18F-FDG PET/CT-based Metabolic-clinical prognostic model could refine risk stratification in high-risk neuroblastoma

Abstract

The prognosis of high-risk neuroblastoma (NB) is highly heterogeneous, underscoring the necessity of refining risk stratification to guide precision treatment and improve outcomes. The aim of this study was to construct an 18 F-FDG PET/CT-based Metabolic-clinical prognostic model to refine risk stratification in high-risk NB. This retrospective study enrolled 72 (41 boys and 31 girls) patients with newly diagnosed high-risk NB. The median age was 4.03 years (interquartile range 2.52–4.95 years). Patients underwent baseline 18 F-FDG PET/CT before receiving any therapy. Two experienced nuclear medicine physicians reviewed the PET/CT images to identify and delineate tumor lesions. Semi-quantitative PET/CT parameters were extracted from lesions, including the maximum standardized uptake value (SUVmax), total lesion glycolysis (TLG) and intra-tumoral metabolic heterogeneity (IMH), et al. Univariate and multivariate Cox proportional hazards regression analyses were used to identify independent prognostic risk factors and construct the Metabolic-clinical model. The optimal cut-off score for Metabolic-clinical model was determined by receiver operating characteristic curve (ROC). 27 (37.5%) patients remained event-free, while 45 (62.5%) patients experienced relapse or death. Multivariate Cox regression analysis revealed that chromosome 11q deletion ( P = 0.016), NSE ( P = 0.003) and IMH ( P = 0.030) were independent prognostic factors for event-free survival. The chromosome 11q, NSE, and IMH were incorporated into Metabolic-clinical scoring system. Based on the Metabolic-clinical scoring system, all high-risk NB patients were stratified into high- and low-score groups. A significant difference in event-free survival (EFS) was observed between the two groups ( P < 0.001). The median survival time was 39.37 months for low score group versus 15.57 months for the high score group. The 18 F-FDG PET/CT-based Metabolic-clinical prognostic model can refine the risk stratification of high-risk NB, enabling more precise risk assessment. Not applicable.