Can Serum Biomarkers Improve Stroke Care?

Stroke is one of the most common causes of hospitalization in high-income countries and among the leading causes of disability worldwide. Yet, unlike many other resource-intensive conditions, there are no established serum biomarkers that can identify stroke, improve its care, or help with prognosis. Biomarkers—including nucleic acids, proteins, hormones, and imaging data—are measurable indicators crucial for understanding biologic or pathologic processes and serve as key tools in diagnosing and classifying diseases. We review the current literature on serum biomarkers to improve stroke care.

Diagnostic Biomarkers

In an era when diagnostic neuroimaging is generally accessible, identifying the location, type, and extent (size) of stroke is relatively straightforward. Nevertheless, in settings with limited imaging resources, or when there is a need to distinguish between transient ischemic attacks (TIA) and stroke mimics, the pursuit of diagnostic stroke serum biomarkers becomes necessary. Much like the application of cardiac troponin in diagnosing cardiac ischemic injury, when imaging is constrained, serum biomarkers could play a pivotal role in improving the diagnosis of cerebral ischemia. Their importance extends to prehospital scenarios, where a reliable biomarker could aid in early diagnosis and triaging, and adequate transportation as required. A list of selected biomarkers by time from symptom onset is presented in Table 1.

Biomarkers to Identify TIA or Minor Ischemic Stroke

Because the risk of an ischemic stroke is high in the first few hours to days after TIA, timely and urgent identification of TIA or minor ischemic strokes could prevent another event in up to 1 in 5 people.¹ Park et al observed increased plasma levels of heart-type fatty acid binding protein (H-FABP) during the acute stage of ischemic stroke.² These elevated levels were positively associated with clinical severity, although they demonstrated limited diagnostic precision in distinguishing ischemic stroke and stroke mimics (Table 1).

Ubiquitin fusion degradation protein 1 (UFD1) is a biomarker of cell stress. Altered UFD1 levels in blood plasma after a stroke may help distinguish it from other conditions, making it a useful tool for rapid stroke diagnosis when imaging is not available.³ A study encompassing 3 distinct cohorts from Switzerland, Spain, and North America underscored the effectiveness of UFD1 as a biomarker, notably in early detection of stroke.

Parkinson disease protein 7 (PARK7) and nucleoside diphosphate kinase A (NDKA) were part of a group of 5 biomarkers (including PARK7, NDKA, interleukin-6 [IL-6], placental glutathione S-transferase (GST-P), and N-terminal pro B-type natriuretic peptide [NT-proBNP]) that showed a significant increase within the first 3 hours of focal neurologic symptoms, a pattern not observed in healthy individuals.³ This may occur because both PARK7 and NDKA are released when damage to the brain occurs.⁴

In a study by An and colleagues, an increase in IL-6 (an inflammatory biomarker) was noted in the bloodstream in the first 24 hours after brain ischemia.⁵ In acute ischemic stroke cases, IL-6 serum levels were substantially higher (average 4.0 pg/mL) compared with stroke mimics (average 1.2 pg/mL). In addition, a study assessing the NMDA receptor NR2 subunit, produced by plasma cells in reaction to peptide fragments released during cerebral ischemia–induced excitotoxicity, showed elevated levels of NR2 degradation products in people with ischemic stroke compared with those without stroke (peaking around 12 hours after onset, with the decisive threshold set at 1.0 mg/L).⁶ Penn et al assessed 16 proteins in people suspected of having TIA, 9 of which were identified as significant predictors of TIA.⁷

Biomarkers to Distinguish Stroke Type

Blood biomarkers have shown some promise in identifying stroke types. In a study of 90 individuals, Marto et al aimed to identify untargeted biomarkers in people suspected of having a stroke, using mass spectrometry to analyze blood samples from people admitted within 6 hours of symptom onset.⁸ The study categorized these participants into ischemic stroke, intracerebral hemorrhage (ICH), or stroke mimic groups (Table 1). Five proteins (complement component 3 [C3], intercellular adhesion molecule 2 [ICAM-2], plasminogen-like A [PLGLA], syntaxin-binding protein 5 [STXBP5], and immunoglobulin heavy variable 3-64 [IGHV3-64]) were noted to help differentiate among these stroke types. A model based on these biomarkers showed 88% sensitivity and 89% negative predictive value in identifying ischemic stroke as opposed to ICH, and 75% sensitivity and 95% negative predictive value when separating ischemic from combined ICH and stroke mimics. These findings introduce promising avenues for acute ischemic stroke diagnosis, but further validation in larger cohorts and varied clinical settings is needed.

MicroRNA (miRNA) products play a role in numerous biologic activities, including cell growth, apoptosis, tissue differentiation, and embryo development. Furthermore, miRNAs have been found to regulate genes in cerebrovascular diseases.⁹ A review of blood levels of miRNA in individuals with stroke reported reduced levels of miR-122, 148a, 19a, 320, and 4429, and elevated levels of miR-363 and 487b.¹⁰ These variations in miRNA levels may be caused by the effects of specific miRNA on vascular and endothelial functions. Large-scale validation studies are needed to determine the diagnostic value of these associations.

While the search for a definitive biomarker for stroke diagnosis continues, recent studies, including the Triage Stroke Panel evaluation, revealed the underlying complexities of establishing a causal or phenomenologic association. The Triage Stroke Panel tested 4 blood biomarkers (NT-proBNP, matrix metalloproteinase 9 [MMP-9], D-dimer, and S100 calcium-binding protein B [S100B]), all previously associated with cerebral ischemia, and combined them into a multimarker index.¹¹ However, its accuracy in diagnosing all stroke cases, including ischemic strokes and ICH, was insufficient. The panel showed some efficacy in identifying people with TIA within the first 3 hours of symptom onset (Table 1), but its effectiveness significantly diminished beyond a certain timeframe.

Prognostic Biomarkers

Several clinical scales have been created and validated to predict outcomes in people with stroke, including the National Institutes of Health Stroke Scale. The identification of robust, accessible, and inexpensive serum markers may serve to increase the prognostic accuracy of these clinical scores.¹² These biomarkers not only can inform and support decision making in acute treatments, but also serve as etiologic markers, and inform secondary prevention to improve functional independence after stroke. In Table 2, we list the potential biomarkers and their association with prognosis or stroke type.

Cardiac Biomarkers

Several studies have evaluated the potential benefit of using biomarkers to determine the ischemic subtype or mechanism. For example, serum NT-proBNP, troponin, and structural measures have been studied previously in the acute phase of stroke care and have suggested a cardiogenic etiology in some people with stroke and identified those with a higher risk of stroke recurrence. In the Warfarin–Aspirin Recurrent Stroke Study (WARSS) randomized trial comparing aspirin with warfarin for recurrent stroke prevention in people with presumed noncardioembolic ischemic stroke, people with elevated serum NT-proBNP levels above the 95th percentile demonstrated a significant reduction in recurrent stroke or death risk with anticoagulation compared with those with NT-proBNP levels ≤750 pg/mL.^13,14 These findings suggest that elevated serum NT-proBNP level may be associated with a cardioembolic mechanism of stroke and may identify people that may benefit from anticoagulation for secondary stroke prevention.

Similarly, the Finding Atrial Fibrillation in Stroke trial (Find-AF) observed that elevated B-type natriuretic peptide (BNP) levels can help identify people who have a higher likelihood of cardiogenic etiology and would benefit from prolonged cardiac monitoring for detection. Higher proBNP and atrial natriuretic factor levels have been associated with large atria, which in turn are seen in people with atrial fibrillation, which is a risk factor for stroke.¹⁵ Elevated serum cardiac-specific troponin levels have been found to be potentially useful in identifying people who would benefit from prolonged cardiac monitoring,¹⁶ but there is no evidence to support anticoagulation therapy in these individuals preemptively. A previous meta-analysis has also suggested that elevated serum cardiac troponin levels may be associated with a higher stroke mortality risk¹⁷; however, a recent small-scale study suggests that higher stroke severity may be associated with higher serum troponin levels, in turn causing higher stroke mortality risk.¹⁸ In another study, 19 participants with elevated troponin levels after acute ischemic stroke and endovascular therapy (EVT) were found to have a higher risk of poor outcome and death at 3 months compared with participants with normal troponin levels. Overall, evidence on use of cardiac biomarkers to change clinical practice is limited.

Inflammatory Biomarkers

Inflammatory biomarkers, including C-reactive protein (CRP) and other proinflammatory cytokines, have been studied for their potential in identifying stroke type.^20,21 In a meta-analysis by Piccardi et al, CRP levels were significantly higher in people with cardioembolic stroke.²² In another study, CRP was associated with a higher risk of recurrence after lacunar stroke but did not inform the response to dual antiplatelet therapy.²³ The evidence surrounding other proinflammatory cytokines in stratifying people by stroke etiology, such as IL-6 or interleukin-1B has been contradictory and inconclusive. In a study by Fornage et al, elevated interleukin-6 levels were associated with MRI-identified white matter lesions and infarcts.²⁴ These findings differed in the study by Licata et al, who found higher levels of the proinflammatory cytokines in people with cardioembolic stroke, and lower levels in those with lacunar stroke.²⁵ Conversely, elevated D-dimer levels formed during activation of the coagulation cascade have been identified in people with cardioembolic stroke in both the acute and subacute phases.²⁶ D-dimer elevations are also observed in other prothrombotic states associated with arterial and venous ischemic stroke (ie, cerebral venous thrombosis), such as malignancy, raising the possibility that these individuals would benefit from early anticoagulation.

Other Biomarkers

When investigating serum long-chain ceramide levels in people undergoing EVT, it was observed that elevated levels of this molecule immediately after EVT were associated with poor functional outcome (modified Rankin Scale [mRS] score of 3 to 6 at 3 months after stroke).²⁷ Similarly, elevated neurofilament light chain levels before and 24 hours after EVT were independent predictors of death or disability (mRS 3 to 6) at 3 months.²⁸ In addition to improving prediction of incident ischemic stroke risk in people with ischemic stroke,²⁹ elevated serum neurofilament light chain levels in people with ischemic stroke were independently associated with lower rates of functional independence at 90 days and higher risk of cardiovascular disease recurrence.³⁰

A 2022 study also noted elevated levels of lipoxin A4 (LXA4) and leukotriene B4 (LTB4) after EVT, which independently predicted unfavorable outcome at 3 months.^31,32 In a meta-analysis, higher platelet-to-lymphocyte ratio (PLR) was associated with greater 90-day mortality risk. PLR after reperfusion was also negatively associated with early neurologic recovery. In a 2023 meta-analysis assessing the prognostic role of neutrophil-to-lymphocyte ratio (NLR) after intravenous (IV) thrombolysis or EVT, elevated NLR was associated with higher risk of poor functional outcome at 3 months in people treated with IV thrombolysis or EVT alone or combination therapy.³³ Mortality risk was also increased in people with higher NLR treated with IV thrombolysis or EVT within this study. Similarly, a significant association was identified between early NLR and mRS score of 3 to 6 at 3 months after EVT.³⁴

Less is known about prognostic biomarkers in people with ICH. In people with ischemic stroke complicated by parenchymal hemorrhage treated with EVT, greater glial fibrillary acidic protein (GFAP) and ubiquitin C-terminal hydrolase L1 (UCHL1) levels were observed compared with individuals without hemorrhagic transformation, raising the possibility that GFAP and UCHL1 may predict the risk of hemorrhagic transformation after EVT and identify individuals who may benefit from increased monitoring.²⁸ A trend toward greater risk of intracranial bleeding noted on brain imaging was seen in people with higher PLR at admission in 1 meta-analysis.³² Increased symptomatic hemorrhage risk has been observed with elevated NLR in people with ischemic stroke treated with IV thrombolysis, EVT, or combination therapy.³³ Among the few systematic reviews on prognostic markers of ICH, 1 study identified 61 different serum biomarkers that were independently associated with poor motor recovery after ICH, including NLR; however, most of these markers are only available in research settings and their clinical role is not entirely clear.³⁵ A meta-analysis of 14 studies on prognostic biomarkers of early neurologic deterioration in people with ICH found that higher fibrinogen and D-dimer concentrations on admission were associated with early deterioration.³⁶

Conclusions

Although serum biomarkers have not become well-established tools in the management of stroke, they hold promise in this area. The identification of biomarkers for any disease condition depends on their intended use, whether for diagnosis or prognosis. The challenge in identifying a diagnostic biomarker for stroke to date has been the lack of specificity with current options. In contrast to treatments for other diseases, such as myocardial infarction, which are standardized regardless of the underlying etiology, stroke management, whether hyperacute or long-term, varies based on the type of stroke (ischemic vs intracerebral hemorrhage) and the stroke’s etiology (small vessel disease or cardioembolic). Hence, the search for serum biomarkers to enhance stroke care continues.

*Drs. Akarsu and Kuczynski are co–first authors.