Blood Biomarkers in Acute Stroke

Introduction

Blood biomarkers may improve the diagnostic workup and may facilitate therapeutic decision making in acute stroke patients. This includes the differentiation of “true” stroke patients from stroke mimics, the discrimination between intracerebral hemorrhage (ICH) and ischemic stroke, the prediction of complications of thrombolysis, the etiologic classification of ischemic stroke, and the prediction of recurrent events in patients having a first transitory ischemic attack (TIA). Since 1990s, significant progress has been achieved in biomarker research in these potential fields of application. Here, an overview on relevant studies is provided.

Methodological Aspects

Most research on blood biomarkers in stroke has been conducted using three different methodological strategies . The first is to search for a single protein biomarker or a protein biomarker panel to be used as a diagnostic test. This hypothesis-driven approach includes proteins associated with glial or neuronal tissue injury (such as glial fibrillary acidic protein (GFAP), S100B, and neurofilament protein), blood–brain barrier damage (matrix metalloprotease 9 (MMP-9)), inflammation (C-reactive protein, CRP, interleukin 6, IL-6), and coagulation (fibrinogen, D-dimer). They are typically measured on standard enzyme-linked immunosorbent assay (ELISA) or multiplex ELISA platforms. The second strategy is based on proteomics technology, which allows a hypothesis-free approach for protein biomarker research. Analyses are typically performed using two-dimensional gel electrophoresis and mass spectrometry. Several promising candidate biomarkers have been identified by this approach, but prospective validation in independent cohorts of stroke patients was not always successful. The third approach is to use modern high-throughput microarray technologies to study RNA expression profiles in whole blood. Interesting findings from explorative studies have been published, but translation into clinical routine is challenging, mostly due to the high methodological efforts associated with the use of these techniques.

Identification of Patients With Ischemic Stroke

The identification of “true” stroke patients in a collective of patients with symptoms suggestive of acute stroke has always been a prioritized field in blood biomarker research. As ischemic stroke is a heterogeneous disease, it was speculated that a biomarker panel rather than a single biomarker might be able to differentiate between ischemic stroke patients and stroke mimics. Pilot studies identified several protein biomarkers that were found elevated in the blood of patients with ischemic stroke, but not in healthy controls. However, a prospective validation study on more than 1000 patients suspected of having acute stroke revealed disappointing diagnostic accuracy values for a four-marker panel including S100B, MMP, D-dimer, and brain natriuretic peptide (BNP) for differentiating “true” stroke patients from stroke mimics . It is likely that the various stroke mimic conditions including migraine, brain tumors, and epileptic seizures also cause alterations in biomarker levels, which reduce the diagnostic potential of the measure. In 2011, caspase-3 and D-dimer were described as candidate proteins for a panel differentiating between ischemic stroke patients and stroke mimics. Other biomarker panels included eotaxin, epidermal growth factor receptor, and prolactin. However, prospective independent validation of these markers is still missing.

RNA expression profiles derived from whole blood were also investigated in this context . A 22-gene panel reached a 78% sensitivity and a 80% specificity for differentiating ischemic stroke from controls. Later on, a study found 1335 genes with a different expression pattern between ischemic stroke and controls. A panel consisting of 18 of these 1335 genes differentiated ischemic stroke patients from controls. A critical point is to compare RNA expression between ischemic stroke patients and patients with stroke mimicking conditions (rather than healthy controls). In this context, a study reported that a 97-gene profile may differentiate ischemic stroke patients from patients with myocardial infarction and patients with vascular risk factors.