Gas Exchange MRI to Visualize Regional Therapy Response in Interstitial Lung Disease

There remains an urgent need to better phenotype interstitial lung diseases, predict trajectories, monitor progression, and measure treatment response. Because ILD pathology is spatially and temporarly heterogeneous, the problem demands a 3-dimensional (3D) lung assessment. To this end, CT imaging patterns are critical in IPF diagnosis, but such structural imaging is insensitive to changing functional status. Thus, we have developed hyperpolarized (HP) 129Xe MRI, whose unique properties enable rapid, non-invasive, 3D functional assessment of inhaled gas distribu-tion in the airspaces, as well as its uptake in the pulmonary interstitium (barrier tissues) and trans-fer to the pulmonary capillary red blood cells (RBCs). This way of probing regional gas exchange confers sensitivity to micron-scale interstitial barrier thickening and locally diminished RBC trans-fer. We have used 129Xe MRI to predict outcomes such as death or the need for transplant and have provided preliminary evidence showing that barrier uptake is a sensitive and early marker of therapy response. These advancements have led to demands for wider dissemination, which drives an urgent need to harmonize acquisition protocols and quantification methods.

This work will be conducted by 4 collaborating centers - Duke University, the University of Cincinnati, Cincinnati Children's Hospital Medical Center, and the University of Iowa. Cincinnati Children's will operate under its own IRB protocol and IND, but each site will share data and imaging techniques in order to establish best practices under an approved data sharing agreement. The specific aims of the work conducted at Cincinnati Children's:

Aim 1 - Establish Harmonized Quantitative Analysis of Gas Exchange MRI/MRS. We will deploy a reconstruction and analysis package enabling users of any of three major MRI vendor platforms to obtain robust, real-time quantitative analysis of images and spectra. We will demonstrate that healthy reference cohorts are equivalent to ±10% across 4 centers and use this to build the de-finitive multi-site healthy reference distributions.

Aim 2 - Deploy a Clinical Framework to Identify Active Fibrosis and Normal Aging. To position the technology for clinical deployment and interpretation, we will develop a physiologic model incor-porating ventilation, barrier and RBC metrics to explain the underlying factors responsible for a given patient's diffusing capacity (DLCO) and transfer coefficient (KCO). This framework will be deployed to readers who will differentiate ILD from normal aging.

Aim 3 - Maximize and Measure Repeatability of 129Xe MRI/MRS Metrics across MRI Platforms. We will improve upon published repeatability of ±15-20% by harmonizing MRI/MRS protocols by improved coaching, optimized dose delivery, and tailoring the inhaled dose volume to the individ-ual patient. Using these approaches, each center will demonstrate coefficients of repeatability of ±6% or better in patients with ILD.