LAUSR

Supplemental Digital Content is available in the text. Purpose The existing periprosthetic joint infection (PJI) models have obvious limitations, and studies of PJI on animal models using PET/computed tomography (CT) for diagnosis are still lacking. Thus, the aim of this study was to establish a new PJI model and 18F-fluorodeoxyglucose (FDG) and 68Ga-fibroblast activation protein inhibitor (FAPI) were employed to study their performance. Methods A novel PJI model of rabbit was developed by placing two screws in the tibia and femur. Based on bacteria concentration, the animals were divided into five groups, control, 104, 105, 106 and 107. 18F-FDG and 68Ga-FAPI PET/CT were performed continuously in next 2 weeks and maximum standardized uptake value (SUVmax), mean standardized uptake value (SUVmean), metabolic target volume (MTV) and total lesion glycolysis/total lesion fibrosis were calculated as the metrics. Results As for SUVmax, all data of 18F-FDG were larger than that of 68Ga-FAPI in the same group for both weeks. For the performance of 18F-FDG, no definitive conclusion could be drawn for SUVmax and SUVmean. As for 68Ga-FAPI, the 104 group was significantly larger than 105, 106 and 107 groups for SUVmax and SUVmean in both weeks (P < 0.05). MTV of 68Ga-FAPI was found to be almost always larger than that of 18F-FDG in the same group. Conclusion The mechanism of 68Ga-FAPI is totally different from 18F-FDG and this unique property of 68Ga-FAPI shows a promising prospect in detecting infection boundary and may even distinguish a small number or a large number of bacterial infections.