Tumor Characterization Using [¹⁸F]FDG PET Radiomics in a PD-L1-Positive NSCLC Cohort.

Area of Science:

• Oncology
• Radiology
• Medical Imaging

Background:

• Durvalumab consolidation is standard for unresectable stage III non-small cell lung cancer (NSCLC).
• [18F]FDG PET/CT provides insights for staging and tumor characterization via radiomics.
• Radiomics may predict histology, immunophenotype, and prognosis in NSCLC.

Purpose of the Study:

• To evaluate the potential of [18F]FDG PET/CT radiomics in differentiating NSCLC histological subtypes.
• To identify PD-L1-associated imaging patterns in NSCLC.
• To explore the association between radiomic features and neutrophil-to-lymphocyte ratio (NLR)-based prognostic groups.

Main Methods:

• Retrospective analysis of [18F]FDG PET/CT scans from 56 stage IIIA-IIIB NSCLC patients.
• Histological classification (squamous vs. adenocarcinoma) and PD-L1 testing of biopsy samples.
• Tumor segmentation using MEDISO InterViewTM FUSION software with an SUVmax threshold of four.
• Extraction and statistical comparison of imaging features based on histology, PD-L1 status, and NLR groups.

Main Results:

• Radiomic features differed between poor and good NLR prognosis groups in PD-L1-positive NSCLC.
• Significant differences in radiomic features were observed between squamous cell carcinoma and adenocarcinoma subtypes, particularly in PD-L1-positive cases.
• Higher SUVmax, mean SUV, SUVpeak, and complexity were found in squamous cell carcinoma compared to adenocarcinoma.
• Specific radiomic features distinguished PD-L1-positive from PD-L1-negative squamous tumors.

Conclusions:

• [18F]FDG PET/CT radiomics shows promise in differentiating NSCLC histological subtypes.
• Radiomics can identify PD-L1-associated imaging patterns, especially in squamous cell carcinoma.
• Certain metabolic radiomic features correlate with NLR-based prognostic groups in PD-L1-positive NSCLC patients.