Abstract
BACKGROUND: Pulmonary exacerbations (PExs) are associated with lung function decline and poor quality of life in people with cystic fibrosis (CF). Standard clinical measurements are limited in predicting future PEx events; pulmonary MRI scan has been shown to be a sensitive measure of CF lung disease, but its ability to predict PExs is currently unknown. RESEARCH QUESTION: Are people with worse structural lung MRI scores and xenon MRI ventilation defects at greater risk of PExs, and are these measurements associated with future PExs? STUDY DESIGN AND METHODS: We retrospectively analyzed clinical, imaging, and spirometry data collected from 106 patients with CF 6 to 45 years of age including ultrashort echo time (UTE) MRI scan and (129)Xe ventilation MRI scan. PEx was defined as either recorded hospitalization for PEx or inpatient treatment with IV antibiotics. UTE MRI scan was visually scored using a modified Brody system, and xenon MRI ventilation defects were measured using ventilation defect percent (VDP). PEx details were obtained from electronic medical records. Statistical analysis included groupwise PEx occurrence and frequency comparisons, correlations between PExs and imaging variables, time-to-event analysis, and multivariable models. RESULTS: Patients who experienced PEx within 2 years after MRI scan had greater VDP, and VDP was higher in those with frequent exacerbations. People with abnormal VDP (> 3%) had a nearly 3 times (2.80; 95% CI, 1.6-4.56) greater PEx incidence rate. There were increased exacerbations among those with the most severe UTE MRI scores (fourth quartile) for consolidation, wall thickening, and bronchiectasis. Multivariable PEx prediction models that included imaging were stronger than those that included clinical information only, and VDP but not spirometry was a significant predictor in a model that controlled for prior exacerbations. INTERPRETATION: Our results show that structural and functional MRI measurements in people with CF were associated with future exacerbations, likely because they directly measure underlying structural and functional aspects of disease.