Blood Biomarkers for the Early Diagnosis of Stroke: The Stroke-Chip Study

Stroke diagnosis could be challenging in the acute phase. We aimed to develop a blood-based diagnostic tool to differentiate between real strokes and stroke mimics and between ischemic and hemorrhagic strokes in the hyperacute phase.

The Stroke-Chip was a prospective, observational, multicenter study, conducted at 6 Stroke Centers in Catalonia. Consecutive patients with suspected stroke were enrolled within the first 6 hours after symptom onset, and blood samples were drawn immediately after admission. A 21-biomarker panel selected among previous results and from the literature was measured by immunoassays. Outcomes were differentiation between real strokes and stroke mimics and between ischemic and hemorrhagic strokes. Predictive models were developed by combining biomarkers and clinical variables in logistic regression models. Accuracy was evaluated with receiver operating characteristic curves.

From August 2012 to December 2013, 1308 patients were included (71.9% ischemic, 14.8% stroke mimics, and 13.3% hemorrhagic). For stroke versus stroke mimics comparison, no biomarker resulted included in the logistic regression model, but it was only integrated by clinical variables, with a predictive accuracy of 80.8%. For ischemic versus hemorrhagic strokes comparison, NT-proBNP (N-Terminal Pro-B-Type Natriuretic Peptide) >4.9 (odds ratio, 2.40; 95% confidence interval, 1.55-3.71; P<0.0001) and endostatin >4.7 (odds ratio, 2.02; 95% confidence interval, 1.19-3.45; P=0.010), together with age, sex, blood pressure, stroke severity, atrial fibrillation, and hypertension, were included in the model. Predictive accuracy was 80.6%.

The studied biomarkers were not sufficient for an accurate differential diagnosis of stroke in the hyperacute setting. Additional discovery of new biomarkers and improvement on laboratory techniques seem necessary for achieving a molecular diagnosis of stroke.