Spinal Cord Stimulation for Freezing of Gait: From Bench to Bedside

Data from a survey of 6,620 Parkinson's disease patients were examined for correlation of freezing with age, sex, duration, subjective severity of Parkinson's disease, and antiparkinsonian medication. Forty-seven percent of the patients reported experiencing freezing regularly. Logistic regression analysis showed that freezing was significantly associated with a longer disease duration and a more advanced stage of the disease. Freezing episodes were more likely in men than in women and in patients taking, in addition to levodopa, Entacapone, Amantadine, or dopamine agonists. Finally, patients considering tremor as their main symptom reported freezing less frequently. Common antiparkinsonian drugs given in combination with levodopa were not negatively correlated with freezing. The results underline the necessity to develop appropriate countermeasures against this phenomenon, which is widely known to cause significant impairment of patients' quality of life and - as our data also showed - may cause traffic accidents in licensed patients. (c) 2007 Movement Disorder Society.

Recent neuroimaging evidence has led to the proposal that freezing of gait in Parkinson's disease is due to dysfunctional interactions between frontoparietal cortical regions and subcortical structures, such as the striatum. However, to date, no study has employed task-based functional connectivity analyses to explore this hypothesis. In this study, we used a data-driven multivariate approach to explore the impaired communication between distributed neuronal networks in 10 patients with Parkinson's disease and freezing of gait, and 10 matched patients with no clinical history of freezing behaviour. Patients performed a virtual reality gait task on two separate occasions (once ON and once OFF their regular dopaminergic medication) while functional magnetic resonance imaging data were collected. Group-level independent component analysis was used to extract the subject-specific time courses associated with five well-known neuronal networks: the motor network, the right- and left cognitive control networks, the ventral attention network and the basal ganglia network. We subsequently analysed both the activation and connectivity of these neuronal networks between the two groups with respect to dopaminergic state and cognitive load while performing the virtual reality gait task. During task performance, all patients used the left cognitive control network and the ventral attention network and in addition, showed increased connectivity between the bilateral cognitive control networks. However, patients with freezing demonstrated functional decoupling between the basal ganglia network and the cognitive control network in each hemisphere. This decoupling was also associated with paroxysmal motor arrests. These results support the hypothesis that freezing behaviour in Parkinson's disease is because of impaired communication between complimentary yet competing neural networks.

Freezing of gait (FOG) has been identified as one of the main contributors to gait disturbances in Parkinson's disease. While the pathophysiology remains enigmatic, several factors such as step length and the sequence effect (step to step reduction in amplitude) may lead to the occurrence of FOG. It was hypothesized that by reducing step length, FOG episodes would present more frequently if a significant sequence effect (measured as a regression slope) was co-existent in the subject. Twenty-six participants with Parkinson's disease were separated clinically into a freezing (PD + FOG, n = 16) and non-freezing (PD-FOG, n = 10) group, with 10 age-matched control participants. Testing involved walking trials where preferred step length was set at 100%, 75%, 50% and 25% of normalized step length. The number of FOG episodes increased in the 50% condition and further increased in the 25% condition compared to other conditions. The participants with FOG also demonstrated a larger average regression slope, with significant differences in the 75%, 50% and 25% conditions when compared to the PD-FOG and control groups. There were no significant differences when comparing the slope of the PD-FOG and control group, indicating the reduced step length and the sequence effect may have led to the occurrence of FOG. These findings support the possible dual requirement of a reduced step length and a successive step to step amplitude reduction to lead to FOG.