Evaluation of Dynamic Contrast‐Enhanced MRI Measures of Lung Congestion and Endothelial Permeability in Heart Failure: A Prospective Method Validation Study

BACKGROUND: Methods for accurate quantification of lung fluid in heart failure (HF) are needed. Dynamic contrast‐enhanced (DCE)‐MRI may be an appropriate modality. PURPOSE: DCE‐MRI evaluation of fraction of fluid volume in the interstitial lung space (v ( e )) and vascular permeability (K (trans)). STUDY TYPE: Prospective, single‐center method validation. POPULATION: Seventeen evaluable healthy volunteers (HVs), 12 participants with HF, and 3 with acute decompensated HF (ADHF). FIELD STRENGTH/SEQUENCE: T (1) mapping (spoiled gradient echo variable flip angle acquisition) followed by dynamic series (three‐dimensional spoiled gradient‐recalled echo acquisitions [constant echo time, repetition time, and flip angle at 1.5 T]). ASSESSMENT: Three whole‐chest scans were acquired: baseline (Session 1), 1‐week later (Session 2), following exercise (Session 3). Extended Tofts model quantified v ( e ) and K (trans) (voxel‐wise basis); total lung median measures were extracted and fitted via repeat measure analysis of variance (ANOVA) model. Patient tolerability of the scanning protocol was assessed. STATISTICAL TESTS: This was constructed as an experimental medicine study. Primary endpoints: K (trans) and v ( e ) at baseline (HV vs. HF), change in K (trans) and v ( e ) following exercise, and following lung congestion resolution (ADHF). K (trans) and v ( e ) were fitted separately using ANOVA. Secondary endpoint: repeatability, that is, within‐participant variability in v ( e ) and K (trans) between sessions (coefficient of variation estimated via mixed effects model). RESULTS: There was no significant difference in mean K (trans) between HF and HV (P ≤ 0.17): 0.2216 minutes(−1) and 0.2353 minutes(−1) (Session 1), 0.2044 minutes(−1) and 0.2567 minutes(−1) (Session 2), 0.1841 minutes(−1) and 0.2108 minutes(−1) (Session 3), respectively. v ( e ) was greater in the HF group (all scans, P ≤ 0.02). Results were repeatable between Sessions 1 and 2; mean values for HF and HV were 0.4946 and 0.3346 (Session 1), 0.4353 and 0.3205 (Session 2), respectively. There was minimal difference in K (trans) or v ( e ) between scans for participants with ADHF (small population precluded significance testing). Scanning was well tolerated. DATA CONCLUSION: While no differences were detected in K (trans), v ( e ) was greater in chronic HF patients vs. HV, augmented beyond plasma and intracellular volume. DCE‐MRI is a valuable diagnostic and physiologic tool to evaluate changes in fluid volume in the interstitial lung space associated with symptomatic HF. LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY STAGE: 2