MRI volumetric and intensity analysis of the cerebellum in Parkinson's disease patients infused with glial-derived neurotrophic factor (GDNF)

Background: A recent human therapeutic trial using intraputaminal infusion of glial cell-derived neurotrophic factor (GDNF) in Parkinson's disease (PD) was abruptly terminated, partly due to safety concerns raised by the finding of cerebellar lesions in monkeys given high-dose GDNF. Methods: Magnetic resonance images from nine PD patients participating in this trial were analyzed to determine whether subtle volumetric or intensity changes could be detected in the cerebellum or elsewhere following GDNF treatment for over 1 year. Subtraction images were compared to a reference standard deviation map constructed by using identically-processed paired scans from 25 normal adults. In a separate voxel-based group morphometric (VBM) analysis of the same patient images, grey matter intensity was compared between pre and post-GDNF infusion scans using a repeated measures ANOVA with family-wise error threshold of P = 0.10. Two expert readers independently reviewed serial FLAIR images from all patients. Results: (1) There were no significant cerebellar differences in any of the nine individual PD patients (difference image analysis), (2) there were no significant morphometric differences between pre- and post-GDNF scans (VBM), and (3) there were no signal abnormalities in the cerebellum detected on the FLAIR images in PD patients (clinical scan review). Conclusions: In concert with lack of evidence of cerebellar dysfunction on clinical examination, we find no imaging evidence of cerebellar injury in human subjects undergoing chronic intracerebral GDNF infusion.