Heart Failure with Preserved and Reduced Ejection Fraction in Hemodialysis Patients: Prevalence, Disease Prediction and Prognosis

• Main Authors: Marlies Antlanger, Stefan Aschauer, Chantal Kopecky, Manfred Hecking, Johannes J. Kovarik, Johannes Werzowa, Julia Mascherbauer, Bernd Genser, Marcus D. Säemann, Diana Bonderman
• Format: Article
• Language: English
• Published: Karger Publishers 2017-04-01
• Series: Kidney & Blood Pressure Research
• Subjects: Hemodialysis; Ejection fraction; Heart failure
• Online Access: http://www.karger.com/Article/FullText/473868
• ISSN: 1420-4096; 1423-0143

Background/Aims: Heart failure (HF) is a main cause of mortality of hemodialysis (HD) patients. While HF with reduced ejection fraction (HFrEF) is known to only affect a minority of patients, little is known about the prevalence, associations with clinical characteristics and prognosis of HF with preserved ejection fraction (HFpEF). Methods: We included 105 maintenance HD patients from the Medical University of Vienna into this prospective single-center cohort study and determined the prevalence of HFpEF (per the 2013 criteria of the European Society of Cardiology) and HFrEF (EF <45%), using standardized post-HD transthoracic echocardiography. We also assessed clinical, laboratory and volume status parameters (by bioimpedance spectroscopy). These parameters served to calculate prediction models for both disease entities, while clinical outcomes (frequency of cardiovascular hospitalizations and/or cardiac death) were assessed prospectively over 27±4 months of follow-up. Results: All but 4 patients (96%) had evidence of diastolic dysfunction. 70% of the entire cohort fulfilled HF criteria (81% HFpEF, 19% HFrEF). Age, female sex, body mass index, blood pressure and dialysis vintage were predictive of HFpEF (sensitivity 86%, specificity 63%; AUC 0.87), while age, female sex, NT pro-BNP, history of coronary artery disease and atrial fibrillation were predictive of HFrEF (sensitivity 85%, specificity 90%; AUC 0.95). Compared to patients without HF, those with HFpEF and HFrEF had a higher risk of hospitalization for cardiovascular reason and/or cardiac death (adjusted HR 4.31, 95% CI 0.46-40.03; adjusted HR 3.24, 95% CI 1.08-9.75, respectively). Conclusion: Diastolic dysfunction and HFpEF are highly prevalent in HD patients while HFrEF only affects a minority. Distinct patient-specific characteristics predict diagnosis of either entity with good accuracy.