A novel analytical approach for outcome prediction in newly diagnosed NSCLC based on [18F]FDG PET/CT metabolic parameters, inflammatory markers, and clinical variables

Abstract

Objectives: To develop a novel analytical approach based on ¹⁸F-fluorodeoxyglucose ([¹⁸F]FDG) positron emission tomography (PET) metabolic parameters, serum inflammatory markers, and clinical variables to improve the outcome prediction in NSCLC.

Methods: A total of 190 newly diagnosed NSCLC patients who underwent pretreatment [¹⁸F]FDG PET/CT were retrospectively enrolled and divided into a training cohort (n = 127) and a test cohort (n = 63). Cox regression analysis was used to investigate the predictive values of PET metabolic parameters, inflammation markers, and clinical variables for progression-free survival (PFS) and overall survival (OS). Based on the results of multivariate analysis, PET-based, clinical, and combined models were constructed. The predictive performance of different models was evaluated using time-dependent ROC curve analysis, Harrell concordance index (C-index), calibration curve, and decision curve analysis.

Results: The combined models incorporating SULmax, MTV, NLR, and ECOG PS demonstrated significant prognostic superiority over PET-based models, clinical models, and TNM stage in terms of both PFS (C-index: 0.813 vs. 0.786 vs. 0.776 vs. 0.678, respectively) and OS (C-index: 0.856 vs. 0.792 vs. 0.781 vs. 0.674, respectively) in the training cohort. Similar results were observed in the test cohort for PFS (C-index: 0.808 vs. 0.764 vs. 0.748 vs. 0.679, respectively) and OS (C-index: 0.836 vs. 0.785 vs. 0.726 vs. 0.660, respectively) prediction. The combined model calibrated well in two cohorts. Decision curve analysis supported the clinical utility of the combined model.

Conclusions: We reported a novel analytical approach combining PET metabolic information with inflammatory biomarker and clinical characteristics, which could significantly improve outcome prediction in newly diagnosed NSCLC.

Key points: • The nomogram incorporating SULmax, MTV, NLR, and ECOG PS outperformed the TNM stage for outcome prediction in patients with newly diagnosed NSCLC. • The established nomogram could provide refined prognostic stratification.

Keywords: Metabolic tumor volume; Non-small-cell lung cancer; Positron emission tomography (PET); Prognosis; TNM stage.