Assessing in-hospital cardiovascular, thrombotic and bleeding outcomes in patients with chronic liver disease undergoing left ventricular assist device implantation.

Abstract

INTRODUCTION\nChronic liver disease (CLD) and advanced heart failure (HF) often co-exist with coagulopathy and hematologic abnormalities being major concerns in this cohort. Perioperative outcomes of patients undergoing LVAD implantation can be affected by coagulopathy, associated with a higher International Normalized Ratio (INR) and cytopenias, as well as pre-operative use of antiplatelet therapy and systemic anticoagulation. Our study is aimed at evaluating the in-hospital mortality and clinical outcomes of patients with CLD who underwent LVAD implantation compared to patients who underwent LVAD implantation without CLD.\n\n\nMETHODS\nThe National Inpatient Sample Database was queried from 2012 to 2017 for relevant International Classification of Diseases (ICD)-9 and ICD-10 procedural and diagnostic codes. Baseline characteristics and in-hospital outcomes were compared in patients with chronic liver disease and those without, who underwent LVAD implantation.\n\n\nRESULTS\nA total of 22,955 patients underwent LVAD implantation, 2200 of which had CLD. There was no difference in mean age between those with and without CLD (52.8\xa0±\xa014.2 vs. 55.7\xa0±\xa015.4\xa0years old, p\xa0<\xa00.001), and 23.7% of patients were female. The proportion of patients with CLD undergoing LVAD implantation trended downward between 2012 and 2017 (average annual growth rate: -14.8% ). In-hospital post-LVAD outcomes revealed: all-cause inpatient mortality (14.8% vs. 11.1%), major bleeding (34.3% vs. 30.2%), transfusion of platelets (18.0% vs. 14.0%), subarachnoid hemorrhage (1.6% vs. 0.7%) and hospital length of stay were greater in patients with CLD (p\xa0<\xa00.001 for all values). LVAD thrombosis (6.6% vs. 9.4%) and postoperative ischemic stroke (3.4% vs. 6.1%) occurred less in patients with CLD (p\xa0<\xa00.001 for both). There were no statistically significant differences in occurrence of post-LVAD gastrointestinal bleeding and transfusion of fresh frozen plasma or packed red blood cells (p\xa0>\xa00.05 for all). Using a multivariate logistic regression model to adjust for confounding factors, CLD was predictive of increased in-hospital all-cause mortality in patients undergoing LVAD implantation (adjusted odds ratio: 1.29, 95% confidence interval [CI]; 1.06 to 1.56, p\xa0=\xa00.010).\n\n\nCONCLUSION\nLVAD implantation in patients with chronic liver disease was associated with increased mortality and post-LVAD major bleeding with increased utilization of platelet products yet comparable thrombotic complications. Further studies are needed to evaluate the balance and pathophysiology of bleeding risks when compared to thrombosis, as well as predictors in patients with chronic liver disease.