Assessment of CF lung disease using motion corrected PROPELLER MRI: a comparison with CT

Objectives To date, PROPELLER MRI, a breathing-motioninsensitive technique, has not been assessed for cystic fibrosis (CF) lung disease. We compared this technique to CT for assessing CF lung disease in children and adults. Methods Thirty-eight stable CF patients (median 21 years, range 6-51 years, 22 female) underwent MRI and CT on the same day. Study protocol included respiratory-triggered PROPELLER MRI and volumetric CT end-inspiratory and - expiratory acquisitions. Two observers scored the images using the CF-MRI and CF-CTsystems. Scores were compared with intra-class correlation coefficient (ICC) and Bland- Altman plots. The sensitivity and specificity of MRI versus CT were calculated. Results MRI sensitivity for detecting severe CF bronchiectasis was 0.33 (CI 0.09-0.57), while specificity was 100 % (CI 0.88-1). ICCs for bronchiectasis and trapped air were as follows: MRI-bronchiectasis (0.79); CT-bronchiectasis (0.85); MRI-trapped air (0.51); CT-trapped air (0.87). Bland-Altman plots showed an MRI tendency to overestimate the severity of bronchiectasis in mild CF disease and underestimate bronchiectasis in severe disease. Conclusions Motion correction in PROPELLER MRI does not improve assessment of CF lung disease compared to CT. However, the good inter- and intra-observer agreement and the high specificity suggest that MRI might play a role in the short-term follow-up of CF lung disease (i.e. pulmonary exacerbations).