Kinetic Analysis of Total-body 18F-FDG PET/CT for the Evaluation of Benign and Malignant Pulmonary Lesions
Nan Meng,Fangfang Fu,Yaping Wu,Zhun Huang,Ting Fang,Pengyang Feng,Yan Bai,Wei,Tao Feng,Tianyi Xu,Meiyun Wang
2021-01-01
Journal of Nuclear Medicine
Abstract:1699 Introduction: Lung cancer remains the cause of the highest morbidity and the main cause of cancer-related mortality worldwide. The EXPLORER consortium has developed the world’s first 2-meter long total-body PET scanner (uEXPLORER) to provide a versatile platform for biomedical research and clinical applications. This work aims to determine if the kinetic analysis of total-body 18F-FDG PET/CT would be useful in differentiating benign and malignant pulmonary lesions. Methods: A total of 44 patients (26 males and 18 females; age range, 41-76 years) with 87 pulmonary lesions were selected in this research, including 46 malignant lesions (squamous cell carcinomas, adenocarcinomas, metastases, and small cell carcinomas) and 41 benign lesions (organizing pneumonia, tuberculoma, and hamartoma). The total-body 18F-FDG PET/CT scanner (uEXPLORER, UIH, Shanghai, China) was used for kinetic analysis on patients. A frontal localizer scan was performed first with an 80kv tube voltage and 20mA tube current, then a low dose CT with a slice thickness of 5mm and slice increment of 5mm to provide the attenuation information for the PET image reconstruction. A dynamic PET scan of 60 minutes was acquired subsequently. Diagnostic CT was scanned at last if needed. The FDG is given on a dosage of 3.7MBq/kg (0.1mCi/kg) and is to be injected right after the PET scan starts and through the injection site on the patient’s leg, preferably on the leg with less varicosity. The 60-minute PET acquisition was reconstructed into the following dynamic frames: starting at 0s post-injection, 2s*50, 10s*20, 30s*10, 1min*10, 5min*8. The reconstructions were using a 192×192 matrix, with both ToF and PSF, using 20 subsets and 3 iterations. An image-derived input function is derived from the dynamic images using a volume-of-interest on the descending aorta. A set of parametric images were produced using two methods. The net flux rate Ki images were generated by the uKinetics software (United Imaging Healthcare) using the linear Patlak model. The transport constant K1 and rate constants K2, an avascular fraction (VB), and the time delay (TD) were estimated jointly via a non-linear 1-tissue compartment model using early phase data (5 minutes post-injection). The regions of interest (ROIs) were drawn manually in target areas over the tumor. SPSS 23.0 and Medcalc 15.0 were used for data analysis. The Mann-Whitney U test was applied for between-group analyses. The ROC curve was generated to evaluate each parameter’s diagnostic accuracy and threshold. The Delong method was used to compare the AUCs of different parameters. P 0.05] (Figure.1. 2). For the diagnosis of benign and malignant lesions, AUC (Ki) > AUC (SUVmax) > AUC (VB) > AUC (K2) > AUC (K1), where the differences between AUC (Ki) and AUC (SUVmax) and AUC (VB) and AUC (K2) and AUC (K1) were statistically significant (Z = 1.997, 3.911, 4.621, and 4.854, all P < 0.05), and the differences between AUC (SUVmax) and AUC (VB) and AUC (K2) and AUC (K1) were statistically significant (Z = 2.955, 3.861, and 4.336, all P < 0.05) (Figure.3).Conclusion: This study demonstrates that the kinetic analysis of total-body 18F-FDG PET/CT can be used to assess benign and malignant pulmonary lesions, and the diagnostic efficiency of Ki is better than SUVmax.Acknowledgments: All team members from Henan Provincial People’s Hospital and United Imaging Healthcare.