Predictive Value of Ventilatory Efficiency Across Categories of Heart Failure

Abstract

Background: In cardiopulmonary stress testing (CPET), minute ventilation-carbon dioxide production relationship (VE/VCO2 slope) strongly predicts outcomes in heart failure (HF). A VE/VCO2 classification system evaluated in patients with reduced left ventricular ejection fraction (LVEF) requires further investigation in those with mid-range and preserved LVEF. Further, whether the current clinical practice of using the same VE/VCO2 threshold for defining normal and abnormal across all types of HF requires investigation. Methods: Single-center retrospective cohort study of 1347 patients with heart failure (60.5% male, age 58.0±14.6 years, LVEF 42% ± 17%) referred for CPET for clinical indications between 2010 and 2016. LVEF was obtained from echocardiogram (n=1309) or cardiac MRI (n=38). All-cause mortality within 2 years of CPET was determined using Research Patient Data Registry at Partners Healthcare, which is linked to National Death Index. HF admission data within in two years of CPET were obtained through chart review. Patients with HF were categorized based on LVEF into heart failure with reduced ejection fraction (HFrEF, LVEF <40%, n=598), heart failure with mid-range ejection fraction (HFmrEF, 40% ≤LVEF <50%, n=164) and heart failure with preserved ejection fraction (HFpEF, LVEF≥50%, n=585). VE/VCO2 slope was divided into four ventilatory categories (VC) - VC-I: VE/VCO2≤ 29, VC-II: 29< VE/VCO2< 36, VC-III: 36≤VE/VCO2<45, VC-IV: VE/VCO2 ≥45. Associations between VE/VCO2 slope category and composite outcome of 2-year all-cause mortality and HF admissions were examined. Receiver operating characteristic (ROC) analysis was performed to detect the optimal VE/VCO2 cutoff for outcome prediction in each HF group. Results: At two-year follow-up post CPET, there were 95 deaths, 134 HF admissions and 197 composite events (death or HF admission). Compared to patients in VC-I, patients with VE/VCO2 ≥ 36 are at increased risks to have a two-year composite event across all three HF cohorts after adjusting for age and gender. VC-II was also associated with increased likelihood of two-year composite outcome in HFrEF (hazard ratio [HR] 2.75, 95% confidence interval [CI]: 1.26 to 5.54, p =0.005) and HFpEF patients (HR 2.68, 95% CI: 1.12 to 6.42, p=0.03). There was a trend of increased likelihood of two-year composite outcome in patients with higher ventilatory categories in the entire cohort. ROC analysis indicated that the optimal VE/VCO2 was 36, 30 and 29 for HFrEF, HFmrEF and HFpEF patients respectively. Conclusions: This study validated the current clinical cut-off of VE/VCO2 of 36 across three HF cohorts; HF patients with VE/VCO2 ≥ 36 are at increased risk to have either death or HF hospitalization occur compared to patients with VE/VCO2 <29 regardless of their LVEF. Furthermore, in patients with HFpEF and HFrEF, VE/VCO2 between 29 and 36 was also found associated with increased risks of two-year all-cause mortality or HF hospitalization. Overall HF patients with higher VC categories are at higher risk to have HF hospitalization or death occur. The optimal cut-off for VE/VCO2 was 36 for patients with HFrEF but a lower cut-off should be considered for patients with HFmrEF and HFpEF.