Structural or Functional Metrics to Assess Motor Impairment in Chronic Stroke?

HomeStrokeVol. 52, No. 1Structural or Functional Metrics to Assess Motor Impairment in Chronic Stroke? Free AccessEditorialPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessEditorialPDF/EPUBStructural or Functional Metrics to Assess Motor Impairment in Chronic Stroke? Assia Jaillard Assia JaillardAssia Jaillard Correspondence to: Assia Jaillard, MD, PhD, SFR1 RMN Biomédicale et Neurosciences, Unité IRM 3 Tesla Recherche, CHU Grenoble Alpes, BP 217 - 38043 Grenoble Cedex 9 France. Email E-mail Address: [email protected] IRMaGe, Inserm-US17 CNRS-UMS3552 UGA CHU Grenoble Alpes (CHUGA), Grenoble, France (A.J.). Université Grenoble Alpes, AGEIS, EA 7407, France (A.J.). Pôle Recherche, CHUGA, France (A.J.). Search for more papers by this author Originally published15 Dec 2020https://doi.org/10.1161/STROKEAHA.120.032992Stroke. 2021;52:250–252This article is a commentary on the followingMultimodal Assessment of the Motor System in Patients With Chronic Ischemic StrokeOther version(s) of this articleYou are viewing the most recent version of this article. Previous versions: December 15, 2020: Ahead of Print Motor deficits following stroke are a leading cause of long-term disability in adults. Although recent advances in the management of acute ischemic stroke have improved patients’ outcome, few treatments are available at the chronic period of stroke, neurorehabilitation remaining the cornerstone of motor impairment treatment.1 There is evidence suggesting that chronic motor impairment is correlated with structural and functional metrics provided by a set of neurophysiological and neuroimaging techniques2–5 that may help clinicians to improve neurorehabilitation impact by attacking the problem of chronic stroke treatment from an original angle. Two types of structural metrics are used. Macrostructural metrics are computed using lesion or damaged corticospinal tract (CST) volumes obtained with morphological computed tomography or magnetic resonance imaging (MRI) images, whereas microstructural metrics, such as fractional anisotropy (FA), are derived from diffusion MRI assessing white matter tracts. The CST, also labeled pyramidal tract since CST fibers cross at the level of the medullar pyramids, is the main motor descending pathway. The CST is the usual target of studies aimed at identifying structural metrics of upper limb motor function and outcome.4 Functional measures can be obtained with several techniques including transcranial magnetic stimulation (TMS), EEG, magnetoencephalography, near infrared spectroscopy, and functional MRI. For instance, TMS, if applied over the head region corresponding to the hand area of the primary motor cortex, can elicit motor evoked potentials (MEPs) in target muscles of the contralateral upper limb. TMS is used to evaluate the excitability in the ipsilesional (damaged) and contralesional (nondamaged) motor cortices using several measures including the resting motor threshold.6,7See related article, p 241There is a debate among the constellation of neurophysiological and imaging approaches on the metrics that may serve as surrogates of upper limb motor function.2,3,8A first question is whether a single measure may predict upper limb motor function, fitting with the one size fits all concept? For example, FA asymmetry measured in the internal capsule of the ipsilesional and contralesional CST appears to be a reliable metric of upper limb motor function explaining around half of its variance.4,9 TMS studies have also shown that MEP parameters correlated with upper limb motor function at the time of testing, providing potential neurophysiological biomarkers, although some limitations might be considered due to methodological heterogeneity between the studies.6,7,10 Conversely, multimodal approaches based on the combination of structural and functional techniques may improve prediction accuracy.3,5,8An alternative perspective proposed by Nazarova et al11 in a report in this issue of Stroke was to develop more sophisticated models at the metric level to evaluate whether combined measures derived from one single modality improve the prediction of upper limb motor performances in small samples. This study of 35 patients with unilateral hemispheric chronic stroke who underwent clinical motor tests as well as diffusion MRI and TMS assessment of the motor system tested if structural and functional multimodal metrics would provide complementary or redundant information on upper limb motor outcome. The authors computed FA measures from the CST and corpus callosum, and TMS-derived parameters such as MEPs and resting motor thresholds recorded in 2 distinct hand muscles (instead of one) to classify the patients into 3 categories assessed as good, moderate, and bad clinical motor outcome. FA metrics included CST FA asymmetry and the Fréchet distance between the ipsilesional (damaged) and contralesional (nondamaged) CST FA profiles. The Fréchet distance represents a measure of similarity between 2 curves accounting for the location and order of the points along these curves (ie, the CST FA profiles), so the greater the distance, the greater the CST damage.The results of this study showed that either FA CST asymmetry in the internal capsule (<0.75) or MEP (absence or presence) were able to differentiate bad from good and moderate paresis but not good from moderate groups. In contrast, the Fréchet distance differed significantly between the three groups. Moreover, a correct classification for each of the 3 groups was only obtained by a metric combining FA asymmetry and Fréchet distance between ipsilesional and contralesional CSTs. Regarding TMS measures, there was a trend between good and moderate outcome when MEPs could be elicited in 2 muscles rather than one and for resting motor thresholds in patients with elicited MEPs. Of note, the corpus callosum FA did not lead to a correct classification of the three clinical groups. However, FA was decreased in the group with bad outcome and in the patients with no elicited MEPs.Overall, Nazarova et al11 showed that simple measures derived from diffusion MRI or TMS are insufficient for a correct outcome classification, in line with some previous reports.3 Furthermore, their findings suggest that the combination of several CST FA measures including the Fréchet distance may provide a better proxy of upper limb motor function. Although these results need to be replicated in larger samples, especially regarding the Fréchet distance, the relevance of microstructural metrics as potential biomarkers can be paralleled with other reports suggesting that diffusion MRI measures from a set of white matter tracts could predict walk recovery.12,13 Finally, Nazarova et al11 showed no benefit of a functional and structural multimodal approach. This may appear counter-intuitive according to the view the more, the better, as multimodal approaches are supposed to provide complementary information.5 Yet, another study comparing several structural and functional approaches has reported that a TMS parameter (resting motor thresholds) was the only predictor of motor outcome.10How practice may change knowing the findings reported in Nazarova et al,11 and what are the issues that remain to be addressed? Above all, the mechanisms of motor recovery need to be better investigated to reduce the fraction of variance unexplained by the current models and evaluate new therapeutics in postacute stroke. Yet, we face contradictory aims. On one hand, functional and structural multimodal neuroimaging with an extensive clinical motor assessment associated with biological patient data appear to be the way to explore brain mechanisms, but such an approach is expansive, time-consuming, and hardly simple. On the other hand, we need standardized biomarkers that can be implemented in clinical practice to select treatment responders and guide individual patient management.A way to reconcile these 2 approaches in practice would be to include a diffusion sequence in the routine MRI protocol performed in patients with subacute and chronic stroke. This would increase the sample sizes, allowing for the validation of microstructural biomarkers among patients assessed at different periods of stroke, with various demographics, biological factors, and clinical outcomes. Indeed, the duration of a diffusion sequence with at least 30 directions is typically 6 minutes, which is the same as 3-dimensional morphological routine sequences. In contrast, the TMS protocol described by Nazarova et al,11 lasting at least 1 hour, cannot be administered as a routine protocol. A more problematic issue is that MEPs cannot be elicited in severe stroke patients who are yet the target population for novel therapeutics, whereas diffusion MRI can be acquired in almost all patients. Finally, it should be reminded that functional MRI based on task or resting sequences also offer the advantage of a relative duration (3–15 minutes) and few limitations, at least for resting state. This approach may be appropriate for exploring functional biomarkers in stroke, as evidenced by the current literature.2,14In conclusion, although diffusion imaging appears to be able to provide robust FA metrics that may help clinicians in stroke management, the reliability and contribution of microstructural and functional metrics remain to be validated in further studies.Sources of FundingNone.DisclosuresNone.FootnotesThe opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.For Disclosures, see page 251.Correspondence to: Assia Jaillard, MD, PhD, SFR1 RMN Biomédicale et Neurosciences, Unité IRM 3 Tesla Recherche, CHU Grenoble Alpes, BP 217 - 38043 Grenoble Cedex 9 France. Email assia.[email protected]frReferences1. Albert SJ, Kesselring J. Neurorehabilitation.Brainin M., Heiss W-D, eds. Textbook of Stroke Medicine. Cambridge University Press; 2010:283–306.Google Scholar2. Salvalaggio A, De Filippo De Grazia M, Zorzi M, Thiebaut de Schotten M, Corbetta M. Post-stroke deficit prediction from lesion and indirect structural and functional disconnection.Brain. 2020; 143:2173–2188. doi: 10.1093/brain/awaa156CrossrefMedlineGoogle Scholar3. Rosso C, Lamy JC. Prediction of motor recovery after stroke: being pragmatic or innovative?Curr Opin Neurol. 2020; 33:482–487. doi: 10.1097/WCO.0000000000000843Google Scholar4. Kumar P, Kathuria P, Nair P, Prasad K. Prediction of upper limb motor recovery after subacute ischemic stroke using diffusion tensor imaging: a systematic review and meta-analysis.J Stroke. 2016; 18:50–59. doi: 10.5853/jos.2015.01186Google Scholar5. Stinear CM, Barber PA, Smale PR, Coxon JP, Fleming MK, Byblow WD. Functional potential in chronic stroke patients depends on corticospinal tract integrity.Brain. 2007; 130(pt 1):170–180. doi: 10.1093/brain/awl333CrossrefMedlineGoogle Scholar6. Bembenek JP, Kurczych K, Karli Nski M, Czlonkowska A. The prognostic value of motor-evoked potentials in motor recovery and functional outcome after stroke − a systematic review of the literature.Funct Neurol. 2012; 27:79–84.MedlineGoogle Scholar7. Rosso C, Lamy JC. Does resting motor threshold predict motor hand recovery after stroke?Front Neurol. 2018; 9:1020. doi: 10.3389/fneur.2018.01020Google Scholar8. Cunningham DA, Machado A, Janini D, Varnerin N, Bonnett C, Yue G, Jones S, Lowe M, Beall E, Sakaie K, et al. Assessment of inter-hemispheric imbalance using imaging and noninvasive brain stimulation in patients with chronic stroke.Arch Phys Med Rehabil. 2015; 96(4 Suppl):S94–103. doi: 10.1016/j.apmr.2014.07.419Google Scholar9. Lindenberg R, Zhu LL, Rüber T, Schlaug G. Predicting functional motor potential in chronic stroke patients using diffusion tensor imaging.Hum Brain Mapp. 2012; 33:1040–1051. doi: 10.1002/hbm.21266CrossrefMedlineGoogle Scholar10. Kemlin C, Moulton E, Lamy JC, Houot M, Valabregue R, Leder S, Obadia MA, Meseguer E, Yger M, Brochard V, et al. Elucidating the structural and functional correlates of upper-limb poststroke motor impairment.Stroke. 2019; 50:3647–3649. doi: 10.1161/STROKEAHA.119.027126LinkGoogle Scholar11. Nazarova M, Kulikova S, Piradov MA, Limonova AS, Dobrynina LA, Konovalov RN, Novikov PA, Sehm B, Villringer A, Saltykova A, et al. Multimodal diffusion tensor imaging-transcranial magnetic stimulation assessment of the motor system in patients with chronic ischemic stroke.Stroke. 2020; 52:241–249. doi: 10.1161/STROKEAHA.119.028832Google Scholar12. Jang SH, Chang CH, Lee J, Kim CS, Seo JP, Yeo SS. Functional role of the corticoreticular pathway in chronic stroke patients.Stroke. 2013; 44:1099–1104. doi: 10.1161/STROKEAHA.111.000269LinkGoogle Scholar13. Soulard J, Huber C, Baillieul S, Thuriot A, Renard F, Aubert Broche B, Krainik A, Vuillerme N, Jaillard A; ISIS-HERMES Group. Motor tract integrity predicts walking recovery: a diffusion MRI study in subacute stroke.Neurology. 2020; 94:e583–e593. doi: 10.1212/WNL.0000000000008755Google Scholar14. Hannanu FF, Goundous I, Detante O, Naegele B, Jaillard A. Spatiotemporal patterns of sensorimotor fMRI activity influence hand motor recovery in subacute stroke: a longitudinal task-related fMRI study.Cortex. 2020; 129:80–98. doi: 10.1016/j.cortex.2020.03.024Google Scholar Previous Back to top Next FiguresReferencesRelatedDetailsCited By Lohkamp K, Kiliaan A, Shenk J, Verweij V and Wiesmann M (2021) The Impact of Voluntary Exercise on Stroke Recovery, Frontiers in Neuroscience, 10.3389/fnins.2021.695138, 15 Related articlesMultimodal Assessment of the Motor System in Patients With Chronic Ischemic StrokeMaria Nazarova, et al. Stroke. 2021;52:241-249 January 2021Vol 52, Issue 1 Advertisement Article InformationMetrics © 2020 American Heart Association, Inc.https://doi.org/10.1161/STROKEAHA.120.032992PMID: 33317413 Originally publishedDecember 15, 2020 Keywordspyramidal tractdiffusion MRImotor cortexEditorialsmagnetic resonance imagingPDF download Advertisement SubjectsIschemic StrokeMagnetic Resonance Imaging (MRI)NeurostimulationPrognosisQuality and Outcomes