Radiomics for Predicting Lung Cancer Outcomes Following Radiotherapy: A Systematic Review

Area of Science:

• Oncology
• Medical Imaging
• Radiomics

Background:

• Lung cancer has a high mortality rate with no validated biomarkers for personalized radical radiotherapy.
• Radiomics offers potential for non-invasive biomarkers to guide treatment decisions.

Purpose of the Study:

• To systematically review radiomics studies in lung cancer to assess their potential for informing radical radiotherapy.
• To identify imaging features associated with clinical outcomes and evaluate the quality of existing research.

Main Methods:

• A PRISMA-guided systematic review of MEDLINE was conducted using terms 'radiomics' and 'lung cancer'.
• Studies analyzing at least 10 quantifiable imaging features and reporting on disease control, survival, or toxicity after radical radiotherapy were included.
• The Radiomics Quality Score was applied to assess study quality.

Main Results:

• 44 studies met eligibility criteria, frequently reporting on overall survival (79%) and distant control (31%).
• Texture features (Gray level run length matrix, Gray level co-occurrence matrix, kurtosis) were strongly associated with outcomes.
• The median Radiomics Quality Score was low (11%), with limited external validation and no patient-reported outcomes.

Conclusions:

• Radiomics-derived indices show promise as non-invasive biomarkers for predicting treatment response, failure patterns, toxicity, and survival in lung cancer.
• Standardization, reporting quality, and external validation in prospective trials are crucial for clinical implementation.