cluster 3, patients with chronic obstructive pulmonary disease and smoking, severe airflow limitation and poor quality of life; cluster 4, patients with asthma-chronic obstructive pulmonary disease overlap syndrome and smoking, airflow limitation and increased total serum IgE levels. The forced expiratory volume in 1 second （FEV1）/ forced vital capacity （FVC） ratio, FEV1/predicted value, rate of FEV1 change, maximal mid-expiratory flow （MMEF）/ predicted value, the diffusion lung capaeity for carbon monoxide （DLCO）/ alveolar volume （VA）/predicted value, residual volume （RV）/ predicted value, total serum Ig E levels, cumulative tobacco cigarette consumption, the St. George＇s Respiratory Questionnaire （SGRQ） score had significant differences in patients before versus after treatment （all P〈0. 05 or P〈0.01）. Conclusions Based on hierarchical cluster analysis, distinct clinical phenotypes of airways disease in elderly patients can be identified. Conclusions With patients having asthma or COPD alone, patients with Asthma-COPD overlap syndrome （ACOS） always experience a more rapid decline in lung function and frequent exacerbations, having poor health related qualit^of-life （HRQOL） outcomes, which deserve our high attention.