# Detection of macrotrabecular-massive hepatocellular carcinoma based on viscoelastic characteristics obtained by multifrequency magnetic-resonance elastography

## TL;DR

This study explores how magnetic resonance elastography can detect a specific type of liver cancer called macrotrabecular-massive hepatocellular carcinoma by measuring tumor stiffness and linking it to gene activity.

## Contribution

The study introduces tumor viscoelasticity measured by MRE as a novel non-invasive biomarker for identifying macrotrabecular-massive HCC and links it to distinct gene expression patterns.

## Key findings

- Tumor viscoelasticity (T-c) is an independent predictor of macrotrabecular-massive HCC with an AUC of 0.818.
- High T-c tumors show upregulated proliferation genes and downregulated immune regulation genes.
- Combining T-c with ≥20% arterial phase hypovascular component improves diagnostic accuracy but not significantly.

## Abstract

To investigate the use of viscoelastic characteristics obtained with magnetic-resonance elastography (MRE) in identifying the macrotrabecular-massive (MTM) subtype of hepatocellular carcinoma (HCC) and its association with gene expression profiles.

Fifty-one patients (mean age, 56.2 ± 12.6 years; 42 men) with histologically proven HCCs (16 with the MTM subtype, and 35 without) and 47 healthy participants (mean age, 54.1 ± 13.7 years, 24 men) underwent preoperative MRI and MRE examinations and were prospectively enrolled. Tumor viscoelasticity (comprising c and φ), imaging features and clinical information were analyzed and diagnostic models developed. Logistic regression and area-under-the-curve (AUC) methodology evaluated the models’ efficacy for determining the MTM-HCC. RNA sequencing and KEGG pathway analyses identified differential gene expression between 12 high-c and 12 low-c tumor samples.

In HCC patients with elevated Edmondson–Steiner grades, satellite nodules, non-smooth margins, fat deficiency, or an arterial phase hypovascular component (APHC) more than 20%, tumor viscoelastic values c or φ were higher, compared with patients without these features (p < 0.05). Tumor c (T-c) was an independent predictor of MTM-HCC (AUC, 0.818; 95% confidence interval: 0.685, 0.950; p < 0.001); Combining T-c with ≥ 20% APHC yielded a higher AUC (0.843), but not significantly different from T-c alone (p = 0.533). RNA sequencing showed high-c tumors upregulated cell proliferation and DNA replication genes but downregulated immune regulation genes.

MRE-derived T-c is a promising non-invasive biomarker for identifying MTM-HCC. HCCs with different T-c levels show distinct gene expression profiles, particularly in proliferation and immune pathways. Research with larger cohorts is needed to validate clinical utility.

Question
Can MRE-based viscoelastic values identify the macrotrabecular-massive (MTM) subtype of HCC?

Findings
Tumor-c based on MRE has a unique diagnostic performance for identifying MTM-HCC; tumor stiffness correlates with proliferative and immune gene expression.

Clinical relevance
MRE-based stiffness is a noninvasive predictor of MTM-HCC, and high-stiffness tumors show upregulation of proliferation genes and downregulation of immune genes. These findings may guide personalized treatment, but larger studies are required to confirm clinical applicability.