P150 - EFFECT OF PEAK FLOW TRIGGERED ADAPTIVE SERVO-VENTILATION (ASVPF) ON MORTALITY IN HEART FAILURE WITH REDUCED EJECTION FRACTION PATIENTS WITH PROLONGED CHEYNE-STOKES RESPIRATORY CYCLE LENGTH.

Thursday, October 23, 2025

1:30pm - 2:00pm ET

Location: Theatre 1

Presenting Author(s)

SI

Sayaki Ishiwata (he/him/his)

post-doctoral research fellow
Sleep Research Laboratories of the University Health Network, Toronto Rehabilitation Institute (KITE) and Toronto General Hospital and University of Toronto
Toronto, Ontario, Canada

Background: In ADVENT-HF trial, treating obstructive sleep apnea (OSA) and central sleep apnea with Cheyne-Stokes respiration (CSA-CSR) in patients with heart failure and reduced ejection fraction (HFrEF) using ASVPF did not affect overall mortality, but there was a trend towards lower mortality in those with CSA-CSR (HR, 0.74, 95%CI 0.44-1.23, p=0.25). Patients with CSA-CSR had more severe HFrEF, suggesting that ASVPF might be more effective in sicker patients. Because CSA-CSR cycle length (CL) is inversely proportional to cardiac output, we hypothesized that: 1) in patients with CSA-CSR in the control arm, mortality would be greater in those with longer than shorter CL; and 2) ASVPF would reduce mortality in those with longer but not shorter CLs.

METHODS AND RESULTS: In 305 patients with HFrEF and predominantly CSA-CSR or predominantly OSA with a central apnea-hypopnea index of ≥10 (OSA-CSA) we measured and averaged CL of 20 CSA-CSR cycles from the beginning of one apnea to the beginning of the next. We divided them into short and long CL groups based on median CL of the CSA-CSR group (54.9 sec). We then compared all-cause mortality between the 2 control groups and effect of ASVPF on all-cause mortality in the short and long CL groups. Among controls, 62 patients with longer CL (67.8±10.9 sec) had higher mortality than 102 patients with shorter CL (43.0±6.7 sec) (HR, 3.91, 95%CI 2.14-7.17, p< 0.0001, Figure 1). In patients with longer CL, mortality was significantly lower in those randomized to ASVPF than to control (HR, 0.56, 95%CI 0.32-0.99, p=0.0438, Figure 2), but ASVPF had no significant effect in the shorter CL group (HR, 1.08, 95%CI 0.52-2.24, p=0.8401).

Conclusion: In HFrEF, all-cause mortality was higher in those with untreated CSA-CSR and OSA-CSA with longer than shorter CL. Most importantly, ASVPF reduced all-cause mortality in the long CL group. This is the first demonstration that treating a particular phenotype of sleep apnea in HFrEF can reduce mortality.