Predicting COPD Exacerbations: Investigators Identify New Biomarkers Using CT


The analysis uses patient data from SPIROMICS, followed by an external validation with the results from COPDGene.

Predicting COPD Exacerbations: Investigators Identify New Biomarkers with CT

Muhammad Chaudhary, BS

Computed Tomography (CT) scanning has become a cornerstone of the characterization of lung disease, and in new research, it's demonstrated usefulness as a tool to identify patients with chronic obstructive pulmonary disease (COPD) who are at high risk of experiencing severe exacerbations. 1

With CT-based prediction models, investigators were able to identify new biomarkers that have the potential to uncover underlying disease mechanisms that are responsible for exacerbations.

Investigators led by Muhammad Chaudhary, BS, of the Roy J Carver Department of Biomedical Engineering, University of Iowa, aimed to develop and validate quantitative CT-based models for predicting severe COPD exacerbations.

Novel CT-Based Prediction Model for COPD

The analysis included 1956 patients with a mean age of 63.1 years, and about half the population was men (52%). According to the findings, history of at least one severe exacerbation applied to 434 (22%) patients.

During the 3-year follow up, the receiver operating characteristic curve (AUC) was significantly higher with CT biomarkers (0.854) compared with exacerbation history (0.823) and the BMI, airflow obstruction, dyspnoea, exercise capacity (BODE) index (0.812).

When using the CT-based models to assess the proportion of patients with ≥1 acute episode in each of the 3 years, or consistent exacerbations, investigators reported an even higher AUC of 0.931.

Additional evidence provided by an external validation cohort of 6965 patients showed the AUC for at least one severe exacerbation was 0.768.

Identifying New Biomarkers

This analysis was derived from information from a cohort in the Subpopulations and Intermediate Outcome Measures in COPD Study (SPIROMICS), which was a multicenter study conducted at 12 clinical sites in the US and included individuals between 40-80 years.

Participants were evaluated based on 4 categories:

  • Those who never smoked
  • Those who smoked but had normal spirometry
  • Those who smoked and had mild-moderate COPD
  • Those who smoked and had severe COPD.

The current investigation used the 3-year follow up data to develop the logistic regression classifiers for predicting severe exacerbations including age, sex, race, BMI, pulmonary function, exacerbation history, smoking status, respiratory quality of life, and CT-based measures of density gradient texture and airway structure.

For the external validation model, investigators used a subset from the Genetic Epidemiology of COPD (COPDGene) cohort.

In addition to the AUC, the BODE index and exacerbation history were compared for discriminative model performance, then calibrated the model using calibration plots and Brier scores.

"We analysed CT biomarkers of lung tissue texture and airway structure in comparison with well known predictors, including exacerbation history and BODE index," investigators wrote. "We found Pi10 and CTDG texture were predictive of severe and consistent exacerbations."


  1. Chaudhary, Muhammad F A et al. Predicting severe chronic obstructive pulmonary disease exacerbations using quantitative CT: a retrospective model development and external validation study. The Lancet Digital Health, 2023; 5 (2): e83 - e92. doi:
Related Videos
Kelley Branch, MD, MSc | Credit: University of Washington Medicine
Sejal Shah, MD | Credit: Brigham and Women's
Video 2 - "Differentiating Medication Non-Adherence From Underlying Comorbidities"
Video 1 - "Defining Resistant Diabetes"
Stephanie Nahas, MD, MSEd | Credit: Jefferson Health
Kelley Branch, MD, MS | Credit: University of Washington Medicine
© 2024 MJH Life Sciences

All rights reserved.