Invasive exercise hemodynamics predict survival in patients with clinical heart failure irrespective of left ventricular systolic function

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Introduction When aiming to stratify heart failure patients for mortality risk, exercise stress testing has proven to be relevant in specific populations of heart failure patients, but is not universally applied. Analysis of left ventricular systolic function (LVEF) remains the cornerstone for stratifying heart failure patients for prognosis and clinical treatment. The aim of the present study is to investigate the prognostic capabilities of invasive exercise testing in a real-world cohort of suspected heart failure patients in whom non-cardiac causes of dyspnea were excluded. Methods We retrospectively analyzed survival of patients who underwent right heart catheterization at rest and during exercise between 2007 to 2017 for dyspnea and expected heart failure. Pulmonary capillary wedge pressure (PCWP) at rest and the PCWP response to exercise, expressed as the ratio of PCWP at peak exercise to workload normalized to body weight [PCWL (mmHg/W/kg)], were determined. Mortality data was deducted from the official German death registry. Results A total of 682 patients [aged 64.1 ± 12 years, 269 (39.2%) women] were included. Based on LVEF, 515 patients (76.4%) had sustained LVEF ≥50%, 71 patients (10.5%) had heart failure with mildly-impaired LVEF (40-49%) and 88 patients (13.1%) had heart failure with reduced LVEF (≤40%). Absolute LVEF was not associated with all-cause mortality (HR 0.995; 95% CI 0.99-1.00; p=0.231). While increases in PCWP at rest an PCWL during exercise were both associated with death, the discriminative properties of pressure measurements derived during exercise were superior in ROC-analysis (AUC 0.637 for PCWP and AUC 0.728 for PCWL, p<0.001) Exercise hemodynamics and systolic function were next used to re-classify patients in four groups based on normal vs. impaired LVEF (≥/< 50%) and response to exercise (PCWL≥/<25.5 mmHg/W/kg for normal LVEF and PCWL≥/<34.7 mmHg/W/kg for impaired LVEF). In multivariate models adjusted for pertinent baseline parameters, patients with pathological response to exercise carried a 2.0-fold (with sustained LVEF, 95% CI 1.17-3.52, p=0.011) to 3.2-fold (with impaired LVEF, 95% CI 1.83-5.91, p<0.001) increased risk for death. Patients with impaired LVEF but featuring a normal response to exercise did not carry an increased risk for death (HR 1.117, 95% CI 0.453-2.757, p=0.810). Discussion: In patients with clinical heart failure, invasive exercise testing improves the prediction of mortality. Subjects with a normal response to exercise have a relatively low mortality irrespective of systolic function.