B78-03 AI-Based Quantitative CT Can Characterize Structural Heterogeneity in Pi*ZZ Alpha-1 Antitrypsin Deficiency

American Journal of Respiratory and Critical Care Medicine, vol. 212

Abstract

Abstract Introduction Alpha-1 antitrypsin deficiency (AATD) is a genetic disorder predisposing to early-onset emphysema, typically with a basal-predominant pattern of lung damage. Artificial intelligence (AI)-based quantitative CT allows for objective, reproducible assessment of emphysema, bronchial and vascular metrics to evaluate wall thickening, bronchial dilatation or collapse, emphysema, and pulmonary blood distribution. These metrics can improve structural phenotyping of and may aid in early detection as well as monitoring of disease progression. We applied AI-based CT analysis in confirmed Pi*ZZ AATD patients to quantify structural abnormalities and inter-individual variability across disease stages. Methods Thirty Pi*ZZ AATD patients from the LUMC Alpha-1 Expertise Center were included in this pilot study. The LungQ platform (Thirona, The Netherlands) was used to analyze lung volume, number of bronchi and BA-pairs, tapering, and the percentage of low attenuation area below −950 HU (%LAA-950). LungQ-BA measured inner and outer bronchial diameters and adjacent arterial diameters to compute Bin/A and Bout/A ratios, as well as the wall-to-outer area ratio (Bwa/Boa) and bronchial tapering percentage (generations 1-6). Measurements were extracted across generations 1-6 and summarized as median(range). LungQ-MP quantified mucus plug number and volume. LungQ-AVX quantified arterial(A) and venous(V) volumes down to 0.2 mm diameter, expressed as percentage of total blood volume in size bins of (0.2-1)mm, (1-2)mm, and (>2)mm. Results are shown as median (range). Results Of 30 patients, 28 CTs were successfully analyzed: age 60 (39-79) years; 61.8% male; 50% former smokers; FEV1 %pred 43% (16-120); FEV1/FVC 0.47 (0.20-0.92); DLCOc %pred 56% (28-92.9); number of bronchi 433 (161-1089); BA-pairs 265 (114-432); %LAA-950 24% (1.2-49.4); arterial blood volume distribution 4.4% (3.2-7.6), 8.3% (6.3-11.6), 40.9% (25.5-45.9) for the (0-1)mm, (1-2)mm, (>2)mm size bins respectively; mucus plugs 0 (0-24) were present in 12 participants; bronchial tapering 21% (10.6-36) ; Bout/A 1.38 (0.88-2.26); Bwa/Boa 0.24 (0.18-0.56). Conclusion This pilot study demonstrated marked variability in parenchymal damage, airway wall thickening, bronchial dilatation, and mucus plugging, both across subjects and across disease stages, highlighting the heterogeneity of structural lung damage in AATD. The variation in structural metrics relative to spirometric and diffusing capacity impairment suggests that AI-based QCT can capture complementary aspects of airway and parenchymal remodeling, supporting its application in future longitudinal studies. This abstract is funded by: None

Authors 6

1. Tjeerd van der Veer corresponding Aachen

   Leiden University Medical Center

   Affiliation as printed

   Leiden University Medical Center (LUMC) , Leiden,

2. C Timmer Aachen

   Leiden University Medical Center

   Affiliation as printed

   Leiden University Medical Center (LUMC) , Leiden,

3. J van der Heijden
   Affiliation as printed

   Thirona , Nijmegen,

4. J.-P. Charbonnier
   Affiliation as printed

   Thirona , Nijmegen,

5. Emily F.A. van ’t Wout Aachen

   Leiden University Medical Center

   Affiliation as printed

   LUMC , Leiden,

6. H A W M Tiddens

   Erasmus MC - Sophia Children’s Hospital

   Affiliation as printed

   Sophia Childrens Hosp , Rotterdam,