Prognostic performance of a two-step method using the Fibro-Scope system for metabolic dysfunction-associated steatotic liver disease

Abstract

Aim

Fibro-Scope is an artificial intelligence/neural network system for the noninvasive diagnosis of hepatic fibrosis in patients with metabolic dysfunction-associated steatotic liver disease. We aimed to examine the diagnostic performance of a two-step method that used the Fibrosis-4 (FIB-4) index and Fibro-Scope system for the assessment of Japanese patients with metabolic dysfunction-associated steatotic liver disease.

Methods

We analyzed a longitudinal study cohort of 796 Japanese patients with biopsy-proven metabolic dysfunction-associated steatotic liver disease during a follow-up period of 6.4 years. The predictive performance of the two-step method of FIB-4 index and Fibro-Scope for liver-related events and prognostic performance of that were assessed in the patients.

Results

In the 796 patients, by classifying the intermediate zone, defined by FIB-4 index 1.30–2.67, using the Fibro-Scope, the final classification was 69.6% low risk, 28.3% high risk, and 2.1% in the middle-risk group. The sensitivity and specificity for predicting advanced fibrosis (≥F3) was 84.0% and 84.0%. During the follow-up period, 52 (6.5%) patients developed liver-related events and 35 died. Multivariate analysis revealed that high-risk patients derived from the two-step method had hazard ratios of 30.1 or the development of liver-related events and 7.8 for outcome.

Conclusions

The two-step method using the FIB-4 index and Fibro-Scope contributed to improving the diagnostic performance by picking up high-risk patients from those classified as intermediate risk with the FIB-4 index. This noninvasive method, which uses a blood sample is a cost-effective screening method, is suitable for clinical practice in Japan.