Professional Documents
Culture Documents
Accepted Manuscript: 10.1016/j.neuint.2017.01.005
Accepted Manuscript: 10.1016/j.neuint.2017.01.005
Geelyn J.L. Ng, Amy M.L. Quek, Christine Cheung, Thiruma V. Arumugam, Raymond
C.S. Seet
PII: S0197-0186(16)30460-0
DOI: 10.1016/j.neuint.2017.01.005
Reference: NCI 3986
Please cite this article as: Ng, G.J.L., Quek, A.M.L., Cheung, C., Arumugam, T.V., Seet, R.C.S., Stroke
biomarkers in clinical practice: A critical appraisal, Neurochemistry International (2017), doi: 10.1016/
j.neuint.2017.01.005.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form.
Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
STROKE BIOMARKERS IN CLINICAL PRACTICE: A CRITICAL APPRAISAL
Geelyn J.L. Ng a,b, Amy M.L. Quek a,b, Christine Cheung c,d, Thiruma V. Arumugam e,
a
Division of Neurology, Department of Medicine, Yong Loo Lin School of Medicine,
b
Division of Neurology, Department of Medicine, National University Health System,
Singapore
c
Institute of Molecular and Cell Biology, Agency for Science, Technology and
Research, Singapore
d
Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore
e
Department of Physiology, Yong Loo Lin School of Medicine, National University of
Singapore, Singapore
Corresponding author:
Singapore 119228
1
ABSTRACT
Biomarkers provide critical mechanistic insights to key biologic processes that occur
during cerebral ischemia which, when carefully applied, can improve clinical
barrier that restricts the release of brain-specific markers into the circulation. The
pathologic and clinical aspects of ischemic stroke are described in this review, where
and brain injury are described. The potential roles of these biomarkers are further
examined under different clinical scenarios aimed at (1) averting the risk of
deterioration and malignant infarction, (3) aiding in the diagnosis of ischemic stroke
and its differentiation from other stroke mimics, (4) guiding the search for stroke
etiology, and (5) assessing stroke risk within the community. Researchers should
By 2050, more than 1.5 billion people in the world will be aged 65 years and older,
and a silent epidemic of stroke is imminent 1. Stroke represents the second leading
cause of death for people older than 60 years, the most important cause of
is a life course disease that begins insidiously with the evolution of risk factors, giving
large intra- and extracranial arteries and small vessels, as well as a result of an
interruption and severe reduction of blood flow within the cerebral circulation 4.
can result, namely the infarct core, where neuronal death occurs within a few
minutes, and a surrounding area called the penumbra which suffers from a moderate
reduction of blood flow and contains functionally impaired but semi-viable brain
tissues5. In the ischemic core region, brain cells undergo necrotic cell death
outcome. If blood flow is not restored within a relatively short time, the penumbra
blood thrombus)7. Early reperfusion can potentially salvage ischemic brain tissues
and, in clinical trials, is associated with a 5-fold increase in the likelihood of a good
functional recovery7. Upon reoxygenation, reactive oxygen species (ROS) and early
reactions take place in the cytosol and/or cellular organelles 8. The ischemic cascade
arterial recanalization and reperfusion are weighed against the dreaded risk of
and neuroimaging features, and initial treatment response. The use of biological
signatures of cerebral ischemia that takes into account the complex biology of stroke
searched medical databases such as MEDLINE, PubMed and Ovid for publications
WHAT IS A BIOMARKER?
processes that can serve a wide range of purposes such as risk stratification,
therapeutic assessment strategies, clinical trial design and drug development 10.
acceptable to the patient, is easy to interpret by clinicians, has a high sensitivity and
in preclinical and human models, and improve the design of clinical trials by
compound of interest11.
epidemiologists as this might shorten the duration of study and reduce the overall
cost of screening the efficacy of a compound of interest. Several features of a good
surrogate biomarker include a biomarker that tracks closely with changes in the
necessary to reflect the different facades of complex disease pathways and yield
from stroke patients who are receiving either antiplatelet or anticoagulation treatment
that puts them at an increased risk of bleeding following lumbar puncture. In patients
lumbar puncture may trigger cerebral herniation especially among patients with
disease, lumbar puncture can be technically difficult to perform which might result in
an unacceptably long delay to obtain an adequate sample for acute clinical decision-
in large amounts in the circulation due to the integrity of the blood-brain barrier
(BBB). By contrast, acute disruptions in the BBB following ischemic stroke can
trigger the release of brain-derived biomarkers into the circulation, thereby providing
a novel window into the pathophysiology of stroke for wider applications of brain-
interleukin-6 (IL-6) that are capable of crossing the BBB into the circulation 15.
Another class of proteins that is also involved in the inflammatory response of stroke
is the Toll-like receptors (TLRs)16. In addition, biomarkers such as D-dimer and those
that play a part in platelet function have been implicated in thrombus formation and
propagation. A major source of thrombus is the heart and large vessels such as the
aorta where failure of the cardiovascular system to deliver sufficient blood to the
brain can exacerbate stroke outcomes. Patients with disorders in cardiac rhythm
such as atrial fibrillation (AF) have been shown to harbor increased stroke risk 17.
(NT-proBNP)18 and high sensitivity troponin T (hsTNT) 19. Due to the heterogeneity
this review covers key biological processes in stroke pathogenesis and is not
intended to be exhaustive.
Oxidative damage results from an imbalance in redox state where the pro-oxidant
effects overwhelm the antioxidant capacity of brain tissues. The burden of oxidative
mellitus, cigarette smoking, advanced age and hypertension, and predisposes these
to its high consumption of oxygen, rich iron content, high concentrations of readily
peroxidizable lipids, and relatively low endogenous antioxidant capacity 21. Following
denaturation22,23. Severe cellular injury occurs, leading to cerebral infarction and, not
damage exist, the roles of two notable biomarkers of oxidative damage for stroke are
Uric acid. Uric acid, which exists almost entirely in its ionized form urate at
physiological pH, is derived from adenine- and guanine-based purine compounds 25.
The ability of uric acid to scavenge peroxynitrite, hydroxyl radical and singlet oxygen,
and chelate transition metals, suggests an antioxidant role of urate in humans25. Uric
acid is degraded in most mammals by the hepatic enzyme, urate oxidase (uricase),
uricase gene non-functional have resulted in urate levels in human plasma that are
approximately 10 times those of most other mammals 26. In vivo, uric acid is a potent
including hydroxyl radicals and superoxide 27,28 but, at higher levels, could exhibit
uric acid in ischemic stroke are largely borne out of observational and interventional
trials. In a prospective study involving 881 consecutive patients, uric acid levels were
inversely associated with the extent of neurological deficits on admission and the
thrombolysis, uric acid levels correlated with better neurological outcomes, whereas
patients with malignant middle cerebral artery infarction and symptomatic intracranial
hemorrhage have significantly lower uric acid levels31. The value of replenishing uric
acid in the latter group of patients has been investigated by trials that examined the
dual administration of uric acid and TPA during acute ischemic stroke 32. Despite
early promising findings, in the definitive URICO-ICTUS phase 2b/3 trial that
recruited 411 patients with acute ischemic stroke, the addition of uric acid to
thrombolytic therapy did not increase the proportion of patients who achieved
and are accurately measured using analytical platforms such as high performance
disease35 and among individuals with vascular risk factors such diabetes mellitus 40,
stroke42. Ward and colleagues supported these findings and similarly showed an
increase in F2-isoprostanes levels during the early course of ischemic stroke 43. Our
Kelly and colleagues observed that early F 2-isoprostanes levels within 8 hours
following the onset of ischemic stroke correlated with MMP-9 levels 44. Taken
together, these data suggest that F 2-isoprostanes are sensitive markers of oxidative
damage in human stroke that are elevated early and remain elevated for several
days.
Biomarkers of inflammation
In the intact brain, trafficking of cellular and molecular components from the
injury, the tight junctions between endothelial cells of the BBB become leaky, which
allow circulatory immune cells to permeate and infiltrate the surrounding brain
mediated by cytokines that increase in the central nervous system (CNS) and the
(IL-6) is a crucial inflammatory cytokine that is released in the post-stroke setting and
and parenchyma resident cells. IL-6 has pleomorphic effects in the brain resulting in
microglia might have dual roles in acute ischemia (with both neurotoxic and
neuroprotective effects attributed to this cytokine), blood levels of IL-6 have been
repeatedly associated with poor outcome after stroke 47. IL-6 concentration in blood
has been correlated with baseline stroke severity, suggesting a plausible role as a
biomarker of acute cerebral injury 48. In the clinical setting, IL-6 has been measured
following the onset of acute ischemic stroke, consistently showing an increase in IL-6
When cerebral ischemia sets in, TLRs recognize DAMP signals that are released as
a result of tissue damage. These molecules, also known as endogenous ligands, are
able to activate TLRs and trigger an inflammatory response 16,52. TLR activation
Among members of the TLR family, TLR2, TLR4, TLR8 and TLR9 have been
closely associated with stroke events. In separate studies, the expressions of TLR2
and TLR4 were associated with poor outcomes in ischemic stroke patients 54,55.
Ischemic stroke induces the release of DAMPs, such as cellular fibronectin (c-Fn),
HSP60, and HSP70, which function as endogenous ligands for TLR2 and TLR4, and
mediated post-stroke inflammation 56, although this has only been demonstrated in
of blood flow into the brain. In a thromboembolic stroke, blood thrombus that is
formed from large arteries of the aorta or the heart can be dislodged into the cerebral
circulation. Aspirin, used alone or in combination with other antiplatelet drugs, has
been shown to confer significant benefit to patients at high risk of vascular events.
Resistance to the action of aspirin can occur and this might decrease its benefit to
mitigate against stroke risks. The definitions of aspirin resistance vary considerably
across studies depending on the type of assays used. Despite the importance of
resistance has led to a lack of consensus on the roles of platelet function testing in
stroke management.
production and fibrinolysis. In the general population, D-dimer levels vary and tend to
correlate positively with other cardiovascular risk factors 57, where the presence of D-
dimer confirms the generation of thrombin and plasmin 58. As fibrin D-dimer is
formation, a rise in fibrin D-dimer levels indicates thrombus formation. While there
are other markers for thrombus formation such as prothrombin fragments 1+2
stroke59.
Platelet Reactivity. Platelets have been shown to contribute to the pathogenesis
inhibition is important, not only in the chronic phase but also in the acute phase of
brain ischemia, to prevent platelet aggregation and early vascular recurrence 60.
Antiplatelet agents such as aspirin and clopidogrel have been shown to reduce the
risk of recurrent vascular events and death in stroke patients 61. The ability to identify
patients who are resistant to aspirin or clopidogrel early during their course of stroke
observations that certain patients do not derive protection from antiplatelet treatment.
dehydrxythrombxane B262. The value of platelet function testing is not clear in stroke
patients given the relative scarcity of data in this area. It remains unclear whether
improved outcomes. One study suggested that in stroke patients with a level of
days in those treated with aspirin plus extended release dipyridamole (Aggrenox)
compared with aspirin alone63. High uric acid levels has been shown to predict
aspirin resistance where a cut-off value of 6.45 mg/dl of uric acid predicted aspirin
resistance with 79% sensitivity and 65% 64. These findings suggest a potential
The brain depends on the heart to effectively pump and deliver sufficient blood to the
thrombus within the cardiac atrium and left atrial appendage. A widely recognized
cardiac risk factor of stroke is AF that results in a chaotic firing of electrical activities,
causing stasis of blood within the heart and, in the presence of thrombogenic factors,
been studied include the brain natriuretic peptide (BNP) and high-sensitivity troponin
T (hsTnT).
neurohormone that is produced by the heart ventricles and the brain 65,66. BNP
promotes natriuresis and diuresis in the body, acting as a vasodilator with countering
which does not have biological activity. Both of these biomarkers are present in
identical concentrations among normal individuals69. BNP has been widely studied in
myocardial necrosis that is used in the diagnosis of acute myocardial infarction and
to stratify patients according to the severity of their acute coronary symptoms 19.
S100β. S100β is a glial protein that belongs to the S100 family that regulates
intracellular calcium. The homodimeric S100β protein consists of two β subunits that
S100β participates in cellular differentiation and motility 76. S100β is protective and
trophic at low concentrations, but can also be toxic and pro-apoptotic at high levels.
When structural damage such as infarction occurs, S100β is released into the
times higher in cerebrospinal fluid than in serum. Previous studies by Aurell et al.
and Büttner et al. have demonstrated that the glial protein S100β was increased in
ischemic stroke patients who present 1 to 7 days after the insult 77,78.
blood81. NSE is released in the systemic circulation after ischemic stroke; therefore,
an increase in NSE levels in the peripheral blood can be taken as an indicator of the
studies have evaluated biomarkers to distinguish ischemic stroke from its mimics,
determine stroke etiology, predict stroke severity and outcomes (e.g. early
functional recovery and vascular events (e.g. recurrent stroke, myocardial infarction
and death)82. Several researchers have investigated biomarkers that estimate infarct
volume and have found significant associations between infarct volume and
findings to ascertain stroke severity 80. Furthermore, stroke volume itself may not be
a reliable indicator of stroke severity as a small stroke may not be any less
innocuous especially when critical or eloquent areas of the brain such as the
brainstem or language areas are affected. By contrast, large areas of injury may
presentation, could have practical value in the clinical setting (e.g. risk of
treatment, the beneficial effects of TPA to achieve arterial recanalization are often
recognition of patients at high risk for SICH is important to triage these patients into
specialized stroke units for closer monitoring and tighter control of risk factors
extracellular matrix84. TPA itself activates MMP-9 that is capable of disrupting BBB
recombinant TPA, serum MMP-9 levels >140 ng/ml predicted the development of
Cellular fibronectin (c-Fn) is another factor that has been associated with
is elevated following ischemic stroke. A study showed c-Fn levels >3.6 µg/ml predict
transformation91. In combination, PAI-1 levels >180% and TAFI levels <21.4 ng/ml
affects up to 40% of stroke patients 92,93. END occurs due to the development of
these complications, early recognition of END and admission of stroke patients into
specialized stroke units have been demonstrated to reduce the rate of poor
important role in this setting to identify these high-risk patients, especially since
ICAM-1, and MMPs have been associated with END 95-97. For example, plasma
glutamate levels of >200 µmol/l predicted END in patients with hemispheric stroke 96
and correlated with infarct growth determined using diffusion-weighted imaging 89.
alongside high-density lipoprotein (HDL) and the presence of internal carotid artery
receptor for advanced glycation end products (sRAGE) have been correlated with 3-
circulation (e.g. internal carotid and middle cerebral artery occlusion) are at risk of
intracranial pressure away from the brainstem101. In one study, plasma S100β level
>0.35 g/l predicted the development of malignant infarction at 12h with a 75%
sensitivity and 80% specificity, and at 24h with a 94% sensitivity and 83% specificity.
In another study, c-Fn and MMP-9 levels were found to be higher in patients with
malignant infarctions. c-Fn >16.6 µg/ml had a sensitivity of 90% and a specificity of
100% for the prediction of a malignant course of infarction 97. Trial data on the use of
may have a stroke or they may have one of several possible “stroke mimic”
conditions, including the aura of migraine, postictal deficits following a focal seizure,
psychogenic spell. In prospective series, stroke mimics can account for up to a third
of patients presenting with stroke-like symptoms 102. Although there may be clinical
following a seizure), these features are often absent and occur unwitnessed. As a
imaging, vascular and perfusion scans) could aid in the differentiation of stroke from
agents to either cease or slow down the detrimental cascade of events triggered by
studies have reported the ability of single and multiple panel biomarkers to
distinguish stroke from controls with high sensitivity and specificity by targeting
peptide105. Reynolds and colleagues had previously screened plasma samples from
subarachnoid hemorrhage) and from 214 healthy individuals for more than 50 serum
neurotrophic growth factor, von Willebrand factor (vWF), MMP-9 and monocyte
hours of symptom onset and 157 healthy controls 107. Twenty-six blood-based
markers with relation to the ischemic cascade were analyzed. Protein S100β, MMP-
data on 130 patients admitted with acute focal neurologic deficits within 6 hours of
symptom onset108. Forty-one patients were later on diagnosed with ischemic stroke.
The predictive model included BNP, CRP, D-dimer, MMP-9, and S100β.
Based on these encouraging results, the diagnostic accuracy of a biomarker
panel including D-dimer, BNP, MMP-9, and S100β was evaluated in a prospective
multicenter trial109. Within a 3-year period, more than 1,100 patients presenting with
symptoms suspicious for stroke were enrolled within 24 hours of symptom onset.
stroke patients and mimics. Setting the threshold of the model to the 25th percentile
unexplored.
disease and lacunar stroke) and determining the underlying cause of stroke is
paramount when planning for secondary prevention treatment. For example, patients
treatment strategies for patients with a large artery disease and lacunar stroke center
around antiplatelet treatment and strict risk factor control 110. Additionally, patients
with a symptomatic large artery disease due to carotid stenosis could benefit from
appealing as this might direct the search for stroke etiology and rationalize the need
prolonged cardiac rhythm monitoring). Both the pro-BNP and D-dimer have been
hours from symptom onset70. Conversely, RNA expression in blood has been shown
samples from 76 patients with acute ischemic stroke, a 40-gene panel was able to
distinguish strokes from a cardioembolic from large artery disease etiology with
loose clinical criteria such as age, the pattern of stroke on neuroimaging studies and
pg/ml was the only variable that independently associated with the risk of developing
119
AF .
Stroke risk. Several biomarkers have been shown to predict the risk of recurrent
recurrent stroke120,121 and high-sensitivity CRP levels have also been shown to
predict the risk of incident stroke122. High levels of oxidized low-density lipoprotein
(oxLDL), a biomarker closely linked with atherosclerosis, has recently been shown to
among patients with large artery and lacunar stroke mechanisms 123. Conversely,
recovery128.
inhibitors of nitric oxide synthase) have been associated with atherosclerosis and
stroke risk factors. A longitudinal study involving 67 ischemic stroke patients has
thromboembolism17. Using data from the ARISTOTLE and STABILITY trials, Hijazi
(Age, Biomarkers, Clinical history of stroke or transient ischemic attack [TIA]) yielded
higher C-indices than the widely used CHA 2DS2-VASc scores to guide selection of
high-risk patients for long-term anticoagulation 130. In a study of 880 patients with
non-valvular AF, CRP levels positively correlated with the CHADS 2 stroke risk
assessment scores and were higher in patients with all-cause mortality and vascular
events. However, no significant correlation was found with stroke risk. 131 Among
individuals with high cardiovascular risk, our group has shown significant elevations
cardiovascular events and all-cause mortality, and the addition of tertiles of urinary
CONCLUSIONS
cerebral infarction and could add to the armamentarium of clinical tools in stroke
management. Although the ischemic cascade has been extensively studied in animal
stroke models, few studies have examined the prognostic significance and temporal
human stroke. More efforts should focus on the standardization of study cohorts and
subjective reasoning. Given the wide stroke heterogeneity, the use of single
biomarkers may not sufficiently capture the different aspects of stroke pathogenesis;
clinical practice is challenging but can be extremely rewarding, especially when such
21,25-33,40,42,64,82,
Purine metabolism product; has both
Uric acid 93
antioxidant and pro-oxidant properties
Oxidative
damage
F2- Prostaglandin isomers, free radical 20,21,34-44,132
isoprostanes product of arachidonic acid
18,65-70,73,82,108,
Neurohormone that promotes
NT-proBNP natriuresis and diuresis in the body 109,113-115,119,130
Cardiac
function Marker of myocardial necrosis; used to
19,71-73,80,82,106,
diagnose acute myocardial infarction
hsTnT and prognostic markers of ischemic 107,109,130
stroke
Abbreviations: IL-6, interleukin-6; CRP, C-reactive protein; TLR, toll-like receptor; NT-proBNP, N-
terminal pro-brain natriuretic peptide; hsTNT, high sensitivity troponin T; NSE, neuron-specific
enolase.
Table 2. Application of stroke biomarkers under different clinical scenarios
18,65-70,73,82,108,
Neurohormone that promotes natriuresis
NT-proBNP 109,113-115,119,130
and diuresis in the body
37
18,65-70,73,82,108,
Neurohormone that promotes natriuresis
NT-proBNP 109,113-115,119,130
and diuresis in the body
Stroke Etiology
Fibrinogen degradation product that
57-59,73,82,104,108,
reflects thrombin production and
D-dimer 109,113-115
fibrinolysis; presence also confirms
generation of plasmin
128
MR-proADM A vasodilatory peptide found in plasma
CRP
An acute-phase pentameric protein 10,18-20,35,40-42,
found in the blood plasma; a type of 44-46,48,49,55,57,
pattern recognition receptors (PRR) 59,64,73,82,98,108,
involved in inflammation response and
116,120-122,131
innate immunity
18,65-70,73,82,108,
Neurohormone that promotes natriuresis
NT-proBNP 109,113-115,119,130
and diuresis in the body
F2- Marker of
myocardial
hsTnT isoprostanes
necrosis; used to
diagnose acute
myocardial infarction and
prognostic markers of ischemic
19,71-73,80,82,106,
stroke
107,109,130
Prostaglandin isomers, free radical
product of arachidonic acid
20,21,34-44,132
ACCEPTED MANUSCRIPT
Abbreviations: MMPs, matrix metalloproteinases; c-Fn, cellular fibronectin; PAI-1, plasminogen
activator inhibitor type 1; TAFI, thrombin-activated fibrolysis inhibitor; GABA, gamma-Aminobutyric
acid; TNF-α, tumor necrosis factor alpha; ICAM-1, intercellular adhesion molecule 1; CRP, C-reactive
protein; vWF, von Willebrand factor; MCP-1, monocyte chemotactic protein-1; VCAM-1, vascular cell
adhesion protein 1; NT-proBNP, N-terminal pro-brain natriuretic peptide; Lp-PLA 2, lipoprotein-
associated phospholipase A2; oxLDL, oxidized low-density lipoprotein; HDL, high-density lipoprotein;
MR-proADM, midregional proadrenomedullin; ADMA, asymmetric dimethylarginine; SDMA, symmetric
dimethylarginine; hsTNT, high sensitivity troponin T.
40
ACCEPTED MANUSCRIPT