*Article* **Diagnostic Performance of Dynamic Whole-Body Patlak [ 18F]FDG-PET/CT in Patients with Indeterminate Lung Lesions and Lymph Nodes**

**Matthias Weissinger 1,2, Max Atmanspacher 1, Werner Spengler 3, Ferdinand Seith 2, Sebastian Von Beschwitz 1, Helmut Dittmann 1, Lars Zender 3, Anne M. Smith 4, Michael E. Casey 4, Konstantin Nikolaou 2,5,6, Salvador Castaneda-Vega 1,7,\* and Christian la Fougère 1,5,6**


**Abstract:** Background: Static [18F]FDG-PET/CT is the imaging method of choice for the evaluation of indeterminate lung lesions and NSCLC staging; however, histological confirmation of PET-positive lesions is needed in most cases due to its limited specificity. Therefore, we aimed to evaluate the diagnostic performance of additional dynamic whole-body PET. Methods: A total of 34 consecutive patients with indeterminate pulmonary lesions were enrolled in this prospective trial. All patients underwent static (60 min p.i.) and dynamic (0–60 min p.i.) whole-body [18F]FDG-PET/CT (300 MBq) using the multi-bed-multi-timepoint technique (Siemens mCT FlowMotion). Histology and follow-up served as ground truth. Kinetic modeling factors were calculated using a two-compartment linear Patlak model (FDG influx rate constant = Ki, metabolic rate = MR-FDG, distribution volume = DV-FDG) and compared to SUV using ROC analysis. Results: MR-FDGmean provided the best discriminatory power between benign and malignant lung lesions with an AUC of 0.887. The AUC of DV-FDGmean (0.818) and SUVmean (0.827) was non-significantly lower. For LNM, the AUCs for MR-FDGmean (0.987) and SUVmean (0.993) were comparable. Moreover, the DV-FDGmean in liver metastases was three times higher than in bone or lung metastases. Conclusions: Metabolic rate quantification was shown to be a reliable method to detect malignant lung tumors, LNM, and distant metastases at least as accurately as the established SUV or dual-time-point PET scans.

**Keywords:** whole-body; dynamic PET; parametric FDG; Patlak; FDG; PET/CT
