Publications
Take a look

James A. Eaden, Ho-Fung Chan, Paul JC. Hughes, Nicholas D. Weatherly, Matthew Austin, Laurie J. Smith, Jim Lithgow, Smitha Rajaram, Andrew J. Swift, Stephen A. Renshaw, Ronald A. Karwoski, Brian J. Bartholmai, Colm T. Leonard, Sarah Skeoch, Nazia Chaudhuri, Geoff JM. Parker, Stephen M. Bianchi, and Jim M. Wild.

Thorax 2019;74:A79. doi: 10.1136/thorax-2019-BTSabstracts2019.131.

Abstract

Introduction: Apparent diffusion coefficient (ADC) is a diffusion-weighted (DW) MRI measure of Brownian gas diffusion in the airspaces, where restrictions by tissue boundaries provide information about lung microstructure. The mean diffusive length scale (LmD), is another DW-MRI lung microstructure measurement calculated using a stretched exponential fit method. Computer-Aided Lung Informatics for Pathology Evaluation and Rating (CALIPER) quantifies various radiological parenchymal features based on histogram signature mapping techniques and is the most widely used quantitative CT image texture analysis software in interstitial lung disease (ILD).

Aim: To evaluate the ability of hyperpolarised 129-xenon (¹²⁹Xe) DW-MRI and high resolution CT (HRCT) to distinguish between ILD subtypes.

Methods: A prospective, multicentre study of patients with ILD including drug induced ILD (DI-ILD), hypersensitivity pneumonitis (HP), idiopathic pulmonary fibrosis (IPF) and connective tissue disease ILD (CTD-ILD). Hyperpolarised ¹²⁹Xe DW-MRI was performed on a 1.5 T scanner. The HRCT scan was performed within a year prior to the MRI scan. Quantitative CT analysis was performed using CALIPER. Semi-quantitative visual CT analysis was performed by two consultant chest radiologists using a scoring system (table 1).

Results: To date, 32 patients (7 DI-ILD, 6 HP, 14 IPF, 5 CTD-ILD) have undergone baseline ¹²⁹Xe DW-MRI and CT analysis. There was a significant difference in LmD between the HP and IPF groups (p=0.048) but this was not observed with ADC (p=0.16). Quantitative CT analysis demonstrated a significant difference between the ILD subtypes in both ground glass (GG) percent (p=0.007) and honeycombing percent (p=0.023), with the significant difference in GG% occurring between the HP and IPF groups (p=0.004). There was no significant difference between the ILD subtypes in reticulation percent (p=0.15). Semi-quantitative visual CT analysis showed a significant difference between the ILD subtypes in GG score (p=0.008), with the significant difference occurring between the HP and IPF groups (p=0.007). There was no significant difference between the ILD subtypes in the reticulation score (p=0.071) or the honeycombing score (p=0.16).

Conclusions: Our findings suggest significant differences in LmD, GG score and CALIPER GG% between IPF and HP. ¹²⁹Xe DW-MRI and quantitative CT could potentially have a role in differentiating between these ILD subtypes. .