Strategic Infarct Locations For Stroke-induced Left Ventricular Dysfunction
Chensheng Pan1, Yuchao Jia1, Zhou Zhu1, Suiqiang Zhu1
1Department of Neurology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology
Objective:
To identify the neural substrates of post-stroke left ventricular dysfunction (PSLVD) using lesion-symptom mapping.
Background:
PSLVD, a subtype of stroke-heart syndrome, is associated with poor outcomes in stroke survivors and is considered a consequence of stroke-induced central autonomic network (CAN) disruption. However, the specific CAN component responsible for PSLVD remains elusive.
Design/Methods:
Patients admitted to our hospital during 2015-2022 with a final diagnosis of ischemic stroke were retrospectively screened. Inclusion criteria: (1) first-ever ischemic stroke confirmed with MRI; (2) echocardiography within three days post-stroke. Exclusion criteria: (1)history of focal brain damage; (2)history of coronary heart disease, heart failure, valvular disease, severe arrhythmia, cardiac surgery, etc. (3)severe systemic conditions e.g. sepsis. Infarcts were manually segmented and spatially normalized. Left ventricular ejection fraction (LVEF) was used to reflect ventricular function. Voxel-based lesion-symptom mapping(VLSM) was performed in FSL Randomise to test the association between lesion location and LVEF while controlling for six covariates (lesion volume, age, sex, history of hypertension, diabetes mellitus, hyperlipidemia). We performed 10-fold cross-validation to test the robustness and generalizability of VLSM results. The intersection between each patient’s lesion and external VLSM map was quantified as “location impact score”. The LVEF residuals (six covariates regressed out) were used to correlate with location impact score.
Results:
1,539 eligible patients were included. VLSM found a significant cluster in cerebellar vermis was negatively correlated with LVEF while clusters in anterior cingulate cortex (ACC) and medial prefrontal cortex (mPFC) were positively correlated with LVEF (TFCE-PFWE<0.05). Results were highly replicable in cross-validation. Location impact score was significantly correlated with LVEF (Spearman r=0.125, P<10-6).
Conclusions:
Lesions involving vermis may increase the risk of ventricular hypofunctioning, while those involving ACC-mPFC may result in ventricular hyperfunctioning. These strategic regions may serve as promising neuromodulation targets to treat ventricular dysfunction caused by stroke or other etiologies.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.