Feasibility of short imaging protocols for [¹⁸F]fluordeprenyl-D2 PET in autoimmune encephalitis and multiple system atrophy.

Abstract

[¹⁸F]fluorodeprenyl-D2 ([¹⁸F]F-DED) positron emission tomography (PET) imaging detects reactive astrogliosis in patients with autoimmune encephalitis (AIE) and multiple system atrophy (MSA). Although dynamic 60-min acquisitions are established, shorter static imaging protocols are desirable for severely impaired patients. This study investigated the feasibility of short static time windows for [¹⁸F]F-DED PET imaging in AIE and MSA. Dynamic 60-min [¹⁸F]F-DED PET scans were analyzed in 20 patients with AIE, 20 patients with MSA (MSA-P/MSA-C), and 16 controls (CTRL). Disease-related lesions were manually segmented based on visually detectable positive PET-signal in AIE and MSA predilection sites (i.e. mesial temporal lobe, posterior putamen, cerebellar deep white matter), and standardized uptake value ratios (SUVr; cerebellar cortex as a reference tissue) were calculated for consecutive 10-min intervals. Advanced kinetic parameters (DVR, VTr) were derived using Logan plot and a one-tissue compartment model (1TC2k) with image-derived input functions, both applying the cerebellar cortex as a reference tissue. Static images acquired between 10 and 60 min p.i. showed good image contrast and signal-to-noise ratio. SUVr of lesions increased over time and approached a plateau at approximately 50-60 min p.i.. The strongest agreement between SUVr and DVR was observed between 30 and 50 min p.i.. Late-phase SUVr outperformed kinetic parameters in discriminating lesions from healthy tissue in both AIE and MSA. Short static [¹⁸F⁸ ]F-DED PET acquisitions are clinically robust for detecting neuroinflammation in AIE and MSA. A late static acquisition between 30-50 min p.i. provides the optimal balance between accuracy and scanning efficiency.