Quantitative Phase-resolved Functional Lung MRI Prediction of Disease Progression in Connective Tissue...

Background Connective tissue disease-associated interstitial lung disease (CTD-ILD) is a spatially and temporally heterogeneous disease. Determining whether CTD-ILD is in a progressive stage is crucial for guiding clinical management. Purpose To determine whether CTD-ILD is associated with ventilation and perfusion parameters quantified with phase-resolved functional lung (PREFUL) MRI and whether these functional parameters can help predict CTD-ILD progression. Materials and Methods In this prospective study performed from May 2024 to April 2025, healthy participants without pulmonary disease and participants with CTD-ILD underwent baseline PREFUL MRI, chest CT, and pulmonary function tests. Participants with CTD-ILD were followed up for 1 year. Disease progression was defined as meeting at least two of the following criteria: an absolute decline in forced vital capacity of more than 5% predicted or an absolute decline in diffusing capacity of lung for carbon monoxide of more than 10% predicted, CT evidence of disease progression, or worsening respiratory symptoms. Least absolute shrinkage and selection operator regression were applied to identify variables associated with disease progression, and predictive performance was assessed using the area under the receiver operating characteristic curve (AUC). Results A total of 172 participants (mean age, 59 years ± 11.8 [SD]; 143 women) were included. Compared with controls, participants with CTD-ILD demonstrated reduced dynamic ventilation (mean, 0.97 arbitrary units [au] ± 0.02 vs 0.95 au ± 0.05, respectively; P < .001) and lower perfusion (mean, 50.3 mL/min per 100 mL ± 16.2 vs 22.0 mL/min per 100 mL ± 9.2; P < .001). At baseline, compared with stable CTD-ILD, progressive CTD-ILD was associated with a higher perfusion defect (mean, 14.7% ± 6.8 vs 26.5% ± 9.1; P < .001) and lower healthy ventilation and perfusion matches (mean, 75.6% ± 8.9 vs 62.1% ± 10.8; P = .001). The multiparametric PREFUL model achieved the highest predictive performance (AUC, 0.87; 95% CI: 0.78, 0.95) compared with other logistic models (P = .004, P < .001). Conclusion In participants with CTD-ILD, ventilation and perfusion parameters derived from PREFUL MRI characterized distinct pulmonary perfusion phenotypes in CTD-ILD and were associated with disease progression over 1 year. © RSNA, 2026 Supplemental material is available for this article. See also the editorial by Rahsepar and Abtin in this issue.