Induced sensorimotor cortex plasticity remediates chronic treatment-resistant visual neglect

Abstract

Right brain injury causes visual neglect - lost awareness of left space. During prism adaptation therapy, patients adapt to a rightward optical shift by recalibrating right arm movements leftward. This can improve left neglect, but the benefit of a single session is transient (~1 day). Here we show that tonic disinhibition of left motor cortex during prism adaptation enhances consolidation, stabilizing both sensorimotor and cognitive prism after-effects. In three longitudinal patient case series, just 20 min of combined stimulation/adaptation caused persistent cognitive after-effects (neglect improvement) that lasted throughout follow-up (18–46 days). Moreover, adaptation without stimulation was ineffective. Thus stimulation reversed treatment resistance in chronic visual neglect. These findings challenge consensus that because the left hemisphere in neglect is pathologically over-excited it ought to be suppressed. Excitation of left sensorimotor circuits, during an adaptive cognitive state, can unmask latent plastic potential that durably improves resistant visual attention deficits after brain injury.

Anatomical predictors of successful prism adaptation in chronic visual neglect

“…In healthy individuals, rightward PA increased BOLD response in the left inferior parietal cortex 30 . Our cortical thickness and diffusion results suggest an implication of left prefrontal cortex in prism-induced neglect compensation, consistent with the observation that effective PA in neglect is facilitated by TMS-based disinhibition of the left motor cortex 31 . Altogether, these findings fit well with the fronto-parietal organization of attention networks in the primate brain 32 .…”

Causal explanation of individual differences in human sensorimotor memory formation

“…Here we exploit this variation to test a causal mechanistic hypothesis about how sensorimotor memories in human brain are formed, retained and enhanced. Primary motor cortex (M1) has been identified as a critical region for the maintenance of newly acquired visuomotor maps following sensorimotor adaptation in both primates (Wise et al 1998;Li et al 2001;Paz et al 2003;Paz et al 2005;Inoue et al 2016) and humans (Hadipour-Niktarash et al 2007;Hunter et al 2009;Galea et al 2010;Landi et al 2011;Leow et al 2016;O'Shea et al 2017). We previously demonstrated that left M1 anodal transcranial direct current stimulation (a-tDCS) applied during adaptation to a 10-degree rightward displacement of the visual field (prism adaptation, PA) enhances consolidation of the prism after-effect (AE) in the healthy brain .…”

“…Due to the size of the voxel, adjacent regions of sensory cortex (S1) were also included in the measure of M1 metabolites. This is a common methodological limitation of MRS studies (Stagg et al 2009;Lunghi et al 2015;Antonenko et al 2017;Ip et al 2017;Kolasinski et al 2017;O'Shea et al 2017). Although we cannot rule out the contribution of S1 to our results, M1 is likely to play a predominant role because of a convergence of studies implicating it in the consolidation of adaptation memory (Wise et al 1998;Li et al 2001;Paz et al 2003;Paz et al 2005;Hadipour-Niktarash et al 2007;Landi et al 2011;Leow et al 2016).…”

“…Within any CLP block, targets appeared either 12 cm to the left or 12 cm to the right of the centre of the screen (50% on each side), following a pseudorandom sequence in which target location transitions were generated if the same location appeared twice in a row. Visual feedback was limited to the last two thirds of the reaching movement in order to limit strategic adjustment(Redding and Wallace 1996;O'Shea et al 2014;Inoue et al 2015;O'Shea et al 2017) and lasted until 500 ms after participants touched the screen to allow for them to perceive their terminal reach endpoint error. After this time, the LC shutter turned opaque and participants had to return to the starting position (i.e.…”

“…The differences in movement speed and target location between prism exposure (CLP trials) and AE probes (OLP) enabled us to assay generalised AE that were not contaminated by extensive training of a specific movement dynamic and local learning at a certain target location(Kitazawa et al 1997). This type of AE is thought to be relevant for neglect rehabilitation(Serino et al 2006;O'Shea et al 2017).In all four PA sessions, the tDCS electrodes were positioned and the stimulation switched on at the end of the baseline phase (20 CLPs, 30 OLPs). In the online condition, the PA phase (E1-7: 10 CLPs per block during the first 2 blocks, 20 CLPs per block in the remaining ones; AE1-7: 15 OLPs per block) started as soon as the stimulation reached its constant level.…”