Exercise oscillatory ventilation in patients with advanced heart failure with and without left ventricular assist device

Background: Exercise oscillatory ventilation (EOV), indicating pathological fluctuations on pulmonary arterial pressure, is associated with mortality in patients with heart failure (HF). Whether left ventricular assist device (LVAD)-induced ventricular unloading can reverse EOV and may prevent short-term rehospitalization has not been investigated.

Methods: We performed a retrospective single-center in- and outpatient analysis of patients with (n = 20, LVAD) and without (n = 27, HF) circulatory support and reduced ejection fraction (EF, 22.8 ± 7.9%). The association of cardiopulmonary exercise testing (CPET) variables and 3 months-rehospitalization (3MR) as a primary outcome was analyzed. Furthermore, CPET variables were compared regarding the presence of EOV (+/-).

Results: Lower VO_2peak (11.6 ± 4.9 ml/kg/min vs. 14.4 ± 4.3 ml/kg/min, p = 0.039), lower increase of PETCO₂ (CI = 0.049-1.127; p = 0.068), and higher VE/VCO₂ (43.8 ± 9.5 vs. 38.3 ± 10.6; p = 0.069) were associated with 3MR. Flattening of O₂ pulse (CI = 0.139-2.379; p = 0.487) had no impact on 3MR. EOV was present in 59.5% (n = 28/47) of patients, without a significant difference between LVAD and HF patients (p = 0.959). Patients with HF/EOV+ demonstrated significantly lower VO_2peak compared with HF/EOV- (p = 0.039). LVAD/EOV+ displayed significantly lower EF (p = 0.004) and fewer aortic valve opening than LVAD/EOV- (p = 0.027).

Conclusions: Lower VO_2peak , but not EOV, was associated with 3MR. EOV occurred at a similar rate in LVAD and HF patients, which may illustrate insufficient unloading during exercise in chronic LVAD therapy and may contribute to the limited exercise capacity following LVAD implantation. Simultaneous CPET and right heart catheterization studies are needed to elucidate whether EOV may serve as a non-invasive predictor of insufficient LV unloading necessitating LVAD reprograming.