ATS 2021 will feature presentations of original research from accepted abstracts. Mini Symposia and Thematic Poster Sessions are abstract based sessions.
Please use the form below to browse scientific abstracts and case reports accepted for ATS 2021. Abstracts presented at the ATS 2021 will be published in the Online Abstract Issue of the American Journal of Respiratory and Critical Care Medicine, Volume 203, May 3, 2021.
Search Tips:
- Use the keyword search to search by keyword or author's name.
- Filter your search results by selecting the checkboxes that apply.
- Click on "Clear" to clear the form and start a new search. .
Search results will display below the form.
Identification of Subjects with High Risk of Disease Progression Using Data from the COPDGene Cohort
Session Title
TP41 - TP041 DIAGNOSIS AND RISK ASSESSMENT IN COPD
Abstract
A2291 - Identification of Subjects with High Risk of Disease Progression Using Data from the COPDGene Cohort
Author Block: R. Palmér1, E. Buehler2, U. Eriksson1, A. Mackay3, F. X. Blé3, M. Strand4, J. E. Hokanson5, B. J. Make4, S. I. Rennard6; 1Clinical Pharmacology and Quantitative Pharmacology, Clinical Pharmacology & Safety Sciences, R&D, AstraZeneca, Mölndal, Sweden, 2Biological and Knowledge Analytics, Data Science & Artificial Intelligence, R&D, AstraZeneca, Gaithersburg, MD, United States, 3Research and Early Development, Respiratory and Immunology (R&I), BioPharmaceuticals R&D, AstraZeneca, Mölndal, Sweden, 4National Jewish Health, Denver, CO, United States, 5Dept of Epidemiology, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, CO, United States, 6University of Nebraska Medical Center, Omha, NE, United States.
RATIONALE
Identification of COPD patients at high risk of disease progression is an important issue when designing interventional randomized COPD clinical trials. Enrollment of rapid progressors may improve chances of identifying disease modifying treatment effects and also ensures trials are conducted in patients with an unmet medical need. We hypothesized that COPDGene® data from baseline and 5-year visits can be used to characterize different levels of COPD disease progression and to identify predictors of rapid progression.
METHODS
Data were analyzed using a multi-response latent class mixed model. Four response variables (pre- and post-bronchodilator FEV1 and PEF) were used to estimate an underlying (latent) lung function metric and its progression between baseline and the 5-year visit. Two classes of progression (normal and rapid) were defined with a target of assigning 10-20% of subjects to the rapid progression class. Predictors of class membership were identified using a combination of statistical and machine learning techniques. Possible predictors included all available baseline patient characteristics (except genomic data) as well as hematology lab variables collected at the 5-year visit (not measured at baseline). In total, 3449 subjects met the data availability criteria to be included in the analysis.
RESULTS
In subjects with a COPD diagnosis at baseline (n=1187), selected predictors associated with an increased probability of rapid progression included: high gas trapping, emphysema, phlegm frequency, tiotropium use, beta-agonist use, lymphocytes (at 5-year visit) and heart rate; low weight and MCHC (at 5-year visit); and current unemployment. In total, 24% of the COPD subjects were classified as rapid progressors. On average, these subjects exhibited a 2-fold more rapid 5-year decline in FEV1 and PEF (16-24% decline from baseline) compared to the normal progression class (8-12%). In subjects without a COPD diagnosis (n=2262), predictors associated to rapid progression included: high gas trapping, monocytes (at 5-year visit), Pi10, MCV (at 5-year visit), blood pressure, eosinophils (at 5-year visit), and heart rate; low MCHC (at 5-year visit); and current smoking and a history of cough. On average, rapid progressors (10% of subjects) showed a 2-fold more rapid 5-year decline in FEV1 and PEF (16-18% compared to 6-9%).
CONCLUSIONS
CT parameters, hematology labs, and vital signs were identified as predictors of rapid disease progression in both COPD and non-COPD subjects and may potentially be used as inclusion criteria in COPD clinical trials investigating disease-modifying treatment effects. The importance of the predictors should be validated using other datasets.
Author Block: R. Palmér1, E. Buehler2, U. Eriksson1, A. Mackay3, F. X. Blé3, M. Strand4, J. E. Hokanson5, B. J. Make4, S. I. Rennard6; 1Clinical Pharmacology and Quantitative Pharmacology, Clinical Pharmacology & Safety Sciences, R&D, AstraZeneca, Mölndal, Sweden, 2Biological and Knowledge Analytics, Data Science & Artificial Intelligence, R&D, AstraZeneca, Gaithersburg, MD, United States, 3Research and Early Development, Respiratory and Immunology (R&I), BioPharmaceuticals R&D, AstraZeneca, Mölndal, Sweden, 4National Jewish Health, Denver, CO, United States, 5Dept of Epidemiology, Colorado School of Public Health, University of Colorado Anschutz Medical Campus, Aurora, CO, United States, 6University of Nebraska Medical Center, Omha, NE, United States.
RATIONALE
Identification of COPD patients at high risk of disease progression is an important issue when designing interventional randomized COPD clinical trials. Enrollment of rapid progressors may improve chances of identifying disease modifying treatment effects and also ensures trials are conducted in patients with an unmet medical need. We hypothesized that COPDGene® data from baseline and 5-year visits can be used to characterize different levels of COPD disease progression and to identify predictors of rapid progression.
METHODS
Data were analyzed using a multi-response latent class mixed model. Four response variables (pre- and post-bronchodilator FEV1 and PEF) were used to estimate an underlying (latent) lung function metric and its progression between baseline and the 5-year visit. Two classes of progression (normal and rapid) were defined with a target of assigning 10-20% of subjects to the rapid progression class. Predictors of class membership were identified using a combination of statistical and machine learning techniques. Possible predictors included all available baseline patient characteristics (except genomic data) as well as hematology lab variables collected at the 5-year visit (not measured at baseline). In total, 3449 subjects met the data availability criteria to be included in the analysis.
RESULTS
In subjects with a COPD diagnosis at baseline (n=1187), selected predictors associated with an increased probability of rapid progression included: high gas trapping, emphysema, phlegm frequency, tiotropium use, beta-agonist use, lymphocytes (at 5-year visit) and heart rate; low weight and MCHC (at 5-year visit); and current unemployment. In total, 24% of the COPD subjects were classified as rapid progressors. On average, these subjects exhibited a 2-fold more rapid 5-year decline in FEV1 and PEF (16-24% decline from baseline) compared to the normal progression class (8-12%). In subjects without a COPD diagnosis (n=2262), predictors associated to rapid progression included: high gas trapping, monocytes (at 5-year visit), Pi10, MCV (at 5-year visit), blood pressure, eosinophils (at 5-year visit), and heart rate; low MCHC (at 5-year visit); and current smoking and a history of cough. On average, rapid progressors (10% of subjects) showed a 2-fold more rapid 5-year decline in FEV1 and PEF (16-18% compared to 6-9%).
CONCLUSIONS
CT parameters, hematology labs, and vital signs were identified as predictors of rapid disease progression in both COPD and non-COPD subjects and may potentially be used as inclusion criteria in COPD clinical trials investigating disease-modifying treatment effects. The importance of the predictors should be validated using other datasets.
