Prognostic Value of a Novel Nomogram Combining 18F-Alfatide PET/CT with Inflammatory Biomarkers in Patients with Advanced NSCLC Receiving Bevacizumab Combination Therapy

Abstract

Imaging of angiogenesis based on 18F-alfatide PET/CT could predict the tumor sensitivity to CCRT and antiangiogenic therapy. We attempted to evaluate its predictive role in advanced NSCLC patients treated by bevacizumab and chemotherapy. Serum inflammatory biomarkers, representing the systemic status, are shown to have complicated relationships with tumor angiogenesis. A novel nomogram integrating molecular imaging with 18F-alfatide PET/CT and serum inflammation biomarkers was built to explore its predictive role. Patients with advanced NSCLC underwent 18F-RGD PET/CT examination and serum biochemistry test before the start of bevacizumab combination therapy. PET/CT parameters including maximum and mean standard uptake values (SUVmax/SUVmean), peak standard uptake values (SUVpeak) and metabolic tumor volume (MTV) of contoured tumor lesions and inflammation biomarkers including neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and lymphocyte-to-monocyte ratio (LMR) were systemic collected. Receiver-operating characteristic (ROC) curve analysis was used to determine accuracy in predicting response. Univariate and multivariate Cox models were used for Progression (PFS) analyses to determine the independent prognosis factors. Nomograms were then established for predicting prognosis. 22 patients with 40 measurable lesions met the inclusion criteria in present study. The median progression-free survival (mPFS) was 3 months (95% confidence interval, 0.37-10.2). The SUVmax, SUVmean, SUVpeak and MTV were significantly higher in responding tumors than in non-responding tumors (p = 0.001, p = 0.007, p = 0.005, p = 0.003, respectively). The area under the ROC curve (AUC) were 0.7294 (p = 0.038), 0.7061 (p = 0.626), 0.7419 (p = 0.029) and 0.7957 (p = 0.008) for SUVmax, SUVmean, SUVpeak and MTV, respectively. Multivariate analysis revealed that SUVmax, MTV and LMR were independent predictors of PFS (p = 0.022, p = 0.032, and p = 0.009, respectively). The nomogram for PFS was established by combining all significant markers. The concordance index (C-index) for PFS was 0.861 (95%CI 0.828-0.894), which showed superior discriminatory power. 18F-RGD uptake on PET/CT imaging can effectively predict the response of bevacizumab combination therapy. The proposed nomogram based on molecular imaging with 18F-RGD PET/CT and serum inflammatory biomarkers with superior prediction ability could provide guidance for treatment decisions.