Liver fibrosis progression analyzed with AI predicts renal decline

The relationship between biopsy-proven liver fibrosis progression and renal function decline in patients with metabolic dysfunction-associated steatotic liver disease (MASLD) has not been fully elucidated. We used an automated quantitative liver fibrosis assessment (qFibrosis) technique to investigate the temporal changes in regional liver fibrosis.

This retrospective longitudinal study included 68 MASLD patients and their paired formalin-fixed sections of liver biopsies. One hundred eighty-four fibrosis parameters were quantified in five different hepatic regions, including portal tract, peri-portal, zone 2, peri-central and central vein regions, and qFibrosis continuous values were calculated for all samples based on 10 fibrosis parameters using qFibrosis assessment. Liver fibrosis progression (QLF⁺, n = 18) and regression (QLF^-, n = 23) was defined as at least a 20% relative change in qFibrosis over a 23-month follow-up. Renal function decline was assessed by estimated glomerular filtration rate (eGFR) changes.

The eGFR decline was greater in the QLF⁺ group (106.53 ± 13.71 ml/min/1.73 m² vs. 105.28 ± 12.46 ml/min/1.73 m²) than in the QLF^- group (110.87 ± 14.58 ml/min/1.73 m² vs. 114.18 ± 14.81 ml/min/1.73 m²). In addition, liver fibrosis changes in the central vein and pericentral regions were more strongly associated with eGFR decline than in periportal, zone 2 and portal tract regions. We combined these parameters to construct a prediction model, which better differentiated eGFR decline (a cut-off value of qFibrosis combined index = 0.52, p <0.001).

A decline in renal function is significantly related to liver fibrosis progression in MASLD. Regional qFibrosis assessment may efficiently predict eGFR decline, thus highlighting the importance of assessing renal function in patients with MASLD with worsening liver fibrosis.

The study shows that liver fibrosis progression assessed by qFibrosis may be associated with renal function decline, which provides a new perspective for understanding complex pathological processes. A combination of artificial intelligence and digital pathology may earlier and more precisely quantify the progression of regional liver fibrosis, thus better identifying changes in renal function. This opens the possibility of early interventions, which are essential to improve patients' outcomes.