Table 1.
Baseline characteristics of patients with HFpEF (n = 142).
Continuous variables are given as medians and inter-quartile ranges. Counts are given as numbers and percentages.
Table 2.
Cox proportional hazard models of non-invasive imaging measurements in patients with HFpEF (n = 142).
Hazard ratios (HR) refer to a 1-SD increase in continuous variables. HRs are adjusted (adj.) for all variables in the clinical confounder model i.e. diabetes, COPD, and NT-ProBNP.
Fig 1.
Kaplan-Meier estimates of primary endpoint (hospitalization for heart failure and/or death for cardiac reason) according to echocardiographic right ventricular function (normal and mildly reduced vs. significantly reduced.
3A; P<0.001, log-rank test), according to invasively measured systolic pulmonary arterial pressure (3B; cut-off = median; P = 0.001, log-rank test), and according to pulmonary artery wedge pressure (3C; cut-off = median; P = 0.006; log-rank test).
Table 3.
Cox proportional hazard models of invasive hemodynamic measurements in patients with HFpEF (n = 142).
Hazard ratios (HR) refer to a 1-SD increase in continuous variables. HRs are adjusted (adj.) for all variables in the clinical confounder model i.e. diabetes, COPD, and NT-ProBNP.
Fig 2.
Left ventricular myocardial biopsies from HFpEF patients.
Panel A shows a representative myocardial biopsy from a patient with a normal stroke volume (70ml) and a small extent of extracellular matrix (12%/mm2). In contrast, panel B shows a representative myocardial biopsy from a patient with a lower stroke volume (51 ml) and a significant amount of extracellular matrix (57%/mm2). Trichrome stain. Extracellular matrix stains blue-purple or green.
Fig 3.
Correlation between the extent of left ventricular extracellular matrix and invasively measured stroke volume (r = -0.53; p = 0.04).
Four patients with myocardial biopsy were not eligible for CMR study due to pacemakers.