Assessment of Solid Pulmonary Nodules or Masses Using Zero Echo Time MR Lung Imaging: A Prospective Head-to-Head Comparison With CT

Objective The aim of this study is to determine the potential of zero echo time (ZTE) MR lung imaging in the assessment of solid pulmonary nodules or masses and diagnostic consistency to CT in terms of morphologic characterization. Methods Our Institutional Review Board approved this prospective study. Seventy-one patients with solid pulmonary nodules or masses larger than 1 cm in diameter confirmed by chest CT were enrolled and underwent further lung ZTE-MRI scans within 7 days. ZTE-MRI and CT images were compared in terms of image quality and imaging features. Unidimensional diameter and three-dimensional volume measurements on both modalities were manually measured and compared using the Wilcoxon signed-rank test, intraclass correlation coefficient (ICC), Pearson’s correlation analysis, and Bland–Altman analysis. Multivariable logistic regression analysis was used to identify the factors associated with significant inter-modality variation of volume. Results Fifty-four of 71 (76.1%) patients were diagnosed with lung cancer. Subjective image quality was superior in CT compared with ZTE-MRI (p < 0.001). Inter-modality agreement for the imaging features was moderate for emphysema (kappa = 0.50), substantial for fibrosis (kappa = 0.76), and almost perfect (kappa = 0.88-1.00) for the remaining features. The size measurements including diameter and volume between ZTE-MRI and CT showed no significant difference (p = 0.36 for diameter and 0.60 for volume) and revealed perfect inter-observer (ICC = 0.975–0.980) and inter-modality (ICC = 0.942–0.992) agreements. Multivariable analysis showed that non-smooth margin [odds ratio (OR) = 6.008, p = 0.015] was an independent predictor for the significant inter-modality variation of volume. Conclusion ZTE lung imaging is feasible as a part of chest MRI in the assessment and surveillance for solid pulmonary nodules or masses larger than 1 cm, presenting perfect agreement with CT in terms of morphologic characterization.