Hepatic T2-times on cardiovascular magnetic resonance imaging reflect liver fibrosis and predict outcome in an all-comer cohort

Abstract

Abstract Background Non-alcoholic fatty liver disease (NAFLD) is associated with dismal outcomes in patients with cardiac disorders but infrequently assessed by cardiologists. Cardiovascular magnetic resonance (CMR) is evolving as one-stop-shop imaging modality in cardiology, allowing for non-invasive myocardial tissue characterization by T1-mapping. On standard CMR exams, hepatic tissue is also assessable on T1-maps. However, it is unknown whether hepatic T1-times are associated with 1) myocardial T1-times, 2) established NAFLD scores, and 3) outcomes in patients referred for CMR. Methods In consecutive patients undergoing CMR we assessed hepatic and myocardial T1-times, and the NAFLD Fibrosis Score (NFS). Correlation analyses were used to test the association between hepatic and myocardial T1-times as well as the NFS. We used Kaplan-Meier estimates and Cox-regression models to investigate the association between hepatic T1-times and a composite endpoint of heart failure hospitalization and cardiovascular death. Results 513 patients were included (57±18 y/o, 49% female). Hepatic T1-times were 588±98ms on average and were correlated with myocardial T1-times (r=0.42, p<0.001) and – weakly – with the NFS (r=0.11, p=0.04). Patients with severe liver fibrosis or cirrhosis (n=47) had significantly higher hepatic T1-times as compared to patients with no or mild fibrosis based on the NFS (635±197ms versus 588±80ms, p=0.02). During follow-up (100±40 months), a total of 137 (27%) events occurred. When stratified by quartiles, patients in the highest hepatic T1-time quartile (>700ms) were at higher risk for events compared to all other quartiles (log-rank, p=0.01), which was consistent across different NAFLD risk groups based on the NFS (no/mild fibrosis, indeterminant score, severe fibrosis/cirrhosis). On Cox regression analyses, higher hepatic T1-times yielded significantly higher risk estimates for events (adj. HR 1.20 [95% CI: 1.04–1.38] per 1-SD increase, p=0.01) even when adjusted for age, sex, left and right ventricular ejection fractions, and myocardial T1-times. Conclusion Hepatic T1-times assessed on standard CMR reflect severity of NAFLD and predict outcome on top of established risk factors, including myocardial T1-times, in an all-comer CMR cohort. Funding Acknowledgement Type of funding sources: Public hospital(s). Main funding source(s): Medical University of Vienna