Prevalence and Prognostic Significance of Microvascular Dysfunction in Heart Failure With Preserved Ejection Fraction.

Background: The pathophysiological and clinical significance of microvascular dysfunction (MVD) in patients with heart failure with preserved ejection fraction (HFpEF) remains uncertain.

Objectives: The aim of this study was to use cardiovascular magnetic resonance to: 1) quantify coronary microvascular function; 2) examine the relationship between perfusion and fibrosis; and 3) evaluate the impact of MVD and fibrosis on long-term clinical outcomes.

Methods: In a prospective, observational study, patients with HFpEF and control subjects underwent multiparametric cardiovascular magnetic resonance (comprising assessment of left ventricular volumetry, perfusion, and fibrosis [focal by late gadolinium enhancement and diffuse by extracellular volume]). The primary endpoint was the composite of death or hospitalization with heart failure.

Results: One hundred and one patients with HFpEF (mean age 73 ± 9 years, mean ejection fraction 56% ± 5%) and 43 control subjects (mean age 73 ± 5 years, mean ejection fraction 58% ± 5%) were studied. Myocardial perfusion reserve (MPR) was lower in patients with HFpEF versus control subjects (1.74 ± 0.76 vs 2.22 ± 0.76; P = 0.001). MVD (defined as MPR <2.0) was present in 70% of patients with HFpEF (vs 48% of control subjects; P = 0.014). There was no significant linear correlation between MPR and diffuse fibrosis (r = -0.10; P = 0.473) and no difference in MPR between those with and without focal fibrosis (mean difference -0.03; 95% CI: -0.37 to 0.30). In the HFpEF group, during median follow-up of 3.1 years, there were 45 composite events. MPR was independently predictive of clinical outcome following adjustment for clinical, blood, and imaging parameters (1 SD increase: HR: 0.673 [95% CI: 0.463 to 0.978; P = 0.038]; HR: 0.694 [95% CI: 0.491 to 0.982; P = 0.039]; and HR: 0.690 [95% CI: 0.489 to 0.973; P = 0.034], respectively).

Conclusions: MVD is highly prevalent among patients with HFpEF and is an independent predictor of prognosis. The lack of correlation between MVD and fibrosis may challenge the assertion of a direct causal link between these entities. (Developing Imaging and Plasma Biomarkers in Describing Heart Failure With Preserved Ejection Fraction [DIAMONDHFpEF]; NCT03050593).