Phase-dependent transcranial magnetic stimulation of the lesioned hemisphere is accurate after stroke

Abstract

Transcranial magnetic stimulation (TMS) can produce plastic changes within descending motor pathways and distributed brain networks [[1]Ziemann U. Paulus W. Nitsche M.A. Pascual-Leone A. Byblow W.D. Berardelli A. et al.Consensus: motor cortex plasticity protocols.Brain Stimulation. 2008; 1: 164-182Abstract Full Text Full Text PDF PubMed Scopus (421) Google Scholar,[2]Wang J.X. Rogers L.M. Gross E.Z. Ryals A.J. Dokucu M.E. Brandstatt K.L. et al.Targeted enhancement of cortico-hippocampal brain networks and associative memory.Science. 2014; 435: 1054-1057Crossref Scopus (268) Google Scholar]. It has been proposed that TMS could enhance post-stroke motor recovery by normalizing imbalanced sensorimotor network function and/or upregulating corticospinal output [[3]Hummel F.C. Cohen L.G. Non-invasive brain stimulation: a new strategy to improve neurorehabilitation after stroke?.Lancet Neurol. 2006; 5: 708-712Abstract Full Text Full Text PDF PubMed Scopus (582) Google Scholar,[4]Di Lazzaro V. Dileone M. Profice P. Pilato F. Cioni B. Meglio M. et al.Direct demonstration that repetitive transcranial magnetic stimulation can enhance corticospinal excitability in stroke. vol. 37. 2006: 2850-2853Google Scholar] but studies using TMS to boost motor recovery have shown heterogeneous results [[5]Smith M.C. Stinear C.M. Transcranial magnetic stimulation (TMS) in stroke: Ready for clinical practice?.J Clin Neurosci. 2016; 31: 10-14Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar]. However, TMS has traditionally been delivered uncoupled from endogenous brain oscillatory activity, leading to indiscriminate application of individual TMS pulses across different, physiologically distinct brain states. Thus, failure to control for endogenous brain states during TMS application may have contributed to the overall weak effect sizes and high response variability often observed in TMS studies [[6]López-Alonso V. Cheeran B. Río-Rodríguez D. Fernández-del-Olmo M. Inter-individual variability in response to non-invasive brain stimulation paradigms.Brain Stimul. 2014; 7: 372-380Abstract Full Text Full Text PDF PubMed Scopus (418) Google Scholar]. Phase-dependent TMS, which involves delivering individual TMS pulses or trains of pulses during pre-defined brain oscillatory phases, attempts to address this limitation. Early results using phase-dependent TMS in healthy individuals are promising: TMS applied during sensorimotor mu (8–12 Hz) trough phases reflecting increased sensorimotor cortical neuronal spiking [[7]Haegens S. Nácher V. Luna R. Romo R. Jensen O. α-oscillations in the monkey sensorimotor network influence discrimination performance by rhythmical inhibition of neuronal spiking.Proc Natl Acad Sci. 2011; 108: 19377-19382Crossref PubMed Scopus (395) Google Scholar] and inter-regional neuronal communication [[8]Stefanou M.I. Desideri D. Belardinelli P. Zrenner C. Ziemann U. Phase synchronicity of μ-rhythm determines efficacy of interhemispheric communication between human motor cortices.J Neurosci. 2018; 38: 10525-10534Crossref PubMed Scopus (21) Google Scholar] enhances corticospinal output to a larger extent than TMS applied irrespective of these phases [[9]Zrenner C. Desideri D. Belardinelli P. Ziemann U. Real-time EEG-defined excitability states determine efficacy of TMS-induced plasticity in human motor cortex.Brain Stimul. 2018; 11: 374-389Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar]. These findings raise the hypothesis that phase-dependent TMS could be more effective than TMS uncoupled from sensorimotor mu phases. Yet, for phase-dependent TMS to be therapeutically useful after stroke, it must first be possible to accurately deliver TMS during pre-defined brain oscillatory phases in the lesioned brain. Why might accurate phase-dependent TMS delivery be challenging after stroke? In order to account for time-delays inherent to signal acquisition and processing, phase-dependent TMS approaches typically use autoregressive forward prediction in either the time or frequency domain to estimate the instantaneous oscillatory phase at some future time point [[9]Zrenner C. Desideri D. Belardinelli P. Ziemann U. Real-time EEG-defined excitability states determine efficacy of TMS-induced plasticity in human motor cortex.Brain Stimul. 2018; 11: 374-389Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar,[10]Madsen K.H. Karabanov A.N. Krohne L.G. Safeldt M.G. Tomasevic L. Siebner H.R. No trace of phase: Corticomotor excitability is not tuned by phase of pericentral mu-rhythm.Brain Stimul. 2019; 12: 1261-1270Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar]. These methods require the presence of a recordable mu rhythm that exhibits consistent and predictable phase progression over time. It is known that brain networks reorganize after cortical and subcortical stroke [[11]Grefkes C. Ward N.S. Cortical reorganization after stroke: how much and how functional?.The Neuroscientist. 2014; 20: 56-70Crossref PubMed Scopus (190) Google Scholar]. Both the initial lesion and this subsequent reorganization could weaken or eliminate the mu oscillation by disrupting its cortical sources [[12]Salmelin R. Hari R. Spatiotemporal characteristics of sensorimotor neuromagnetic rhythms related to thumb movement.Neuroscience. 1994; 60: 537-550Crossref PubMed Scopus (590) Google Scholar], their pacemaker cells [[13]Hughes S.W. Crunelli V. Thalamic mechanisms of EEG alpha rhythms and their pathological implications.The Neuroscientist. 2005; 11: 357-372Crossref PubMed Scopus (324) Google Scholar], or connections between them. Even if a mu oscillation is present after stroke, it may exhibit highly variable phase progression over time, preventing reliable phase targeting. In light of these potential issues, the accuracy of phase-dependent TMS in the lesioned brain has not been established. Here, we sought to address this gap in knowledge by examining the accuracy of delivering phase-dependent TMS to the lesioned hemisphere after stroke. Specifically, we evaluated the ability of a real-time EEG analysis algorithm commonly used during phase-dependent TMS [[9]Zrenner C. Desideri D. Belardinelli P. Ziemann U. Real-time EEG-defined excitability states determine efficacy of TMS-induced plasticity in human motor cortex.Brain Stimul. 2018; 11: 374-389Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar] to reliably target sensorimotor mu oscillation peak and trough phases in chronic stroke patients with persistent upper limb motor deficits. The study involved three sessions, including (1) a complete neurological examination, (2) structural magnetic resonance imaging, and (3) an experimental session during which the ability to deliver phase-dependent TMS was evaluated by examining the accuracy of phase targeting using real-time EEG analysis. Eight chronic stroke patients were screened for inclusion in this study, and 3 met all eligibility criteria and completed all procedures (Patients A, B and C). Patient A was a 74 y/o female with a large left hemispheric cortical lesion and a Fugl-Meyer Assessment Upper Extremity (FMA-UE) score of 33/66. Patient B was a 57 y/o male with a right subcortical lesion and an FMA-UE score of 59/66. Patient C was a 62 y/o female with a left subcortical lesion and an FMA-UE score of 60/66. Motor-evoked potentials (MEPs) could be reliably elicited in the first dorsal interosseous muscle of the affected hand in each patient. Accuracy of phase-dependent TMS was quantified during a resting, eyes-open EEG recording during which no TMS pulses were delivered [[9]Zrenner C. Desideri D. Belardinelli P. Ziemann U. Real-time EEG-defined excitability states determine efficacy of TMS-induced plasticity in human motor cortex.Brain Stimul. 2018; 11: 374-389Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar,[10]Madsen K.H. Karabanov A.N. Krohne L.G. Safeldt M.G. Tomasevic L. Siebner H.R. No trace of phase: Corticomotor excitability is not tuned by phase of pericentral mu-rhythm.Brain Stimul. 2019; 12: 1261-1270Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar]. See Supplementary Information for a detailed description of eligibility criteria, experimental procedures, patient characteristics and anatomical MR images. Each of the three patients exhibited a clear mu oscillation, reflected as periodic 8–12 Hz activity that exceeded the aperiodic 8–12 Hz component of EEG signals recorded over the lesioned hemisphere (p < 0.001 for all patients; Fig. 1a). Further, the real-time EEG analysis algorithm accurately identified and targeted mu oscillatory peak and trough phases in each patient (see Fig. 1b for filtered and raw time-series data). Phase angle distributions at the time of peak and trough targeting significantly deviated from uniformity in the 90° and 270° directions (p < 0.001 and V-statistic > 43.81 for all patients) and also from each other (p < 0.001 and F > 208.50 for all patients; Fig. 1c–d). As expected, phase angle distributions at the time of random phase targeting did not deviate from uniformity in any direction (p > 0.05 and M-value > 35 for all patients). These findings demonstrate for the first time that sensorimotor mu oscillation phase-dependent TMS can be accurately delivered to the lesioned hemisphere after stroke. The patients tested here showed variability in lesion location and motor impairment. Patients B and C had subcortical lesions and mild motor deficits, while Patient A had a very large cortical lesion and moderate to severe deficits. Because Patient A’s lesion produced large brain volume loss in the regions thought to generate the mu rhythm (primary somatosensory cortex [[12]Salmelin R. Hari R. Spatiotemporal characteristics of sensorimotor neuromagnetic rhythms related to thumb movement.Neuroscience. 1994; 60: 537-550Crossref PubMed Scopus (590) Google Scholar]), it is surprising that we could both measure a mu oscillation and accurately target mu phases in this patient. Based on the size and extent of the lesion and the characteristics of this patient’s power spectra (see Fig. 1a, left panel), it is possible that the recorded oscillation originated from a non-motor cortical or subcortical brain region and was shunted through high-conductivity cerebrospinal fluid within the lesion space [[14]Baumann S.B. Wozny D.R. Kelly S.K. Meno F.M. The electrical conductivity of human cerebrospinal fluid at body temperature.IEEE Trans Biomed Eng. 1997; 44: 220-223Crossref PubMed Scopus (332) Google Scholar]. Future studies delivering mu phase-dependent TMS to the lesioned brain would benefit from individualizing filtering procedures used during phase targeting to more precisely isolate the mu oscillation in both the frequency and spatial domain. Given that sensorimotor rhythms are strongest at rest and desynchronize with movement [[15]Pfurtscheller G. Lopes da Silva F.H. Event-related EEG/MEG synchronization and desynchronization: basic principles.Clin Neurophysiol. 1999; 110: 1842-1857Crossref PubMed Scopus (4587) Google Scholar], individual filter parameters could be constructed using a functional localization task requiring patients to alternate between (attempted) movement and rest. In conclusion, we report that TMS can be accurately delivered during pre-defined brain oscillatory phases in the lesioned brain. Combined with recent work documenting the advantages of phase-dependent TMS in healthy individuals [[9]Zrenner C. Desideri D. Belardinelli P. Ziemann U. Real-time EEG-defined excitability states determine efficacy of TMS-induced plasticity in human motor cortex.Brain Stimul. 2018; 11: 374-389Abstract Full Text Full Text PDF PubMed Scopus (119) Google Scholar], these results pave the way for future phase-dependent TMS studies aimed at enhancing sensorimotor function after stroke. SJH: conceptualization, data curation, formal analysis, investigation, methodology, project administration, software, validation, visualization, writing (original draft). WH: data curation, formal analysis, investigation. FF: data curation, investigation, project administration. CZ: methodology, software. UZ: methodology, software, writing (review and editing). ERB: methodology, software, writing (review and editing). MKH: data curation, investigation, project administration, supervision. LGC: conceptualization, funding acquisition, resources, supervision, writing (review and editing). UZ has received grants from European Research Council ( ERC ), German Research Foundation ( DFG ), German Federal Ministry of Education and Research ( BMBF ), Bristol Myers Squibb , Janssen Pharmaceutica NV, Servier , Biogen Idec GmbH, and personal fees from Bayer Vital GmbH , Pfizer GmbH, CorTec GmbH, all not related to this work. CZ is coordinator of and partially funded through an EXIST Transfer of Research grant by the German Federal Ministry for Economic Affairs and Energy (grant 03EFJBW169 ). The goal of this grant is the commercialization of a real-time EEG analysis device through a spin-off start-up to enable therapeutic brain-oscillation synchronized stimulation. This work was supported by the Intramural Research Program of the National Institute of Neurological Disorders and Stroke . This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health , under Contract No. HHSN261200800001E and No. 75N91019D00024 , Task Order No. 75N91019F00129. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. S.J.H. is supported by an NINDS Intramural Competitive Fellowship.

Topics & Concepts