Functional Liver Imaging Score derived from gadoxetic acid-enhanced MRI predicts outcomes in patients with chronic liver disease

Background and aims:

Non-invasive methods for stratifying the risks of first hepatic decompensation and mortality in patients with compensated (cACLD) and decompensated (dACLD) advanced chronic liver disease, respectively, might facilitate individualized therapy. The Functional Liver Imaging Score (FLIS) derived from the hepatobiliary phase (HBP) of gadoxetic acid-enhanced MRI (GA-MRI) does neither require tedious measurements or calculations, nor specific hard- or software, and thus, is easily applicable in clinical routine.

We aimed to investigate the predictive value of FLIS for hepatic decompensation and transplant-free survival in a large series of patients with chronic liver disease (CLD).

Methods:

Our study comprised 262 consecutive patients with CLD but without malignancy who underwent a standardized GA-MRI protocol. Three radiologists blinded to the clinical data evaluated the FLIS components (0 – 2 points each) in the HBP 20 min after contrast administration: (1) contrast enhancement and (2) excretion as well as (3) portal vein sign.

Patients were stratified into three groups according to FIB-4 and the history/presence of hepatic decompensation: non-advanced CLD (FIB-4≤1.45), cACLD (FIB-4> 1.45), and dACLD. Within these strata, patients with FLIS indicative of poor (0 – 3 points) and good (4 – 6 points) liver function were compared.

Results:

Intra- (κ= 0.983, 95% confidence interval [95%CI]: 0.971 – 0.991) and inter-observer (κ= 0.931, 95%CI: 0.898 – 0.954) agreement for FLIS was excellent.

Patients with non-ACLD (n = 56; 0 – 3 points: 25%), cACLD (n = 110; 0 – 3 points: 15%), and dACLD (n = 99; 0 – 3 points: 46%) were followed for a median of 40.7, 40.6, and 13.7 months, respectively. As expected, patients with non-advanced CLD had a negligible risk of clinical events. In cACLD patients, the FLIS was independently predictive of hepatic decompensation (0 – 3 vs. 4 – 6 points; adjusted hazard ratio [aHR]: 3.724,95%CI: 1.098 – 12.635; P= 0.035; panel A). Moreover, a FLIS score of 0 – 3 points emerged as an independent risk factor for transplant-free mortality in both cACLD (aHR: 7.437,95% CI: 2.742 – 20.17; P< 0.001; panel B) and dACLD patients (aHR: 3.839,95%CI: 1.155 – 9.477; P= 0.004; panel C).

Fig. 1

Conclusions:

The FLIS is an easy-to-use and highly reproducible imaging biomarker for the development of first hepatic decompensation and mortality in patients with ACLD.