Clinical Features and Voxel-Based-Symptom-Lesion Mapping of Silent Aspiration in Acute Infratentorial Stroke

Abstract

AbstractPost-stroke dysphagia (PSD) is a severe and common complication after ischemic stroke. The role of silent aspiration as an important contributing factor in the development of a dysphagia-associated complications, in particular aspiration-associated pneumonia has been insufficiently understood. The aim of this study was to investigate the characteristics and risk factors of silent aspiration in patients with acute infratentorial stroke by FEES and to identify culprit lesions in stroke patient with a high risk of silent aspiration via voxel-based-symptom-lesion mapping (VBS/ML). This study is a retrospective observational study based on a prospectively collected FEES and stroke database. Consecutive patient cases with acute ischemic infratentorial stroke and FEES examination between 2017 and 2022 were identified. Group allocation was performed based on PAS scores. Imaging analysis was performed by manual assignment and by VBS/ML. Group comparisons were performed to assess silent aspiration characteristics. Binary logistic regression analysis was performed to determine if baseline clinical, demographic, and imaging parameters were helpful in predicting silent aspiration in patients. In this study 84 patient cases with acute infratentorial stroke who underwent FEES examination were included. Patients were moderately affected at admission (mean NIH-SS score at admission 5.7 SD ± 4.7). Most lesions were found pontine. Overall 40.5% of patients suffered from silent aspiration, most frequently in case of bilateral lesions. Patients with silent aspiration had higher NIH-SS scores at admission (p < 0.05), had a more severe swallowing disorder (p < 0.05) and were 4.7 times more likely to develop post-stroke pneumonia. Patients who underwent FEES examination later than 72 h after symptom onset were significantly more likely to suffer from silent aspiration and to develop pneumonia compared to patients who underwent FEES examination within the first 72 h (p < 0.05). A binary logistic regression model identified NIH-SS at admission as a weak predictor of silent aspiration. Neither in manual assignment of the lesions to brain regions nor in voxel-wise statistic regression any specific region was useful in prediction of silent aspiration. Silent aspiration is common in patients with infratentorial stroke and contributes to the risk for pneumonia. Patients with silent aspiration are more severely affected by stroke, but cannot reliably be identified by NIH-SS at admission or lesion location. Patients suffering from acute infratentorial stroke should been screened and examined for PSD and silent aspiration.