Clinical tool for disease phenotyping in granulomatous lung disease.

Autor: Silveira LJ; Division of Biostatistics and Informatics, National Jewish Health, Denver, Colorado, United States of America.; Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States of America., Strand M; Division of Biostatistics and Informatics, National Jewish Health, Denver, Colorado, United States of America.; Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States of America., Van Dyke MV; Colorado Department of Public Health and Environment, Denver, Colorado, United States of America., Mroz MM; Division of Environmental and Occupational Health, National Jewish Health, Denver, Colorado, United States of America., Faino AV; Division of Biostatistics and Informatics, National Jewish Health, Denver, Colorado, United States of America., Dabelea DM; Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States of America., Maier LA; Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States of America.; Division of Environmental and Occupational Health, National Jewish Health, Denver, Colorado, United States of America.; Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States of America., Fingerlin TE; Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States of America.; Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora, Colorado, United States of America.; Center for Genes, Environment and Health, National Jewish Health, Denver, Colorado, United States of America.
Jazyk: angličtina
Zdroj: PloS one [PLoS One] 2017 Nov 16; Vol. 12 (11), pp. e0188119. Date of Electronic Publication: 2017 Nov 16 (Print Publication: 2017).
DOI: 10.1371/journal.pone.0188119
Abstrakt: Background: Exposure to beryllium may lead to granuloma formation and fibrosis in those who develop chronic beryllium disease (CBD). Although disease presentation varies from mild to severe, little is known about CBD phenotypes. This study characterized CBD disease phenotypes using longitudinal measures of lung function.
Methods: Using a case-only study of 207 CBD subjects, subject-specific trajectories over time were estimated from longitudinal pulmonary function and exercise-tolerance tests. To estimate linear combinations of the 30-year values that define underlying patterns of lung function, we conducted factor analysis. Cluster analysis was then performed on all the predicted lung function values at 30 years. These estimates were used to identify underlying features and subgroups of CBD.
Results: Two factors, or composite measures, explained nearly 70% of the co-variation among the tests; one factor represented pulmonary function in addition to oxygen consumption and workload during exercise, while the second factor represented exercise tests related to gas exchange. Factors were associated with granulomas on biopsy, exposure, steroid use and lung inflammation. Three clusters of patients (n = 53, n = 59 and, n = 95) were identified based on the collection of test values. Lower levels of each of the factor composite scores and cluster membership were associated with baseline characteristics of patients.
Conclusions: Using factor analysis and cluster analysis, we identified disease phenotypes that were associated with baseline patient characteristics, suggesting that CBD is a heterogeneous disease with varying severity. These clinical tools may be used in future basic and clinical studies to help define the mechanisms and risk factors for disease severity.
Databáze: MEDLINE
Externí odkaz: