Can MRI features predict clinically relevant hepatocellular carcinoma genetic subtypes?

Area of Science:

• Hepatocellular Carcinoma Genomics
• Medical Imaging
• Oncology

Background:

• Hepatocellular carcinoma (HCC) pathogenesis is increasingly understood through cancer genomics.
• Identifying genetic subtypes of HCC is crucial for prognosis and treatment.
• Non-invasive methods for predicting HCC genetic subtypes are needed.

Purpose of the Study:

• To investigate if Magnetic Resonance Imaging (MRI) features can serve as non-invasive markers for predicting common genetic subtypes of HCC.
• To correlate specific MRI findings with prevalent genetic mutations in HCC.

Main Methods:

• 43 HCC patients underwent contrast-enhanced MRI and genetic sequencing of 447 cancer-implicated genes.
• Retrospective evaluation of MRI features including tumor margin, enhancement patterns, and heterogeneity.
• Statistical analysis (Fisher's exact test) to correlate genetic mutations (TP53, CTNNB1) with MRI features.

Main Results:

• TP53 mutation (30%) was associated with an infiltrative tumor margin on MRI (p=0.01), with high inter-reader agreement (kappa=0.95).
• CTNNB1 mutation (40%) correlated with peritumoral enhancement on MRI (p=0.04), showing substantial inter-reader agreement (kappa=0.74).
• MRI features demonstrated predictive accuracy for these genetic subtypes.

Conclusions:

• An infiltrative tumor margin on MRI is a significant predictor of TP53 mutation in HCC.
• Peritumoral enhancement on MRI is associated with CTNNB1 mutation in HCC.
• These MRI findings offer potential non-invasive biomarkers for HCC genetic subtyping, guiding prognosis and treatment decisions.