Biomarkers have various potential uses in stroke management [4] (Table 1).

The actively investigated areas for the application of biomarkers include the use of blood biomarkers for the early diagnosis of stroke and determining the status of its emergency. For example, blood levels of creatine kinase MB (CK-MB) fraction and troponin have been evaluated for the early diagnosis of myocardial infarction. However, compared to myocardial infarction, no effective biomarkers have been validated for stroke. For early diagnosis, biomarker levels must increase in the blood early after stroke and be specific to not only brain tissues but also the pathophysiology of stroke. However, it is difficult to identify the increase in blood concentrations of brain-specific biomarkers at early phases due to the blood-brain barrier, which continues to be a limitation. Furthermore, clinically, as diffusion-weighted magnetic resonance imaging has high imaging speed; and high sensitivity and specificity, it is difficult for novel biomarkers to replace this diagnostic modality.

However, biomarkers provide additional information that can assist with more rapid and accurate decision-making for disease management. For example, information from biomarker panels obtained at admission accelerated decision-making for the use of thrombolytics [5]. Biomarkers can also be useful in rehabilitation to predict the response to specific treatments or for stratification based on the risk of complications.

Studies have investigated the potential of blood biomarkers for the prediction of functional prognosis after stroke. For example, fibrinogen [6] and CRP [7] levels are associated with a poor prognosis. A systematic review by Whiteley et al. [8] in 2009 indicated that while some biomarkers were consistently correlated with stroke prognosis in studies published between 1996 and 2007, these did not increase the predictability in clinical settings.

In their 2012 prospective study using various blood biomarkers to predict the prognosis of 270 stroke patients, Whiteley et al. [9] measured blood biomarker concentrations within 24 hours of the onset of ischemic stroke and evaluated mortality rate or correlation with disability after 3 months. While they reported a significant correlation between interleukin-6 and N-terminal-pro hormone B-type natriuretic peptide levels, the predictability did not increase for these two biomarkers compared to prognosis prediction based on the clinical information. Thus, while biomarkers significantly associated with poor prognosis have been identified, superior clinical prediction models are already available, namely the National Institutes of Health Stroke Scale and age.

In consideration of the complex pathophysiology of stroke and resulting disabilities, there are clear limitations of blood biomarkers. First, biomarkers do not provide information on the lesion location. Even if the infarct volume may be the same, the functional prognosis differs greatly between those occurring in the parietal association area or the internal capsule posterior limb. However, it is difficult to predict this differential information based only on blood biomarker concentrations. Second, the blood-brain barrier results in a significant delay in the expression of brain tissue-specific biomarkers in the blood. Third, study results indicate that it is difficult for a single biomarker to have sufficient sensitivity and specificity for clinical applications. Therefore, approaches for biomarker panels using multiple important biomarkers will be required.

Unfortunately, there have been very little literature on biomarkers related to stroke rehabilitation. Clinical and neurophysiological measurements, as well as imaging study results, have previously been used to predict patients’ functional recovery [10,11]. In order to prepare for the coming era of precision medicine, it is essential to not only adequately evaluate and quantify various functional outcomes in detail but also to understand the biological changes that occur at the beginning of the disease and throughout rehabilitation therapy. For this reason, prospective clinical studies that not only evaluate the functional level of patients but also collect blood biomarkers should be actively conducted. It would be ideal to form a multicenter research consortium including investigators from various areas. Furthermore, large-scale studies such as the KOSCO project need to fully consider, from the planning stage, whether to include blood biomarker collection.