FACE collected real life clinical data and long-term follow-up (FU) in a CHF population with SDB eligible for ASV. We previously identified 6 clinically meaningful clusters using Latent class analysis. We investigate herein whether these clusters may have different morbi-mortality outcomes.

Reduced (HFrEF), middle range (HFmrEF) and preserved (HFpEF) CHF classes were included. Eligible SDB were central sleep apnea (CSA), coexistent obstructive central SA not controlled under CPAP and CPAP emergent CSA. ASV was offered to all. Controls were patients who refused ASV at inclusion. Morbi-mortality events were collected for 2 years. Primary endpoint in the time-to event analysis was the first event of death, unplanned hospitalization for worsening HF or cardiac transplant.

503 patients were included in the ITT analysis at 2-year FU. ASV was protective for primary outcome (P=0.01, Fig. 1A). Cluster behavior regarding primary outcome was different, the worst prognosis was for HFrEF with CSA (Fig. 1B). ASV did not benefit in the HFrEF with CSA or OSA clusters (Fig. 1 C) but was protective for 2 clusters: obese HFpEF with CSA and severe hypoxemic patients with OSA (Fig. 1D).

These results align with recent studies that showed no impact on morbi-mortality of ASV in HFrEF but a protective effect in two HFpEF clusters.