A Practical, Short, [¹⁸F]F-DOPA PET/CT Acquisition Method for Distinguishing Recurrent Brain Metastases from...

Background/Objectives: Determining whether 3,4-dihydroxy-6-[¹⁸F]fluoro-L-phenylalanine positron emission tomography/computed tomography ([¹⁸F]F-DOPA PET/CT) data indicate brain metastasis progression (MP) or brain radionecrosis (RN) is challenging. The aim of this study was to present a method usable in the clinical setting without dedicated software that relies on less than five minutes of SiPM PET/CT data collected immediately after [¹⁸F]F-DOPA injection. Methods: We prospectively enrolled 15 patients with 19 lesions. Each acquisition was conducted in list mode (LM) for 25 min using a four-ring SiPM PET/CT system. We reconstructed three volumes from the LM file: acquisition duration of 120 s (V120), 270 s (V270), and 10 min for the standard clinical volume (Vclin). We measured each lesion's maximum voxel activity (LS_max) and the corresponding mean activity with its standard deviation (CL_mean, CL_sd). We then calculated the LS_max/CL_mean ratio and the coefficient of variation (COV), defined as 100 × (CL_sd/CL_mean). Results: Lesions were classified as RN (n = 7) and MP (n = 12). For all volumes, LS_max/CL_mean differed significantly. The COV parameter exhibited significant differences in all comparisons except for V120 vs. V270. The best diagnostic performances were observed for V120 and V270, with an accuracy of 94.7%. The AUC values were 97.6% in both cases. Conclusions: A simple, static [¹⁸F]F-DOPA PET/CT acquisition, starting 1.5 min after injection and lasting less than five minutes, permitted reaching excellent diagnostic performance (100% sensitivity, 91.7% specificity, and 97.6% AUC) in discriminating between RN and MP.