Reduction of spontaneous cortical beta bursts in Parkinson’s disease is linked to symptom severity

Abstract

Parkinson’s disease is characterized by a gradual loss of dopaminergic neurons, which is associated with altered neuronal activity in the beta-band (13–30 Hz). Assessing beta-band activity typically involves transforming the time-series to get the power of the signal in the frequency domain. Such transformation assumes that the time-series can be reduced to a combination of steady-state sine- and cosine waves. However, recent studies have suggested that this approach masks relevant biophysical features in the beta-band—for example, that the beta-band exhibits transient bursts of high-amplitude activity. In an exploratory study, we used magnetoencephalography to record beta-band activity from the sensorimotor cortex, to characterize how spontaneous cortical beta bursts manifest in Parkinson’s patients on and off dopaminergic medication, and compare this to matched healthy controls. We extracted the time-course of beta-band activity from the sensorimotor cortex and characterized bursts in the signal. We then compared the burst rate, duration, inter-burst interval and peak amplitude between the Parkinson’s patients and healthy controls. Our results show that Parkinson’s patients off medication had a 5–17% lower beta bursts rate compared to healthy controls, while both the duration and the amplitude of the bursts were the same for healthy controls and medicated state of the Parkinson’s patients. These data thus support the view that beta bursts are fundamental underlying features of beta-band activity, and show that changes in cortical beta-band power in Parkinson’s disease can be explained—primarily by changes in the underlying burst rate. Importantly, our results also revealed a relationship between beta burst rate and motor symptom severity in Parkinson’s disease: a lower burst rate scaled with increased severity of bradykinesia and postural/kinetic tremor. Beta burst rate might thus serve as a neuromarker for Parkinson’s disease that can help in the assessment of symptom severity in Parkinson’s disease or in the evaluation of treatment effectiveness.

Individual Magnetoencephalography Response Profiles to Short-Duration L-Dopa in Parkinson’s Disease

“…Additionally, the present analysis included both hemispheres to compare predictions against the largely bilateral set of clinical factor scores considered, thus neglecting potential relationships between the lateralized clinical factors of upper extremities and the laterality of MEG in PD (Heinrichs-Graham et al, 2017). Finally, while our recording duration (∼3 min) was consistent with prior resting-state spectral analyses (Pollok et al, 2012;Hall et al, 2014;Cao et al, 2020;Vinding et al, 2020), other relevant metrics, such as resting-state canonical brain networks, may require longer recording durations (Liuzzi et al, 2017).…”

“…Notably, motor factor scores did not correlate with each other, and this independence amongst scores challenges the use of summated MDS-UPDRS III ratings as a primary clinical outcome measure, as motor factor scores may provide distinct clinical information about therapeutic response. While our PD cohort was of comparable size to previous MEG studies ( Heinrichs-Graham et al, 2014 ; Cao et al, 2020 ; Vinding et al, 2020 ), future studies that leverage the factor score methodology in larger cohorts will be crucial to expand the needed quantitative information of therapeutic response in SDR, long-duration L-Dopa response ( Anderson and Nutt, 2011 ; Cilia et al, 2020 ), and other interventional therapies.…”

“…Notably, motor factor scores did not correlate with each other, and this independence amongst scores challenges the use of summated MDS-UPDRS III ratings as a primary clinical outcome measure, as motor factor scores may provide distinct clinical information about therapeutic response. While our PD cohort was of comparable size to previous MEG studies (Heinrichs-Graham et al, 2014;Cao et al, 2020;Vinding et al, 2020), future studies that leverage the factor score methodology in FIGURE 2 | Individual profiles of MEG responses in terms of frequencies that best distinguished between LEVODOPA-OFF and LEVODOPA-ON within subjects. (A) Accuracy of within-subject linear support vector machine (SVM) in distinguishing between LEVODOPA-OFF and LEVODOPA-ON with 10-fold cross validation.…”

“…However, similar to clinical symptoms, non-invasive neuroimaging metrics also differ across individuals ( Finn et al, 2015 ; Yang et al, 2017 ) and the therapeutic implications of such inter-subject variability in PD are currently not well characterized ( Marras et al, 2020 ). Previous studies using non-invasive magnetoencephalography (MEG) in PD to image motor cortex reported SDR effects in a variety of neuroimaging metrics such as spectral power ( Heinrichs-Graham et al, 2014 ; Cao et al, 2020 ; Vinding et al, 2020 ), cortico-muscular coherence ( Salenius et al, 2002 ; Hirschmann et al, 2013 ), and interhemispheric coherence ( Heinrichs-Graham et al, 2014 ). However, conflicting findings reflect a lack of robustness of any single metric when averaged across multiple subjects ( Boon et al, 2019 ; Vinding et al, 2020 ), and the role of inter-subject variability remains unclear.…”

“…Previous studies using non-invasive magnetoencephalography (MEG) in PD to image motor cortex reported SDR effects in a variety of neuroimaging metrics such as spectral power ( Heinrichs-Graham et al, 2014 ; Cao et al, 2020 ; Vinding et al, 2020 ), cortico-muscular coherence ( Salenius et al, 2002 ; Hirschmann et al, 2013 ), and interhemispheric coherence ( Heinrichs-Graham et al, 2014 ). However, conflicting findings reflect a lack of robustness of any single metric when averaged across multiple subjects ( Boon et al, 2019 ; Vinding et al, 2020 ), and the role of inter-subject variability remains unclear. Novel features may help account for outcome differences across individuals, as highlighted in recent MEG studies of healthy subjects using novel whole-brain modeling and feature extraction techniques ( Vidaurre et al, 2018 ; Becker et al, 2020 ; Hirschmann et al, 2020 ).…”