The Level of Cortical Afferent Inhibition in Acute Stroke Correlates With Long-Term Functional Recovery in Humans
Background and Purpose—Using transcranial magnetic stimulation, we investigated short-interval intracortical inhibition and short-latency afferent inhibition in acute ischemic stroke.
Methods—We evaluated short-interval intracortical inhibition and short-latency afferent inhibition in the affected hemisphere and unaffected hemisphere in 16 patients and correlated electrophysiological parameters with outcome at 6 months.
Results—Affected hemisphere short-latency afferent inhibition was significantly reduced in patients, and short-latency afferent inhibition level correlated with functional outcome.
Conclusions—Reduced afferent inhibition in acute stroke correlates with long-term recovery.
Changes in gamma-aminobutyric acid (GABA)-ergic activity in perilesional cortex after stroke have a central role in recovery.1 Inhibitory circuits of human cerebral cortex can be evaluated using paired–pulse transcranial magnetic stimulation, short-interval intracortical inhibition (SICI), or by coupling peripheral nerve stimulation with transcranial magnetic stimulation in short-latency afferent inhibition (SAI).2 Both inhibitory phenomena are mediated by inhibitory interneurons that use GABAA receptors, but different receptor subtypes are involved in SICI and SAI.2
We investigated SICI and SAI in acute stroke and evaluated the correlation between the level of cortical inhibition and functional outcome at 6 months.
Methods and Patients
Sixteen patients (mean age, 66.8±13.4 years) with first-ever stroke were recruited. Acute-phase evaluation was based on the National Institutes of Health Stroke Scale. Outcome at 6 months was assessed using modified Rankin Scale (mRS). This study was performed according to the Declaration of Helsinki and was approved by the local ethics committee. Patients gave their informed consent before participation.
Patients underwent brain magnetic resonance imaging. Seven patients had a subcortical stroke, whereas 9 patients showed cortical and subcortical involvement. To evaluate whether SAI changes were correlated with structural abnormalities of cholinergic systems, we estimated the damage extent of pathways emanating from nucleus basalis of Meynert: medial pathway, Capsular Lateral pathway, and Perisylvian Lateral pathway.3 For further details, see Supplemental Methods and Supplemental Figure I (http://stroke.ahajournals.org).
We evaluated active motor threshold and resting motor threshold, amplitude of motor-evoked potentials (MEP), SICI at 2 ms interstimulus interval, and SAI at interstimulus intervals from N20 latency plus 2, 3, and 4 ms. We evaluated both affected hemispheres (AH) and unaffected hemispheres (UH).
Because it has been suggested that a change in the slope of input–output curve may influence the amount of cortical inhibition,4 we also obtained AH input–output curve using increasing stimulus intensities and evaluated whether there was a correlation between slope of input–output curves and amount of AH-SAI.
Data obtained in patients were compared with those obtained in 13 healthy subjects (mean age, 70.4±11 years).
Comparison between AH and UH was performed by means of paired t-test, after checking frequency distributions and, eventually, transformed raw values, to achieve a better fit to gaussianity and a reduction of biasing effects of outliers (such as for MEP values). Comparisons of stroke patients versus healthy subjects were performed by means of t-test for independent samples.
Associations between electrophysiological findings and clinical outcome (mRS at 6 months) were assessed by means of nonparametric Spearman's rho. The potential effect of the lesion site on mRS was assessed with Mann-Whitney U test. Electrophysiological measures associated with clinical outcome were further investigated through partial correlation analysis. More specifically, the correlation between AH-SAI and mRS-6-months was controlled for baseline National Institutes of Health Stroke Scale with the following formula:
Here, the left term indicates partial correlation between SAI and mRS, controlling for National Institutes of Health Stroke Scale, and the right term comprises the usual bivariate nonparametric correlations. This measure, after the appropriate transformation, follows a t-distribution with (n−3) degrees of freedom.
Correlation between AH-SAI and the recruitment slope (indexed by the linear increase of MEP amplitude with respect to stimulation increase) was evaluated with Spearman's rho. Because stroke-induced functional changes in cortical excitability may be influenced by stroke location and distribution,5 we evaluated the effect of lesion site (subcortical or cortical–subcortical) on SAI using Mann-Whitney U test.
Nonparametric Spearman's ρ was used to correlate AH-SAI with percentages of lesional voxels in cholinergic pathways.
Significance levels were adjusted according to Bonferroni procedure to control the risk of α-inflation.
Results are summarized in Figure. AH-SAI and AH-MEP amplitude were lower than corresponding UH and control values.
No evidence of association between electrophysiological parameters and stroke severity in the acute phase was found (consistently P>0.05). Looking at correlations with clinical status at 6 months, the only significant associations were found with AH-SAI (Table).
When the effect of AH-SAI on mRS was adjusted for the confounding effect of baseline clinical status (National Institutes of Health Stroke Scale at T0), the nonparametic partial correlation remained significant (rho=0.66; P=0.016), suggesting its relevance even equalizing for baseline clinical status.
There was no correlation between AH-SAI and either slope of the input–output curve (Spearman's rho=0.12; P=0.676) or site of the lesion (Mann-Whitney U, P=0.958). Also, there was no correlation between site of the lesion and recovery at 6 months (Mann-Whitney U, P=0.99).
Involvement of cholinergic pathways was limited (Supplemental Table I), and there was no correlation between AH-SAI and either percentage of lesional voxels of medial pathway (ρ=−0.39; P=0.52), lateral cholinergic pathways (ρ=−0.47; P=0.264), and lateral perforant pathway (ρ=−0.43; P=0.376) or percentage of lesional voxels in the 3 pathways considered together (Spearman's ρ=−0.5; P=0.17).
We report for the first time a suppression of afferent inhibition in acute stroke. AH-SAI level was correlated with recovery at 6 months.
SAI is produced by afferent inputs, and central cholinergic pathways are involved in SAI2; thus, a lesion of these circuits might explain its reduction. However, the absence of consistent sensory deficits and/or abnormalities of N20 wave of somatosensory evoked potentials, the limited involvement of cholinergic pathways, and the absence of any correlation between involvement of cholinergic pathways and level of SAI, make this hypothesis unlikely.
We speculate that SAI suppression might be produced by functional changes in central inhibitory circuits.2 Because SAI is probably mediated by the α5-subunit,2 we suggest that its suppression might be related to a reduction of activity related to this subunit. Interestingly, a recent experimental study showed that pharmacological antagonization of α5-subunit activity promotes functional recovery after stroke.1 Long-term potentiation can be induced in motor cortex by stimulation of sensory cortex,6 and it has been proposed that long-term potentiation produced by sensory inputs might promote cortical reorganization after a lesion.7 Thus, it can be speculated that reduced SAI level could enhance sensory stimuli-related long-term potentiation phenomena in the motor cortex with a positive effect on relearning related recovery.
The online-only Data Supplement is available at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.111.631085/-/DC1.
- Received July 1, 2011.
- Accepted July 21, 2011.
- © 2012 American Heart Association, Inc.