EEG-based biomarkers for optimizing deep brain stimulation contact configuration in Parkinson's disease.

Objective: Subthalamic deep brain stimulation (STN-DBS) is a neurosurgical therapy to treat Parkinson's disease (PD). Optimal therapeutic outcomes are not achieved in all patients due to increased DBS technological complexity; programming time constraints; and delayed clinical response of some symptoms. To streamline the programming process, biomarkers could be used to accurately predict the most effective stimulation configuration. Therefore, we investigated if DBS-evoked potentials (EPs) combined with imaging to perform prediction analyses could predict the best contact configuration. Methods: In 10 patients, EPs were recorded in response to stimulation at 10 Hz for 50 s on each DBS-contact. In two patients, we recorded from both hemispheres, resulting in recordings from a total of 12 hemispheres. A monopolar review was performed by stimulating on each contact and measuring the therapeutic window. CT and MRI data were collected. Prediction models were created to assess how well the EPs and imaging could predict the best contact configuration. Results: EPs at 3 ms and at 10 ms were recorded. The prediction models showed that EPs can be combined with imaging data to predict the best contact configuration and hence, significantly outperformed random contact selection during a monopolar review. Conclusion: EPs can predict the best contact configuration. Ultimately, these prediction tools could be implemented into daily practice to ease the DBS programming of PD patients.

Towards biomarker-based optimization of deep brain stimulation in Parkinson's disease patients.

Background: Subthalamic deep brain stimulation (DBS) is an established therapy to treat Parkinson's disease (PD). To maximize therapeutic outcome, optimal DBS settings must be carefully selected for each patient. Unfortunately, this is not always achieved because of: (1) increased technological complexity of DBS devices, (2) time restraints, or lack of expertise, and (3) delayed therapeutic response of some symptoms. Biomarkers to accurately predict the most effective stimulation settings for each patient could streamline this process and improve DBS outcomes. Objective: To investigate the use of evoked potentials (EPs) to predict clinical outcomes in PD patients with DBS. Methods: In ten patients (12 hemispheres), a monopolar review was performed by systematically stimulating on each DBS contact and measuring the therapeutic window. Standard imaging data were collected. EEG-based EPs were then recorded in response to stimulation at 10 Hz for 50 s on each DBS-contact. Linear mixed models were used to assess how well both EPs and image-derived information predicted the clinical data. Results: Evoked potential peaks at 3 ms (P3) and at 10 ms (P10) were observed in nine and eleven hemispheres, respectively. Clinical data were well predicted using either P3 or P10. A separate model showed that the image-derived information also predicted clinical data with similar accuracy. Combining both EPs and image-derived information in one model yielded the highest predictive value. Conclusion: Evoked potentials can accurately predict clinical DBS responses. Combining EPs with imaging data further improves this prediction. Future refinement of this approach may streamline DBS programming, thereby improving therapeutic outcomes. Clinical trial registration: ClinicalTrials.gov, identifier NCT04658641.

Dopamine substitution alters effective connectivity of cortical prefrontal, premotor, and motor regions during complex bimanual finger movements in Parkinson's disease.

Bimanual coordination is impaired in Parkinson's disease (PD), affecting patients' quality of life. Besides dysfunction of the basal ganglia network, alterations of cortical oscillatory coupling, particularly between prefrontal and (pre-)motoric areas, are thought to underlie this impairment. Here, we studied 16 PD patients OFF and ON medication and age-matched healthy controls recording high-resolution electroencephalography (EEG) during performance of spatially coupled and uncoupled bimanual finger movements. Dynamic causal modeling (DCM) for induced responses was used to infer task-induced effective connectivity within a network comprising bilateral prefrontal cortex (PFC), lateral premotor cortex (lPM), supplementary motor area (SMA), and primary motor cortex (M1). Performing spatially coupled movements, excitatory left-hemispheric PFC to lPM coupling was significantly stronger in controls compared to unmedicated PD patients. Levodopa-induced enhancement of this connection correlated with increased movement accuracy. During performance of spatially uncoupled movements, PD patients OFF medication exhibited inhibitory connectivity from left PFC to SMA. Levodopa intake diminished these inhibitory influences and restored excitatory PFC to lPM coupling. This restoration, however, did not improve motor function. Concluding, our results indicate that lateralization of prefrontal to premotor connectivity in PD can be augmented by levodopa substitution and is of compensatory nature up to a certain extent of complexity.

Ageing changes effective connectivity of motor networks during bimanual finger coordination.

Bimanual finger coordination declines with age. However, relatively little is known about the neurophysiological alterations in the motor-system causing this decline. In the present study, we used 128-channel electroencephalography (EEG) to evaluate causal interactions of cortical, motor-related brain areas. Right-handed young and elderly subjects performed complex temporally and spatially coupled as well as temporally coupled and spatially uncoupled finger tappings. Employing dynamic causal modelling (DCM) for induced responses, we inferred task-induced effective connectivity within a core motor network comprising bilateral primary motor cortex (M1), lateral premotor cortex (lPM), supplementary motor area (SMA), and prefrontal cortex (PFC). Behavioural analysis showed significantly increased error rates and performance times for elderly subjects, confirming that motor functions decrease with ageing. Additionally, DCM analysis revealed that this age-related decline can be associated with specific alterations of interhemispheric and prefrontal to premotor connectivity. Young and elderly subjects exhibited inhibitory left to right M1-M1 coupling during performance of temporally and spatially coupled movements. Effects of ageing on interhemispheric connectivity particularly emerged when movements became spatially uncoupled. Here, elderly participants still expressed inhibitory left to right M1-M1 coupling, whereas no such connection was present in the young. Furthermore, ageing affected prefrontal to premotor connectivity. In all conditions, elderly subjects showed significant couplings from left PFC to left lPM. In contrast, young participants exhibited left PFC to SMA connections. These results demonstrate that (i) in spatially uncoupled movements interhemispheric M1-connectivity increases with age and (ii) support the idea that ageing is associated with enhanced lateral prefrontal to premotor coupling (PFC to lPM) and hypoactivation of a medial pathway (PFC to SMA) within the dominant hemisphere.