P6326Which results of the cardiopulmonary exercise test deserve greatest attention to establish the prognosis in heart failure?

Abstract

Abstract Background and aim Limitation of exercise tolerance is one of the cardinal manifestations of heart failure (HF). Cardiopulmonary exercise testing (CPET) provides a thorough assessment of exercise integrative physiology involving the pulmonary, muscular, and oxidative cellular systems. We aimed to identify which data collected during a CPET shows the best prognostic performance with respect to predicting mortality or the need for heart transplantation (HT). Methods Single-centre retrospective cohort study of consecutive HF patients performing a CPET for functional and prognostic HF evaluation from October 1996 till May 2018. Left ventricular ejection fraction was not an exclusion criterion. A Cox model was fit with time to death or heart transplantation (whichever recorded first within 5 years) as the dependent variable and CPET parameters as the independent variables. Both unadjusted and adjusted covariate Cox regressions were performed. ROC curve analysis was used to determine whether the significant variables, as a model, could reliably predict the study endpoint. Results The study population consisted of 513 patients, median age 58 (IQ 16) years, and 74.9% male. The majority had reduced ejection fraction (75.4%), and the most common HF aetiology was ischemic heart disease (55.8%). During the 5-years follow up, 126 patients died and 60 underwent heart transplantation. In unadjusted Cox regression, nearly all CPET variables were significantly associated with the study endpoint. After covariate adjustment, with prior exclusion of redundant variables, three measures remained associated with the study endpoint: peak VO2 consumption (hazard ratio [HR] 0.85; 95% confidence interval [CI], 0.81–0.90); VE/VCO2 slope (HR 1.02; 95% CI, 1.00–1.02); presence of oscillatory ventilatory pattern (HR 3.73; 95% CI, 2.43–5.72). As a model, these 3 variables showed a strong discriminatory ability (c-statistic 0.87; 95% CI, 0.83–0.90) (see figure) for the study endpoint. Figure 1 Conclusion When using the CPET for prognostic stratification of HF patients, the presence of an oscillatory ventilatory pattern, the peak VO2 and the VE/VCO2 slope are the most important tools on which clinicians should focus.