Plasma brain-derived tau correlates with cerebral infarct volume

Abstract

Background: A blood-based biomarker that accurately reflects neuronal injury in acute ischemic stroke could be an easily accessible and cost-effective complement to clinical and radiological evaluation. Here, we investigate whether plasma levels of the novel biomarker brain-derived tau (BD-tau) reflect cerebral infarct volumes and whether BD-tau can improve clinical outcome prediction.

Methods: The present study included 713 consecutive cases from two different hospital-based cohorts, the Sahlgrenska Academy Study on Ischemic Stroke (SAHLSIS) and SAHLSIS phase 2 (SAHLSIS2). Acute stroke severity was determined by the Scandinavian Stroke Scale converted to the National Institutes of Health stroke scale (NIHSS) in SAHLSIS and by the NIHSS in SAHLSIS2. All participants were assessed for functional outcome 3 months after stroke by the modified Rankin Scale, and 254 participants in SAHLSIS had quantitative neuroimaging available.

Findings: Plasma BD-tau concentrations and cerebral infarct volumes were highly correlated (ρ 0.72, p < 0.001). BD-tau improved the prognostic accuracy of suffering an unfavorable outcome over age and stroke severity in the whole cohort. However, the gain in predictive power was dependent on stroke severity and infarct location. The largest improvement was observed for mild ischemic strokes (NIHSS <5; area under the curve [AUC] = 0.73 for age + NIHSS versus AUC = 0.84 with addition of BD-tau; DeLong p 0.02), posterior circulation stroke (AUC = 0.75 vs. AUC = 0.84; DeLong p 0.06) and more specifically for infarcts in the brainstem/cerebellum (AUC = 0.74 vs. 0.87; DeLong p 0.009).

Conclusion: Plasma BD-tau can provide information on the extent of acute neuronal damage in ischemic stroke and adds prognostic value for outcome, especially for mild and posterior circulation strokes.

Keywords: blood biomarkers; brain‐derived tau; infarct volume; stroke; stroke severity.

Fig. 1

Plasma brain‐derived tau (BD‐tau) concentrations are correlated with infarct volume in acute ischemic stroke. (a) Box plots of BD‐tau concentrations (top) and infarct volumes (bottom) stratified by stroke severity, magnetic resonance imaging (MRI)‐based vascular territory, and anatomical location. (b) Scatter plots of BD‐tau versus infarct volumes stratified by stroke severity and infarct location.

Fig. 2

Receiver operating characteristic curves for accuracy of brain‐derived tau (BD‐tau) and clinical variables for predicting unfavorable outcome. Analyses were performed in the combined Sahlgrenska Academy Study on Ischemic Stroke (SAHLSIS) + SAHLSIS phase 2 (SAHLSIS2) cohort (a) for all ischemic stroke and stratified based on; (b) clinical stroke severity (i.e., NIHSS < or ≥5); (c) Oxfordshire Community Stroke Project (OCSP) classifications (i.e., vascular territory); and (d) by symptoms‐ and neuroimaging‐based anatomical location. Table 3 provides detailed receiver operating characteristics curve (ROC) analysis.