Increased Neuron-Specific Enolase Level Predicts Symptomatic Intracranial Hemorrhage in Patients with Ischemic Stroke Treated with Endovascular Treatment

Abstract

Background: Neuron-specific enolase (NSE), which is a highly specific marker for neurons, could be a predictor for prognosis in patients with symptomatic intracranial hemorrhage (sICH) with acute ischemic stroke who are receiving endovascular treatment (EVT). This study aimed to investigate the relationship between NSE and sICH in patients with acute anterior circulation stroke undergoing EVT.

Methods: A total of 215 consecutive patients with acute stroke treated with EVT were included. Patients with stroke and acute anterior circulation occlusion, receiving EVT treated at our hospital, were enrolled between January 2017 and August 2021. NSE level was measured on arrival at the neurology intensive care unit after EVT. The patients were divided into 2 groups according to whether sICH was present. Univariate and multivariate analyses were performed. NSE level was also incorporated into the TAG score (modified Thrombolysis in Cerebral Infarction score, Alberta Stroke Program Early CT Score, and glucose level), which was developed as a scoring system to predict sICH, and the prediction capability was compared with the TAG score alone. Causal inference was performed using the package DoWhy in Python to evaluate the causal relationship between NSE and sICH.

Results: The area under the curve (AUC) value of NSE showed moderate accuracy, with an AUC value of 0.729 (95% confidence interval, 0.655-0.795; P < 0.001). The NSE cutoff value was set at 23.88 ng/mL. When the NSE level ≥23.88 ng/mL, the sensitivity was 58.33% and the specificity was 78.72% (P < 0.001). The AUC for the TAG + NSE score was 0.801 compared with an AUC of 0.632 for the TAG score (Z = 2.034; P = 0.042). A causal inference model using the DoWhy library shows a proportional relationship between NSE and the diagnosis of sICH.

Conclusions: This study is the first to show that increased NSE level is an independent predictor of sICH in patients with acute anterior circulation stroke who are undergoing endovascular treatment.

Keywords: Acute anterior circulation stroke; Endovascular treatment; Neuron-specific enolase; Symptomatic intracranial hemorrhage.