CT-Based Nested Habitats Analysis for Early Recurrence Prediction and Risk Stratification in Hepatocellular...

Area of Science:

• Radiology and Oncology
• Medical Imaging Analysis
• Cancer Research

Background:

• Hepatocellular carcinoma (HCC) poses a significant challenge due to early recurrence.
• Identifying aggressive tumor subregions is crucial for predicting patient outcomes.
• Current methods may not fully capture intratumoral heterogeneity.

Purpose of the Study:

• To develop and validate a computed tomography (CT)-based nested habitats analysis.
• To identify aggressive tumor subregions within HCC.
• To predict early recurrence in HCC patients.

Main Methods:

• A nested habitats analysis was performed on venous-phase CT images from training, internal validation, and external validation cohorts.
• Support vector machine and k-means clustering were used to identify aggressive micro-regions.
• A nested habitats score was generated by integrating features from high-risk micro-regions.

Main Results:

• The nested habitats score showed strong predictive ability for early recurrence with AUCs ranging from 0.832 to 0.896 across cohorts.
• The score was an independent predictor of recurrence-free survival (RFS) in multivariable analysis.
• Kaplan-Meier analysis confirmed significantly shorter RFS in patients with high nested habitats scores.

Conclusions:

• CT-based nested habitats analysis effectively captures intratumoral heterogeneity in HCC.
• This method accurately predicts early recurrence, enabling precise postoperative risk stratification.
• The technique offers a valuable tool for personalized HCC management.