Abstract
BACKGROUND: The metabotropic glutamate receptor 4 (mGlu4) has been proposed as a target for Parkinson’s disease to measure levodopa-induced dyskinesia. [(11)C]PXT012253 is a PET radioligand for mGlu4 (3.4 nM), previously characterized in non-human primates. We aimed to determine the optimal method for quantification, duration for acquisition, and test-retest reliability of the binding parameters for [(11)C]PXT012253 in healthy volunteers. RESULTS: Six subjects (4 females) completed. [(11)C]PXT012253 displayed high uptake and rapid wash-out. Unchanged [(11)C]PXT012253 at 20 min was 10–20%. V(T) in subcortical regions was higher than in cortical regions. 2TC provided better fits than 1TC. V(T) by Logan GA and MA1 analysis correlated with that of 2TC-CM. MA1 showed better identifiability and standard error than Logan. The test-retest metrics in pons, putamen and thalamus showed absolute variability of V(T)<7% and ICC > 0.93 using the 2TC, Logan and MA1 graphical analyses. Time stability analysis showed that V(T) values estimated using 63 min of imaging were within 10% of the values obtained with 93 min with all three models. CONCLUSION: [(11)C]PXT012253 showed a high brain uptake, with rapid washout and metabolism. V(T) was reliably estimated using 2TC, Logan GA and MA1. The test-retest metrics showed high repeatability, indicating [(11)C]PXT012253 to be a suitable PET radioligand for mGlu4. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s13550-025-01266-y.