Can Dynamic Whole-Body FDG PET Imaging Differentiate between Malignant and Inflammatory Lesions?

Background: Investigation of the clinical feasibility of dynamic whole-body (WB) [F-18]FDG PET, including standardized uptake value (SUV), rate of irreversible uptake (Ki), and apparent distribution volume (Vd) in physiologic tissues, and comparison between inflammatory/infectious and cancer lesions. Methods: Twenty-four patients were prospectively included to undergo dynamic WB [F-18]FDG PET/CT for clinically indicated re-/staging of oncological diseases. Parametric maps of Ki and Vd were generated using Patlak analysis alongside SUV images. Maximum parameter values (SUVmax, Ki(max), and Vd(max)) were measured in liver parenchyma and in malignant or inflammatory/infectious lesions. Lesion-to-background ratios (LBRs) were calculated by dividing the measurements by their respective mean in the liver tissue. Results: Seventy-seven clinical target lesions were identified, 60 malignant and 17 inflammatory/infectious. Ki(max) was significantly higher in cancer than in inflammatory/infections lesions (3.0 vs. 2.0, p = 0.002) while LBRs of SUVmax, Ki(max), and Vd(max) did not differ significantly between the etiologies: LBR (SUVmax) 3.3 vs. 2.9, p = 0.06; LBR (Ki(max)) 5.0 vs. 4.4, p = 0.05, LBR (Vd(max)) 1.1 vs. 1.0, p = 0.18). LBR of inflammatory/infectious and cancer lesions was higher in Ki(max) than in SUVmax (4.5 vs. 3.2, p < 0.001). LBRs of Ki(max) and SUVmax showed a strong correlation (Spearman's rho = 0.83, p < 0.001). Conclusions: Dynamic WB [F-18]FDG PET/CT is feasible in a clinical setting. LBRs of Ki(max) were higher than SUVmax. Ki(max) was higher in malignant than in inflammatory/infectious lesions but demonstrated a large overlap between the etiologies.