A Novel Method for Prognostic Risk Classification After Carbon-Ion Radiotherapy for Hepatocellular Carcinoma Using...

Area of Science:

• Oncology
• Radiotherapy
• Data Mining

Background:

• Hepatocellular carcinoma (HCC) treatment lacks established prognostic models for carbon-ion radiotherapy.
• Predicting cancer-specific survival (CSS) is crucial for optimizing patient management.

Purpose of the Study:

• To develop a risk classification system for CSS following carbon-ion radiotherapy in HCC patients.
• To utilize decision tree analysis (DTA) for identifying significant prognostic factors.

Main Methods:

• Retrospective analysis of 90 HCC patients treated with carbon-ion radiotherapy (2018-2022).
• Univariate, multivariate analyses, and DTA were employed to identify prognostic factors for CSS and progression-free survival (PFS).
• Irradiation doses were standardized at 60 Gy (relative biological effectiveness [RBE]) with varied fractionation based on tumor proximity to the gastrointestinal tract.

Main Results:

• Multivariate analysis identified dose fractionation and pretreatment alpha-fetoprotein levels as significant prognostic factors for PFS and CSS.
• Clinical stage and pretreatment protein induced by vitamin K absence or antagonist II values were significant prognostic factors for CSS.
• DTA successfully stratified patients into low-risk (100% 3-year CSS), intermediate-risk (73.3% 3-year CSS), and high-risk (44.4% 3-year CSS) groups.

Conclusions:

• Decision tree analysis offers a novel approach for risk stratification in HCC patients undergoing carbon-ion radiotherapy.
• The developed classification system, based on tumor markers and clinical stage, can aid in predicting CSS.
• This method provides a foundation for personalized treatment strategies and improved patient outcomes.