Neurophysiological Predictors of Aphasia Recovery in Patients with Large Left-Hemispheric Infarction: A Mismatch Negativity Study
Introduction
Large hemispheric infarction (LHI), which includes sub-total or complete infarction in the territory of the middle cerebral artery (MCA) with or without involvement of the anterior or posterior cerebral artery territories, occurs in up to 10% of patients with supratentorial ischemia. Global aphasia, often accompanying severe hemiplegia in patients with left LHI, significantly impacts the efficacy of rehabilitation therapy and quality of life. Despite considerable attention to aphasia recovery, reliable predictors of language outcomes remain elusive.
Clinical factors such as age, lesion characteristics, education, and possibly sex have been established for predicting aphasia recovery, but they are either insensitive or contradictory, explaining only about 40% of the variance. Anatomical predictors have also been studied, but functional magnetic resonance imaging (MRI) examinations are difficult or unfeasible for critically ill patients with LHI. Therefore, alternative predictors are needed for these severely aphasic patients.
Electroencephalography (EEG), a non-invasive functional imaging technique with high temporal resolution, is a suitable choice for studying auditory discrimination. Mismatch negativity (MMN), an automatic brain response to deviant or infrequent changes in the acoustic environment, occurs 100 to 250 ms after stimulus onset. MMN has good replicability at the individual level and has been used to probe speech processing, cognitive functions, and even predict disorders of consciousness. MMN has also been confirmed as a useful tool for mapping functional language recovery after left-hemisphere stroke, particularly during the chronic recovery stage. The potential of MMN as a biomarker for predicting neurofunctional outcomes and monitoring neurorehabilitation is largely unexplored. MMN may provide valuable information regarding the integrity of auditory processes in the early stages of stroke when patients are not yet able to fully cooperate. Therefore, this study hypothesized that MMN could serve as a neurophysiological predictor of aphasia in patients with left LHI.
Methods
Ethical Approval and Participants
This study was approved by the Ethical Committee of Xuanwu Hospital. Informed consent was obtained from the closest relatives of all subjects. Eighteen patients with left LHI were prospectively and consecutively included based on specific inclusion and exclusion criteria. Inclusion criteria included left LHI with a volume of at least two-thirds of the left MCA territory, first onset, right-handedness, age between 18 and 80 years, native Mandarin Chinese speakers living in Beijing or surrounding areas, an aphasia severity rating of 0 according to the Boston Diagnostic Aphasia Examination (BDAE) within the first 20 days, and no obvious cognitive dysfunction before the stroke. Exclusion criteria included patients who received decompressive craniectomy, those with a history of neuropsychiatric disorders, left LHI accompanied by right hemispheric or posterior circulation infarction, no N100 in the bilateral hemispheres, administration of sedatives within 24 hours before EEG data collection, severe complications such as electrolyte and metabolic disturbances, seizures, or loss to follow-up.
Clinical information, including age, sex, intra-vascular therapy, low-density volume on computed tomography (CT), National Institutes of Health Stroke Scale (NIHSS) score, and whether the patient underwent decompressive craniectomy, was collected. All patients were followed up for three months, and the aphasia severity rating scale (ASRS) of the BDAE was used to determine the severity of aphasia at the second collection session and three months post-stroke. Patients were divided into two groups based on their ASRS score at three months: the good recovery group (ASRS score ≥1) and the poor recovery group (ASRS score = 0).
Stimulus Paradigm
An oddball paradigm with pure tones was used to elicit MMN. The stimuli consisted of standard (1000 Hz) and deviant (1500 Hz) pure tones, presented at 90% and 10% of the total stimuli, respectively. Both tones were at an intensity of 70 dB sound pressure level (SPL), with a duration of 50 ms and an inter-stimulus interval of 600 ms.
EEG Data Collection and Analysis
EEG data were collected over two sessions: the first session within the first seven days post-onset and the second session between 10 to 20 days post-onset. Continuous EEG data were recorded using a 64-electrode EEG system according to the extended international 10–20 system. Data were online-filtered with a bandpass of 0.05 to 70 Hz, sampled at 512 Hz, and online referenced to CPz.
EEG data were analyzed as event-related potentials (ERPs) using the EEGLAB toolbox in MATLAB. Pre-processing included bandpass filtering (0.05–30.00 Hz), referencing to the averaged left and right mastoids, and independent component analysis to remove eye movement and blinking artifacts. Epochs from 100 ms pre-stimulation to 500 ms post-stimulation were baseline corrected, and epochs containing artifacts greater than ±100 mV were rejected. Individual sweeps were averaged for standard and deviant stimuli separately, and difference waveforms were obtained by subtracting the standard ERP from the deviant ERP. MMN peaks were visually identified as the most negative deflection between 100 and 200 ms post-stimulation.
Statistical Analysis
Mixed three-way analysis of variance (ANOVA) was used to compare MMN amplitudes and laterality indexes (LIs) between the two groups at each session, with session and region of interest (ROI) as within-subject factors and group as the between-subject factor. Receiver operating characteristic (ROC) curves were used to identify the most predictive factor. Statistical analysis was performed with SPSS, and significance was defined as P < 0.05.
Results
Clinical Characteristics
The lesion sites in all patients included the frontal, temporal, insular, and parietal cortices, basal ganglia, and sub-cortical structures to various degrees. The ASRS score at three months post-stroke was 0 in eight patients, 1 in seven patients, and 2 in three patients. No significant differences were found in sex, age, or collection time between the two groups. All patients received no recanalization therapy or failed to achieve recanalization.
Mismatch Negativity
No significant differences were found in the numbers of standard or deviant trials between the two groups. The mixed three-way ANOVA showed no interaction effect of session × ROI × group or ROI × group for MMN amplitude, suggesting that the absolute amplitude of MMN was not a good predictor of language outcome.
For LIs of MMN, no interaction effect of session × ROI × group was found, but an interaction effect of session × group was observed. In the poor recovery group, the LIs at all ROIs except the parietal area became more negative at session 2 compared to session 1, but this effect was not observed in the good recovery group. Significant differences were observed in the LIs at session 2 between the two groups, suggesting that the LIs at session 2 were predictors of the language outcome.
ROC curve analysis showed that the LI of the perisylvian electrodes had the greatest predictive value, with the highest area under the curve (AUC) of 0.963. The LI cut-off value of -0.36 over the perisylvian area had the highest sensitivity (90.0%) and specificity (87.5%) for predicting a good language outcome at three months post-stroke.
Discussion
Time Point for Language Outcome Prediction
MMN at approximately two weeks, but not within seven days, was more valuable for predicting language outcome. This observation may be due to brain edema in the first days post-stroke, which dampens neural processes and affects brain electrical activity. After brain edema subsides, MMN amplitudes can reflect language function more accurately.
Brain Hemisphere(s) Associated with Language Recovery
The LIs of MMN amplitudes at 10 to 20 days were good predictors of the three-month language recovery, indicating that LIs were more sensitive than absolute amplitudes. A higher LI indicates better function of the left hemisphere or weaker compensation of the right hemisphere, suggesting that language function is related to the activity of the left hemisphere and the right homologous regions. Successful restoration of language networks depends on efficient reintegration of ipsilateral language-related areas or their neighboring regions, rather than recruitment of new regions.
Brain Regions Related to Language Recovery
LIs of MMN over any selected ROIs at two weeks post-infarction were predictors of language outcome, indicating the essential roles of these regions in language processing. The perisylvian area, including the inferior frontal gyrus, superior temporal gyrus, angular gyrus, and supramarginal gyrus, plays the most crucial role in overall language recovery.
Conclusion
MMN amplitudes at approximately two weeks post left-hemisphere stroke serve as sensitive predictors of language outcome, with the LI over the perisylvian area exhibiting the best predictive value. This study highlights the importance of considering both the timing and specific brain regions when predicting aphasia recovery in patients with large left-hemispheric infarction.
doi.org/10.1097/CM9.0000000000000459
Was this helpful?
0 / 0