Non-Invasive Diagnosis of Pulmonary Vascular Disease Using Inhaled 129Xe Magnetic Resonance Imaging

Overview

This study seeks to deploy several forms of 129Xe MRI contrast as well as emerging conventional proton MRI technqiues for imaging lung structure and perfusion. Specifically, the 129Xe MRI scans will provide 3D images of ventilation and gas exchange, and spectroscopic indices will be evaluated to test gas exchange dynamics with high temporal resolution. The conventional 1H MRI scans will include a free-breathing ultra-short echo time (UTE) scan that provides images similar to that of a CT scan. In addition, to characterize perfusion and vascular dimensions directly, patients will undergo a gadolinium-enhanced perfusion scan.

This study seeks to deploy several forms of 129Xe MRI contrast as well as emerging conventional proton MRI technqiues for imaging lung structure and perfusion. Specifically, the 129Xe MRI scans will provide 3D images of ventilation and gas exchange, and spectroscopic indices will be evaluated to test gas exchange dynamics with high temporal resolution. The conventional 1H MRI scans will include a free-breathing ultra-short echo time (UTE) scan that provides images similar to that of a CT scan. In addition, to characterize perfusion and vascular dimensions directly, patients will undergo a gadolinium-enhanced perfusion scan. The overall aims of the study are as follows: Aim 1. Perform 129Xe MRI scans in cohorts of patients with PVD, isolated left sided heart failure and isolated lung disease: 1.1 Deploy and optimize 129Xe gas exchange MR spectroscopy and imaging, 129Xe ventilation MRI, structural 1H UTE MRI, and breath-hold 1H perfusion MRI. 1.2 Conduct comprehensive 1H-129Xe MRI in 10 patients with pure pulmonary arterial hypertension (PAH) 1.3 Conduct comprehensive MRI in 10 patients with pure left heart failure 1.4 Conduct comprehensive MRI in 10 patients with pure lung disease but no pulmonary hypertension 1.5 Conduct comprehensive MRI in 5 patients with Chronic thromboembolic pulmonary hypertension (CTEPH) pre and post PTE surgery. Aim 2. Develop diagnostic criteria for optimizing the sensitivity and specificity of 129Xe MRI for the diagnosis of PVD: 2.1 Develop quantification/scoring methods based on PAH, left heart disease, and lung disease MRI 2.2 Develop reader training materials and train 3 expert readers 2.3 Begin prospective recruitment of 92 patients being evaluated for PAH for MRI scans. These patients may be asked to return for a second scan in the study, at a time no earlier than 48hr after their last xenon dose. Aim 3. Perform a larger, single-blind study testing the diagnostic accuracy of 129Xe MRI for diagnosis of PVD: 3.1 Trained readers will evaluate 129Xe MRI, while blinded to the subject's disease state and will determine the presence and severity of PVD 3.2 Diagnostic accuracy of 129Xe MRI will be compared to the gold standard of hemodynamic and clinical criteria of PAH 1. For this PFTs are not necessary, so they are ancillary 2. However, they are a standard way to characterize patients for publications, etc, so we would like to get them if we can. Getting them from the medical record is perfectly adequate for this pur-pose. i. If PFTs are not available from the medical record, they are not so critical as to pre-vent us from enrolling the subject. In publications we will simply acknowledge the realities of COVID for this missing data. This will be an single-blinded, open-label study enrolling volunteers and patients with pure PAH (10 subjects), pure left heart disease (10 subjects) and pure lung disease (10 subjects) \[Aim 1\] followed by a larger cohort of 92 subjects being evaluated for PAH \[Aim 2\]. The investigators plan to consent to 127 subjects. The sample size calculation is based on testing whether the proposed diagnostic test based on Xe MRI has accuracy comparable to imaging diagnostic procedures in clinical practice, such as mammography. To this end, the investigator proposes to test the hypotheses H0: AUC = 0.72 vs. H1: AUC \> 0.85, where area under the curve (AUC) is a summary measure of diagnostic accuracy obtained as the the area under the receiver operating characteristic (ROC) curve which displays the tradeoff between sensitivity and specificity for our proposed Xe MRI based diagnostic test. Here AUC = 0.72 represents a moderately accurate test while AUC = 0.85 represents accuracy comparable to clinical mammography. Based on a one sided, one sample test with variances approximated using the binormal model, the investigator estimates a minimum sample size of 46 normals and 46 with disease will be required to test the above hypotheses at a significance level of 5% with 80% power.