A Preoperative MRI-Based Score to Predict High-Risk Pathology and Stratify Adjuvant Therapy Benefit in BCLC 0/A
Huiying Li1, Xinxin Wang1, Can Yu1
1Department of Radiology, Harbin Medical University Cancer Hospital, Harbin, Heilongjiang, 150081, People's Republic of China.
Purpose:
This study aims to develop and validate a magnetic resonance imaging (MRI)-based preoperative scoring model for predicting microvascular invasion (MVI) or Edmondson-Steiner grade 3/4 in BCLC 0/A hepatocellular carcinoma (HCC) and explore its potential for early recurrence risk stratification and as a hypothesis-generating tool for identifying patients who may benefit from adjuvant therapy.
Materials And Methods:
This multicenter retrospective study comprised consecutive patients with BCLC 0/A HCC who underwent contrast-enhanced MRI followed by curative resection at three different institutions between January 2015 and November 2024. To identify imaging predictors of MVI and Edmondson-Steiner grade 3/4, multivariable logistic regression was applied to the training set. Imaging features and AFP levels were combined to create a scoring system (HRI-score), and the model's ability to predict recurrence and stratify recurrence-free survival (RFS) was assessed in internal and external sets.
Results:
Overall, 994 patients were included (training: n = 601; internal validation: n = 257; external outcome set: n = 136). The HRI-score included irregular tumor morphology, nonperipheral washout, intratumoral blood products, and AFP (≥ 400 ng/mL). The training and internal validation sets had AUCs of 0.813 and 0.747, respectively. Across all sets, RFS was significantly lower in HRI-positive patients than for HRI-negative patients (all P < 0.05). Adjuvant therapy was associated with improved RFS in the HRI-positive subgroup; however, this effect was not observed in the HRI-negative subgroup, based on exploratory analyses.
Conclusion:
The HRI-score enables preoperative prediction of high-risk pathological features and risk stratification in BCLC 0/A HCC. It may also serve as a hypothesis-generating tool to identify patients who could potentially benefit from adjuvant therapy, warranting further prospective multicenter validation.