Probing the intravascular and interstitial compartments of remodeled myocardium in heart failure patients with preserved and reduced ejection fraction: a CMR study

Abstract

Background: Recent autopsy studies found microvascular rarefaction in remodeled myocardium of patients who died of heart failure with preserved ejection-fraction (HFpEF). This condition has not been investigated so far by non-invasive methods in patients with HFpEF. The aim was to quantify the intravascular volume (IVV) compartment by CMR in HFpEF patients.

Methods: In two separate CMR examinations, HFpEF patients (n = 6; 12 examinations) and post-myocardial infarction patients (post-MI; n = 6; 12 examinations) were studied with T₁-mapping (MOLLI-sequence) before and after IV bolus of 0.03 mmol/Kg of the intravascular contrast-medium (CM) Gadofosveset and 0.2 mmol/Kg of the extravascular CM Gadobutrol yielding IVV and extracellular volume (ECV), respectively. Healthy controls (n = 10 with Gadofosveset only, n = 10 with Gadobutrol only) were also studied with the same protocol. IVV and ECV were measured in the basal septum (without ischemic scar in post-MI patients). In post-MI patients, ECV and IVV were also measured in the ischemic scar. Left ventricular (LV) volumes, mass, and ejection-fraction were measured by standard protocol. LV global longitudinal strain (GLS) was calculated by feature tracking on long-axis cine acquisitions.

Results: LV mass to end-diastolic volume ratio and GLS in HFpEF were higher and lower, respectively, than in healthy controls and post-MI patients, whereas the post-MI patients showed lower LV ejection-fraction. Compared to healthy myocardium of controls, IVV in scar was reduced (0.135 ± 0.018 vs 0.109 ± 0.008, respectively, p = 0.005), while ECV was increased (0.244 ± 0.037 vs 0.698 ± 0.106, respectively, p < 0.001). However, IVV did not differ among HFpEF, post-MI, and healthy controls (0.155 ± 0.033, 0.146 ± 0.038, and 0.135 ± 0.018, respectively, p = 0.413), whereas ECV was higher in HFpEF than in post-MI and healthy controls (0.304 ± 0.159, 0.270 ± 0.017, and 0.244 ± 0.037, respectively, p = 0.003).

Conclusions: The T₁-mapping technique combined with an intravascular CM shows potential to measure IVV. In infarct scar with substantially increased ECV, IVV was significantly reduced. Unlike in infarct scar, in remodeled myocardium of HFpEF patients, increased ECV was not accompanied by a reduction of IVV.

Fig. 1

Native and post-contrast T₁-maps in healthy volunteers, HFpEF, and post-MI patients. First and second rows, basal short-axis native (a, d), post-gadofosveset (b), post-gadobutrol (e) T₁-maps and LGE image (c, f) in two healthy volunteers. The region-of-interest (ROI) was traced within the basal interventricular septum in all volunteers and patients. Third and fourth rows, in HFpEF and post-MI patients, respectively, show basal short-axis native (g, l), post-gadofosveset (h, m), post-gadobutrol (I, n) T₁-maps, and LGE images (k, o). In the HFpEF patient (third row), non-ischemic scars (red arrow in K) was included in the septal ROI for the calculation of IVV and ECV. In post-MI patients (bottom row), a second ROI was traced within the ischemic scar taking as reference the LGE image (o, with red arrows indicating the scar)

Fig. 2

Top Panel: Bars and 95% confidence intervals of intravascular (IVV; a) and extracellular (ECV; b) volumes in healthy controls (green) and ischemic scar (red). IVV was reduced and ECV was strongly augmented in the ischemic scar as compared to normal myocardium. Bottom Panel: Bars and 95% confidence intervals of IVV (c) and ECV (d) in healthy (green), post-MI patients (remodeled myocardium not including the scar, light blue) and HFpEF patients (blue). Bonferroni’s post-hoc analysis, † P < 0.05