AI-Powered Quantitative CT Analysis in Search of Predictive Biomarkers for Biologic Treatment in Severe Asthma 

Biologics therapy is the targeted injectable treatment for severe asthma designed to block specific immune pathways that drive airway inflammation. As one type of the biologicsBenralizumab effectively targets the IL-5 receptor to reduce eosinophilic inflammation. However, clinical responses in severe asthma patients remain highly heterogeneous. To bridge this gap, CHU Montpellier is collaborating with Thirona to utilize AI-driven quantitative CT analysis to evaluate the treatment effects of biologic therapy, identify patients most likely to benefit, and determine which metrics of airway structural changes best predict optimal responders. Ultimately, these findings will refine patient selection based on unique airway characteristics, paving the way for more precise and personalized therapeutic decisions. 

Project highlights

  • 59 severe eosinophilic asthma patients involved who were starting benralizumab therapy with 177 CT scans in total for the assessment  
  • Longitudinal monitoring of airway structural changes after the treatment at week 24 and week 48 
  • Objective measurement of structural changes in airway dimensions, mucus plugs (number and volume) and trapped air  

Contribution by Thirona

In this project, our AI-enabled quantitative imaging analysis was used to measure bronchial widening (Bout/A and Bin/A), wall thickening (Bwt/A and Bwa/Boa) using LungQ® BA, quantify mucus plug count and volumes using LungQ® MP and assess trapped air through low-attenuation region quantification using LungQ® MP.  

These quantitative insights into structural changes offer significantly higher precision and reproducibility than conventional visual assessments. By demonstrating greater sensitivity than standard lung function tests in detecting early-stage or subtle changes, these measurements can serve as robust, reliable metrics for both clinical trial endpoints and the development of personalized treatment strategies. 

Relevant Publications

No relevant publications found.

Do you want to stay up to date with our new developments?

Sign up to receive product and technology updates.