Field strength dependence, physiologic correlates, and prognostic significance of ventricular blood pool T2 mapping on cardiovascular magnetic resonance imaging.

Abstract

Background and purpose: Cardiovascular magnetic resonance (CMR) T2 mapping is sensitive to blood oxygenation. The purpose was to evaluate determinants of the right-to-left ventricular (RV/LV) blood pool (BP) T2 ratio, determine field strength-specific cut-points for prediction of major adverse cardiac events (MACE), and assess physiologic correlates.

Methods: Adults undergoing CMR (at 1.5T or 3T) for evaluation of cardiomyopathy between 2018 and 2020 were retrospectively evaluated along with healthy controls. RV and LV BP T2 values were measured on a mid-ventricular short-axis T2 map with calculation of the RV/LV BP T2 ratio. Reproducibility was assessed in a random subset of 100 patients. Associations with cardiopulmonary exercise testing (CPET) parameters and major adverse cardiac events (MACE) were evaluated using Spearman correlation, Cox regression, and receiver operating characteristic analysis.

Results: A total of 718 subjects were included: 678 with cardiomyopathy (mean age 50±16 years; 58% men) and 40 controls (mean age 45±15 years; 55% men). Over a median follow-up of 3.2 years (IQR 2.3-3.8), 53 patients experienced MACE. RV/LV BP T2 ratios were 18%-31% higher at 1.5T versus 3T (cardiomyopathy: 0.77±0.14 vs 0.59±0.13, p<0.001; controls: 0.79±0.11 vs 0.67±0.09, p<0.001) with excellent reproducibility (intra-observer ICC 0.96 and inter-observer ICC 0.95). Each 0.1-unit decrease in RV/LV BP T2 ratio was associated with 41%-42% increased hazard of MACE in multivariable analysis (HR 0.59, 95%CI 0.45, 0.78, p<0.001 at 1.5T and HR 0.58, 95%CI 0.40, 0.85, p=0.005 at 3T). Optimal cut-points for prediction of MACE were <0.70 at 1.5T (AUC 0.727) and <0.55 at 3T (AUC 0.689). Absolute RV BP T2 demonstrated similar prognostic performance. Among patients with CPET (n=150), lower RV/LV BP T2 ratios correlated significantly with reduced percent-predicted oxygen uptake (VO₂), impaired heart rate recovery, and worse ventilatory efficiency (VE/VCO₂) at anaerobic threshold.

Conclusions: The RV/LV BP T2 ratio reflects clinically meaningful cardiopulmonary physiology and independently predicts MACE, but is strongly influenced by MRI field strength, challenging assumptions of inherent field-independence and necessitating field strength-specific thresholds. Absolute RV BP T2 demonstrated comparable prognostic performance and may serve as a practical complementary metric. These findings establish RV BP T2 metrics as promising biomarkers for improved risk stratification in clinical CMR.