Modulating motor function after stroke: Differential effects of high-frequency rTMS on movement kinematics and neural activity in subcortical and cortical MCA stroke

Objective: Following the concept of interhemispheric competition, up-regulating the cortical excitability of the ipsilesional motor cortex (M1) may improve dexterity of the affected hand after stroke. We here examined the effects of 10Hz rTMS over ipsilesional M1 on both movement kinematics and neural activity assessed with 3 Tesla fMRI after (i) subcortical and (ii) cortical stroke.

Methods: 15 subjects with impaired dexterity due to subcortical middle cerebral artery (MCA) stroke and 13 subjects with cortical MCA stroke received 10Hz rTMS (5s stimulation, 25s break, 1000 pulses, 80% of the resting motor threshold of the unaffected hemisphere) for ten minutes applied over: (i) ipsilesional M1 (ii) and vertex (control stimulation). For behavioural testing, subjects performed finger, hand and elbow tapping and grasping movements as well as a simple reaction time task with both hands prior to and following each rTMS application. For fMRI, subjects performed index finger tappings with their affected or unaffected hand before and after each rTMS application using an on-off block-design (gradient echo, TR=1.82s, TE=30ms, analysis with SPM5).

Results: 10Hz rTMS applied over ipsilesional M1 transiently improved movement kinematics of the affected hand in subcortical stroke (p<0,05), but had no improving effect with a tendency to transient deterioration on movement kinematics of the affected hand in cortical stroke (P<0.05). The DTI analysis showed that all subcortical stroke lesions involved the corticospinal tract (CST), while the cortical lesions had only minor anatomical relationships to the CST. At the neural level, rTMS applied over ipsilesional M1 caused changes not only in M1, but increased neural activity in a wide network of motor regions, especially in the basal ganglia, frontal operculum and cingulate motor areas. A correlation analysis testing for voxels correlating with behavioral improvement rates demonstrated a significant correlation (P<0.05) of ipsilesional M1 activity and increase in maximum finger tapping speed after rTMS over M1.

Conclusion: These data argue against a non-critical use of rTMS in the rehabilitation of stroke. High frequency rTMS may develop both beneficial and detrimental effects depending on lesion site, extension and involvement of fibre tracts. A critical re-evaluation of the interhemispheric competition model is essential in order to implement rTMS as a safe and effective procedure in current concepts for stroke rehabilitation.

Fig. 1: Effects of rTMS over (i) ipsilesional M1 and (ii) vertex (control stimulation) on neural activity while tapping with the index finger of the stroke-affected hand.