Abstract Background Cardiopulmonary exercise testing (CPET) has an important role in mortality prediction in heart failure (HF) and patient selection for heart transplant. New indices as circulatory power (CP) and ventilatory power (VP) have been proposed as predictors of cardiac events. In addition, VP predicts mean pulmonary artery pressure (mPAP) in patients with pulmonary arterial hypertension. Purpose We aimed to analyse the prognostic value of classic and new CPET variables in patients with HF. Methods We retrospectively assessed consecutive patients with HF who underwent CPET in a single-centre between 2013 and 2017. New CPET variables were collected: CP was defined as the product of peak O2 uptake and peak systolic blood pressure (SBP), while VP was defined as peak SBP divided by the minute ventilation–CO2 production (VE/VCO2) slope. The primary endpoint was a composite of all-cause mortality, heart transplant, or HF hospitalization. Survival analysis was performed using Kaplan-Meier curves and multivariable Cox regression. Results Overall, 216 patients (mean age 55.4±10.9, 77.3% male) were included, 38.4% with ischemic HF, and mean left ventricle ejection fraction (LVEF) 30±14%. Most patients were evaluated through the modified Naughton (76.3%), the original Naughton (19.0%), and Bruce protocols (4.7%). Regarding classic CPET variables: mean pVO2 16.8±6.0 mL O2 kg–1 min–1, mean percent-predicted pVO2 62.6±23.9%, median VE/VCO2 slope 37.3 [32.6–44.5], exercise oscillatory ventilation (EOV) present in 13.9%, resting partial pressure of end-tidal carbon dioxide (PETCO2) ≥33 mmHg with an increase of 3–8 mmhg during exercise in 17.1%, and mean peak SBP 128.8±27.2 mmHg. Median circulatory power was 1927 [1404–2694] mmHg·min/mL/kg and mean ventilatory power 3.47±1.32 mmHg. After a median follow-up of 5 [4–6] years, the primary endpoint occurred in 66.2% of patients (rehospitalization, heart transplant, and all-cause death occurred in 57.0%, 25.9%, and 32.4%, respectively). In Cox regression multivariate analysis, the primary endpoint was predicted by pVO2 (HR 0.90, 95% CI: 0.87–0.93), percent-predicted pVO2 (HR 0.97, 95% CI: 0.96–0.98), VE/VCO2 slope (HR 1.04, 95% CI: 1.03–1.06), VP (HR 0.62, 95% CI: 0.52–0.73) but not CP (HR 0.99, 95% CI: 0.98–1.01). Kaplan-Meier curves according to the LVEF are depicted in Fig. 1A. ROC analysis (Fig. 1B) revealed that VP (AUC 0.768) has higher discriminative power for the primary endpoint, compared to pVO2 (AUC 0.741). One hundred and twenty-seven patients also underwent right heart catheterization: mean mPAP was 30.6±12.9 and it was not correlated with VP (r=−0.06, p=0.47). Conclusion CPET variables are good predictors of all-cause mortality, heart transplant, or HF hospitalization. Ventilatory power (but not circulatory power) is an additional useful variable in event prediction. On the other hand, VP is not correlated with mPAP in patients with HF. Funding Acknowledgement Type of funding sources: None.
Read full abstract