A CT-based Radiomics Signature Is Associated with Response to Immune Checkpoint Inhibitors in Advanced Solid Tumors.

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• Immune checkpoint inhibitors (ICIs) have revolutionized cancer treatment.
• Reliable pretreatment imaging biomarkers are crucial for predicting patient response to ICIs.
• Current methods lack sufficient predictive accuracy for diverse solid tumors.

Purpose of the Study:

• To develop and validate a computed tomography (CT)-based radiomics signature.
• To predict treatment response to anti-programmed cell death-1 or programmed cell death ligand-1 therapies in advanced solid tumors.
• To assess the signature's association with tumor immunophenotype.

Main Methods:

• Retrospective development of a radiomics signature in 85 patients with advanced solid tumors (cohort 1).
• Validation in independent cohorts of bladder (n=46, cohort 2) and lung cancer (n=47, cohort 3) patients.
• Radiomics variables extracted from pretreatment CT scans of metastases, selected using an elastic-net model.
• A regression model integrated radiomics and clinical variables (albumin, lymphocyte count).
• Biologic validation performed using RNA profiling of cytotoxic cells (n=20, cohort 4).

Main Results:

• The radiomics signature showed significant association with clinical response (AUC, 0.70).
• Validation cohorts demonstrated consistent performance (AUCs of 0.67 for cohorts 2 and 3).
• A combined radiomics-clinical signature improved predictive performance (AUC, 0.74).
• The radiomics score correlated with RNA profiling of cytotoxic cells, suggesting immunophenotypic relevance.

Conclusions:

• A pretreatment CT-based radiomics signature can predict response to immune checkpoint inhibitors in advanced solid tumors.
• The signature likely reflects the underlying tumor immune microenvironment.
• Integrating clinical variables enhances predictive accuracy, supporting its potential clinical utility.