Hypoxaemia in patients with heart failure and preserved ejection fraction

Aims: It is widely held that heart failure (HF) does not cause exertional hypoxaemia, based upon studies in HF with reduced ejection fraction, but this may not apply to patients with HF and preserved ejection fraction (HFpEF). Here, we characterize the prevalence, pathophysiology, and clinical implications of exertional arterial hypoxaemia in HFpEF. Methods and results: Patients with HFpEF (n = 539) and no coexisting lung disease underwent invasive cardiopulmonary exercise testing with simultaneous blood and expired gas analysis. Exertional hypoxaemia (oxyhaemoglobin saturation <94%) was observed in 136 patients (25%). As compared to those without hypoxaemia (n = 403), patients with hypoxaemia were older and more obese. Patients with HFpEF and hypoxaemia had higher cardiac filling pressures, higher pulmonary vascular pressures, greater alveolar–arterial oxygen difference, increased dead space fraction, and greater physiologic shunt compared to those without hypoxaemia. These differences were replicated in a sensitivity analysis where patients with spirometric abnormalities were excluded. Regression analyses revealed that increases in pulmonary arterial and pulmonary capillary pressures were related to lower arterial oxygen tension (PaO₂), especially during exercise. Body mass index (BMI) was not correlated with the arterial PaO₂, and hypoxaemia was associated with increased risk for death over 2.8 (interquartile range 0.7–5.5) years of follow-up, even after adjusting for age, sex, and BMI (hazard ratio 2.00, 95% confidence interval 1.01–3.96; p = 0.046). Conclusion: Between 10% and 25% of patients with HFpEF display arterial desaturation during exercise that is not ascribable to lung disease. Exertional hypoxaemia is associated with more severe haemodynamic abnormalities and increased mortality. Further study is required to better understand the mechanisms and treatment of gas exchange abnormalities in HFpEF.