[¹⁸F]FDG PET/CT tumor habitat and metabolic hot spot analysis predicts progression in unresectable LA-NSCLC

Unresectable locally advanced non-small cell lung cancer (LA-NSCLC) presents significant challenges in treatment response prediction and risk stratification, leading to suboptimal patient outcomes. Current standard-of-care imaging and clinical markers often lack the precision needed to identify patients at high risk of progression or those who will benefit most from specific therapies. There is a critical need for non-invasive, robust biomarkers that can provide deeper insights into tumor biology and spatial heterogeneity. Advanced [¹⁸F]FDG PET/CT analysis, moving beyond simple SUV metrics, offers a promising avenue to characterize the tumor microenvironment and metabolic activity, potentially filling this prognostic gap.

This retrospective two-center study analyzed 403 patients with unresectable LA-NSCLC who underwent baseline [¹⁸F]FDG PET/CT before antitumor therapy. The cohort was split into training (n=241), validation (n=81), and held-out testing (n=81) groups. Researchers delineated four metabolic-density habitats using the Otsu thresholding algorithm to generate a Habitat Score. Metabolic Hot Spot Displacement (MHSD) and Edge Proximity Score (EPS) were quantified to describe the spatial localization of metabolic dominance. The predictive performance of the habitat model was compared against whole-tumor radiomics and clinical PET/CT models using the area under the receiver operating characteristic curve (AUC). Independent predictors of radiographic progression were identified via Fine-Gray competing risk regression.

The integrated tumor habitat model demonstrated superior predictive performance for treatment response and progression. It achieved AUCs of 0.843 (95% CI, 0.789-0.895) in the training cohort, 0.822 (95% CI, 0.725-0.904) in validation, and 0.847 (95% CI, 0.744-0.935) in the held-out testing cohort. This performance significantly outperformed both conventional radiomics and clinical PET/CT models. In multivariable Fine-Gray analysis, several factors emerged as independent predictors of radiographically confirmed progression:

Elevated pro-gastrin-releasing peptide (subdistribution hazard ratio [sHR], 1.86; 95% CI, 1.23-2.81; P=0.004), a higher Habitat Score per 0.1-unit increase (sHR, 1.49; 95% CI, 1.31-1.70; P<0.001), and EPS-SUVmax (sHR, 1.80; 95% CI, 1.13-2.88; P=0.014) were all strongly associated with increased risk of progression.