Abstract
BACKGROUND: While abnormalities of liver function and imaging are common in patients with end-stage heart failure, advanced fibrosis is uncommon. Liver biopsy (LB) is used to identify advanced fibrosis in heart transplant (HT) candidates but can delay or limit access to definitive therapies and cause complications. We sought to develop and determine the utility of a clinical risk score for advanced fibrosis in HT candidates. METHODS: We conducted a retrospective, single-center review of patients evaluated for HT between 2012 and 2019 (n = 1,651) and identified those who underwent LB (n = 137) as well as a matched control cohort (n = 160). Patients with congenital heart disease were excluded. All biopsies were reviewed by a liver pathologist. Univariate logistic modeling was used to identify factors predictive of advanced liver fibrosis. Simulation using synthetic data bootstraps was performed to determine the utility of using a score-based approach to trigger LB. Kaplan-Meier curves were used to assess survival. RESULTS: We identified 32 (23%) patients with stage 0, 79 (58%) with stage 1 to 2, and 26 (19%) with stage 3 to 4/advanced fibrosis. The factor most associated with pursuit of LB was abnormal liver parenchyma on ultrasound. We found that a score combining severe tricuspid regurgitation, alcohol use, and low-density lipoprotein improved specificity and reduced the number of LBs required. We found no difference in survival at 3 years post-HT based on pre-HT fibrosis stage. CONCLUSIONS: A score composed of noninvasive factors may help reduce the number of patients who require LB for diagnosis of advanced fibrosis. Additional multicenter studies are needed to validate this score.