Liver stiffness as measured by transient elastography is an independent predictor of outcomes in patients with chronic heart failure

Abstract

Listen

Translate article

Abstract Background Patients with chronic heart failure (HF) may have liver abnormalities due to systemic congestion. Transient elastography (TE) has been used to assess liver stiffness (LS) in patients with primary liver disease. Some studies have suggested that congestion can influence liver stiffness. Purposes We sought to assess the role of LS in the prediction of outcomes in patients with chronic HF. Methods Nine-three consecutive patients with chronic HF were screened and fulfilled the inclusion criteria which consisted of signs or symptoms of HF and left ventricle ejection fraction (LVEF) less than 50%. Patients with concomitant liver disease were excluded or techical problems during TE. Patients underwent routine laboratory tests, liver function tests and TE. The primary endpoint was time to first event, which was defined as a composite of cardiovascular death or HF hospitalization. Results Mean follow up was 219±86 days. Nine patients were excluded due to exclusion criteria or technical problems during the TE. Eighty-four patients were included in the final analysis. Mean age was 63.2±12.2 and 57 (67.8%) were male. Mean ejection fraction and median NT-proBNP were, respectively, 38.7±14.3% and 1,140 pg/mL (interquartile range 224.3–2,810.3). Median LS for the entire population was 5.35 (3.7–10.65) kPa. LS correlated with NT-proBNP (r=0.54; p<0.0001), total bilirubin (r=0.47; p<0.001), direct bilirubin (r=0.66; p<0.0001), alkaline phosphatase (r=0.57; p<0.0001); γ-glutamyl-transpeptidase (r=0.59; p<0.0001), and age (r=−0.22; p=0.03). A Receiver Operating Characteristic (ROC) curve was performed and a cut point of 5.9 kPa showed sensitivity of 80% and specificity of 64.1% with area under the curve of 0.73. Mean event-free survival time for patients above and below this cutoff was 215.2±20 vs 302.8±7.2 days (p=0.0001; log rank test). Using Cox proportional hazard model (independent variables: LS as a continuous variable, age, gender, NT-proBNP, LVEF, and creatinine), only LS was independently associated with the primary endpoint (hazard ratio 1.05, 95% confidence interval 1.01–1.09; for each increment of one unit of LS). Conclusion LS correlates with biomarkers of myocardial stretch and several liver function tests and is an independent predictor of outcomes in patients with chronic HF. Funding Acknowledgement Type of funding source: Public hospital(s). Main funding source(s): Own funding and examinations carried out with a welfare fund.