Evaluation of Recurrence Risk in Irreversible Electroporation-Treated Pancreatic Adenocarcinoma Patients Using...

Area of Science:

• Radiology
• Oncology
• Medical Imaging Analysis

Background:

• Locally advanced pancreatic cancer (LAPC) presents treatment challenges.
• Irreversible electroporation (IRE) is a therapeutic option for LAPC.
• Predicting treatment effectiveness and patient outcomes is crucial for personalized care.

Purpose of the Study:

• To determine if radiomics signatures from longitudinal CT scans can predict IRE treatment effectiveness in LAPC patients.
• To correlate radiomics features with key patient outcomes such as recurrence and survival.

Main Methods:

• Retrospective analysis of 50 LAPC patients treated with IRE.
• Extraction of 2078 radiomics features (shape, texture, filter, intensity, local texture) from preoperative and follow-up CT scans.
• Application of Principal Component Analysis (PCA) for composite feature generation.
• Correlation of radiomics signatures with time to recurrence (TTR), recurrence-free survival (RFS), and overall survival (OS) using hazard ratios (HRs) and log-rank tests.

Main Results:

• Significant association found between radiomics features and TTR risk groups.
• Gray-level co-occurrence matrix features showed a significant HR of 2.65 (p < 0.01).
• Composite radiomics features, particularly intensity and filter features, demonstrated strong predictive power (HRs ranging from 2.27 to 3.13, p < 0.01).

Conclusions:

• Pre-treatment radiomics signatures derived from CT scans are significantly associated with LAPC patient outcomes after IRE.
• These radiomics features can predict individual patient risk of disease recurrence, enabling risk stratification.
• The findings highlight the potential of radiomics in personalizing treatment strategies for LAPC.