Acute imaging does not improve ASTRAL score's accuracy despite having a prognostic value

Area of Science:

• Neurology
• Radiology
• Medical Imaging

Background:

• The ASTRAL score is a validated tool for predicting functional outcomes in acute ischemic stroke patients at three months.
• Assessing the added value of multimodal imaging to existing prognostic scores is crucial for refining stroke management.

Purpose of the Study:

• To determine if multimodal imaging data enhances the predictive accuracy of the ASTRAL score for functional outcomes in acute ischemic stroke.
• To identify specific imaging biomarkers that independently predict outcomes in stroke patients.

Main Methods:

• A multivariate logistic regression analysis was performed on patients from the ASTRAL registry.
• Covariates from parenchymal, vascular, and perfusion imaging were added to the ASTRAL score model.
• Area Under the Curve (AUC) was calculated and compared to assess improvements in predictive accuracy.

Main Results:

• Most multimodal imaging covariates did not significantly improve the ASTRAL score's predictive accuracy (AUC).
• Alberta Stroke Program Early CT score and extracranial arterial pathology were independent predictors but did not significantly increase AUC.
• Exploratory analyses in subgroups also failed to show improved AUC with added imaging covariates.

Conclusions:

• Multimodal imaging, as applied, does not enhance the ASTRAL score's accuracy for predicting functional outcomes in acute ischemic stroke.
• Selected radiological parameters in specific patient subgroups may offer additional prognostic value, warranting further investigation.