Characterization of Connective Tissue Disorders related Interstitial Lung Disease (CTD-ILD) with Computed Tomography (CT) and Magnetic Resonance Imaging (MRI)

Purpose/Objectives Interstitial Lung Disease (ILD) is a major complication of Systemic Sclerosis (SSc), with high morbidity and mortality. Computed tomography is the gold standard for ILD imaging assessment. However, recent Ultra-Short Echo-Time (UTE) Magnetic Resonance Imaging (MRI) sequences are promising for lung parenchyma evaluation. Moreover, different Dynamic Contrast Enhancement (DCE) MRI patterns seems able to identify prevalent inflammatory and fibrotic ILD, with possible implications in the management of the therapy. We aimed to verify the reliability of respiratory triggered UTE Spiral VIBE-MRI sequence in SSc-ILD assessment, compared to Computed Tomography (CT). Moreover, we tested the feasibility of DCE-MRI ILD analysis. Materials and methods 54 SSc patients performed chest CT and MRI in the same day. The MRI protocol included a UTE Spiral VIBE sequence before contrast administration and a VIBE sequence, scanned pre and post contrast administration (after 1, 3, 5 and 10 minutes). Two radiologists, independently and in consensus, analyzed CT and UTE verifying ILD presence/absence and performing a semiquantitative analysis (sQA) of ILD, Ground Glass Opacities (GGO), Reticulations and Honeycombing (HC) extents on CT and UTE. A CT software quantitative texture analysis (QA) of alterations extents was also performed. Then, in patients with ILD on CT, the two readers assessed the prevalent CE pattern as following: pattern 1) early enhancement and washout with discernible peak enhancement at 1 or 3 min; pattern 2) slight enhancement with no discernible peak enhancement at a specific time-point; and pattern 3) delayed persistent enhancement with discernible peak enhancement at 5 or 10 min. Pattern 1 was considered inflammatory, patterns 2 3 and 3 fibrotic. For ILD detections as well as for DCE MRI analysis, intra- and inter-reader agreements were computed with Cohen’s K coefficient. UTE sensitivity and specificity for ILD detection against CT were assessed. For extents analysis, intra-/inter-reader agreements and UTE performance against CT were computed by Lin’s concordance coefficient (CCC). Results 51 subjects were included for visual assessment, as 3 UTE were discarded for low quality. ILD was diagnosed in 39/51 CT. 16 out of the 51 included patients accepted to underwent VIBE pre and post contrast administration scans: 14 patients had ILD on CT and DCE-MRI were analyzed (no one discarded for low quality). UTE intra and inter-reader K in ILD diagnosis was 0.56 and 0.26, respectively. UTE showed 92.8% sensitivity and 75.0% specificity in ILD detection. 42 out of 51 QA segmentations were accepted. ILD, GGO and Reticulations extents were 14.8%, 7.7% and 7.1% on CT sQA and 13.0%, 11.2% and 1.6% on CT QA. HC was <1% and not further considered. UTE intra- and inter-reader CCC was 0.92 and 0.89 for ILD extent and 0.84 and 0.79 for GGO extent, respectively. UTE RET extent intra and inter-reader CCC was 0.22 and 0.18, respectively. ILD and GGO extents CCC were 0.95 and 0.93 against CT sQA and 0.89 and 0.88 against QA. RET extent CCC were 0.35 against CT sQA and 0.22 and against CT QA. DCE-MRI analysis has intra- and inter- reader K of 1 and 0.63 (disagreement in one patient), respectively. Only one patients demonstrated a prevalent inflammatory pattern. Conclusions In SSc patients, UTE Spiral VIBE-MRI sequence may suffer of low inter-reader agreement in ILD detection, but it is a reliable tool in assessing ILD and GGO extents in ascertained ILD. DCE-MRI ILD evaluation seems feasible could add information on inflammatory ILD activity.