Comparing Heterogenous Phenotypes of Chronic Obstructive Pulmonary Disease: Network Analysis and Penalized Generalized Linear Model

Abstract

Hyeon-Kyoung Koo,1 Sung Jun Chung,1 Dongil Park,2 Ho Cheol Kim,3 Hyewon Seo,4 Hyun Jung Kim,5 Hyoung Kyu Yoon,6 Chin Kook Rhee,7 Kwang Ha Yoo,8 Deog Kyeom Kim9 1Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Ilsan Paik Hospital, Inje University College of Medicine, Goyang, Republic of Korea; 2Department of Internal Medicine, Division of Pulmonology, College of Medicine, Chungnam National University, Deajeon, Republic of Korea; 3Department of Internal Medicine, Gyeongsang National University Changwon Hospital, Gyeongsang National University School of Medicine, Changwon, Republic of Korea; 4Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, School of Medicine, Kyungpook National University, Kyungpook National University Hospital, Daegu, Republic of Korea; 5Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Keimyung University School of Medicine, Dongsan Hospital, Daegu, Republic of Korea; 6Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, College of Medicine, Yeouido St Mary’s Hospital, The Catholic University of Korea, Seoul, Republic of Korea; 7Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea; 8Division of Pulmonary, Allergy and Critical Care Medicine, Department of Internal Medicine, Konkuk University School of Medicine, Seoul, Republic of Korea; 9Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul Metropolitan Government-Seoul National University Boramae Medical Center, Seoul National University College of Medicine, Seoul, Republic of KoreaCorrespondence: Deog Kyeom Kim, Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul Metropolitan Government-Seoul National University Boramae Medical Center, 20 Boramaero-5-Gil, Dongjak-Gu, Seoul, 07061, Republic of Korea, Tel [phone], Fax [phone], Email [email protected] and Objective: Chronic obstructive pulmonary disease (COPD) is a heterogeneous disease, with chronic bronchitis (CB) and emphysema phenotypes. The aim of our study was to compare the distinct patterns of correlation networks for respiratory symptoms and predictors of future exacerbations of different COPD phenotypes.Methods: CB and emphysema were identified using a questionnaire and computed tomography images, respectively, and also included patients with preserved ratio impaired spirometry (PRISm). We constructed separate correlation networks for each subgroup using Spearman correlation coefficients. Predictors of future exacerbations were selected via least absolute shrinkage and selection operation regression analyses in multivariable analysis.Results: Among the 3436 patients, 2232 were non-CB, 1131 were CB, 1116 were emphysema, and 73 were PRISm groups. The forced expiratory volume in one second (FEV1) and respiratory symptoms worsened in the following order: PRISm, non-CB, emphysema, and CB groups. During the 1-year follow-up, 17.3%, 21.3%, and 18.9% of patients in the non-CB, CB, and emphysema groups, respectively, experienced exacerbation. Each group showed a distinct correlation pattern between demographic characteristics, comorbidities, pulmonary function, blood biomarkers, respiratory symptoms, and exercise capacity. Across all groups, lower FEV1 (%), higher white blood cell count, higher erythrocyte sedimentation rate, and worse Saint George’s Respiratory Questionnaire symptom and total scores were identified as common risk factors for future exacerbations. However, each group showed distinct predictors for future exacerbations.Conclusion: The correlation network patterns and predictors of future exacerbations varied significantly depending on the COPD phenotype. Further research is required to understand the heterogeneous COPD pathophysiology and facilitate personalized medicine.Plain Language Summary: COPD has various subtypes, including chronic bronchitis, emphysema, and PRISm phenotypes. This study compared symptom patterns and predictors of future exacerbation in these groups. We analyzed data from over 3400 patients and observed that lung function and symptoms worsened in the following order: PRISm, non-chronic bronchitis, emphysema, and chronic bronchitis. Each group showed distinct patterns of relationships between demographics, lung function, biomarkers, and respiratory symptoms. Although some risk factors overlapped, each group had its own predictors for future exacerbation. Understanding these differences among subtypes could lead to better personalized treatments for COPD patients.Keywords: COPD, exacerbation, heterogeneous, network analysis

Keywords

• copd
• exacerbation
• heterogeneous
• network analysis