Cholinergic patterns correlate with dopamine medication ON freezing of gait in Parkinson's disease

Gait problems in people with Parkinson's disease are increasingly common as disease progresses. Symptoms include freezing of gait (FoG) and a predisposition to falls. The causative pathophysiology is not completely understood. In this study, PET with 18F-fluoroethoxybenzovesamicol (18F-FEOBV), a presynaptic marker of cholinergic terminal density, and 18F-fluorodeoxyglucose (18F-FDG) was performed in a cohort of people with Parkinson's disease and gait disorder to derive spatial covariance networks of cholinergic and metabolic activity and to evaluate the correlation of such networks against the frequency of FoG and other gait measures. Fourteen patients with Parkinson's disease and FoG in the ON motor state underwent PET using 18F-FEOBV and 18F-FDG on two separated days. Following spatial normalization, functional networks were derived by principal component analysis. The individual expression of linear combinations of principal components was subsequently correlated with measures of FoG in the ON motor state (ON-FoG) and a lower body and gait subsection of the Unified Parkinson's Disease Rating Scale part III. Gait measures were derived from home-worn measures using a triaxial accelerometer. We found a derived pattern of 18F-FEOBV binding that was correlated with ON-FoG (R2 = 0.46975, P = 0.045) and with other lower body and gait signs (R2 = 0.78591, P = 0.0077). Lower levels of cholinergic activity in the thalamus, hippocampus, striatum, anterior cingulate and areas of the brainstem consistent with the mesencephalic locomotor region were associated with worse ON-FoG and gait disturbances. The derived pattern was not associated with overall disease duration or progression as assessed by standard motor scores. There was no correlation between 18F-FEOBV and OFF-FoG. For 18F-FDG, no correlation between covariance patterns and gait assessments could be found. However, a statistically significant correlation was found for a subset of lower body and gait symptoms (R2 = 0.78306, P = 0.002). These results exhibit a correlation between lower levels of cholinergic function in locomotor-related areas of the brainstem and objective measures of dopamine medication ON-FoG, potentially indicating a causative link between the two. No association was found with OFF-FoG. Taken together, our results provide support for the role of the cholinergic system in the occurrence of dopamine medication ON-FoG.