Keywords vascular cell markers - cardiology - management of disease
Introduction
In cardiovascular disease (CVD), biomarkers (i.e., “biological markers”) could have
multiple roles in understanding the complexity of cardiovascular (CV) pathophysiology
and to offer an integrated approach to management. The Food and Drug Administration
(FDA) defines a biomarker as any measurable indicator that is potentially useful throughout
the whole spectrum of the disease process; research and development of new therapies;
diagnosis, prognosis, and monitoring progression of a disease; or response to treatment.[1 ] Biomarkers could help in daily practice as a diagnostic tool, to monitor therapy
response, to assess prognosis and as early marker of CV damage, or to stratify risk.[2 ] However, the implementation of routine strategies in a cost-effective manner is
restricted by the limited predictive value of current risk assessment models using
biomarkers.[3 ]
In recent years, the role of biomarkers in CVD is even more relevant and some have
recently been included in clinical management guideline recommendations. The aim of
this review is to discuss the recommendations in clinical guidelines of various biomarkers
and to review their usefulness in daily clinical practice.
Biomarkers in Heart Failure
Biomarkers in Heart Failure
Heart failure (HF) is associated with many different cardiac and extracardiac pathophysiological
mechanisms leading to a complex clinical syndrome with multiple phenotypes. The diagnosis
of HF can be difficult because the clinical presentation does not necessarily imply
structural cardiac involvement. It may be important to develop specific and immediate
tests to perform a rapid “rule-out” of HF in the emergency department.
Recent studies have highlighted the role of different biomarkers to assess patients
with HF, which have focused on different pathways: myocardial stress, neurohormonal
activation, inflammatory state, or remodeling. Hence, these biomarkers are not only
related to the diagnosis in the acute stress phase of HF but also to the chronic state[4 ] which could help in daily clinical practice for monitoring the response of a new
therapy or to stratify the prognosis related to future admissions. Moreover, these
markers can identify early heart transplant rejection or cardiotoxicity with antineoplastic
agents.[5 ]
The most important biomarkers in HF patients are the natriuretic peptides (NPs) and
troponin levels. Plasma levels of the NPs (B-type NP [BNP] and N-terminal fraction
of BNP [NT-proBNP]) are the result of end-diastolic stress due to an increase in volume
or pressure. BNP is a neurohormone synthesized by myocytes in response to increased
cardiac wall tension.[6 ] In the setting of volume expansion or pressure overload, the resulting wall stress
initiates the synthesis of pre-proBNP in the ventricular myocardium, mainly in the
left ventricular (LV) myocardium.[7 ] After that, the peptide is cleaved in the active BNP and in the inactive amino-terminal
NT-proBNP form.[6 ]
[8 ] However, the use of BNP to guide treatment in HF patients treated with sacubitril/valsartan
is controversial. NPs are substrates of neprilysin; hence, BNP concentrations rise
with neprilysin inhibition. The clinical utility of BNP in sacubitril/valsartan-treated
patients has been questioned, and NT-proBNP has been recommended. Myhre et al[9 ] showed that despite an initial increase in BNP after initiation of sacubitril/valsartan
in approximately 60% of patients, which may lead to clinical confusion, BNP remained
a reliable prognostic marker before and during treatment with sacubitril/valsartan.
Cardiac troponins are sensitive and specific markers of myocardial injury. Elevations
of troponin I and/or troponin T are observed in a many of HF patients without acute
coronary syndrome (ACS) due to the stress and damage of myofibrillar proteins.
NT-proBNP for Diagnosis-Making Process and Prognosis in Acute and Chronic Heart Failure
(Formal Clinical Guidelines Recommendations in [Table 1 ])
Table 1
Main current clinical guidelines recommendations for biomarkers use in heart failure
ESC (5)
AHA/ACC(9)
Heart failure
• Measurement of plasma NP level (BNP, NT-proBNP, or MR-proANP) is recommended in
all patients with acute dyspnea to role-in AHF
• Measuring NP biomarkers (NT-proBNP, or BNP) is recommended to support a clinical
diagnosis of HF
Class I, level of evidence A
Class I, level of evidence A
• In the acute phase of HF, cardiac troponin blood assessment is recommended to dismiss
an ACS
• Measurement of baseline levels of NPs biomarkers and/or cardiac troponin on admission
to the hospital is useful to establish a prognosis in acutely decompensated HF
Class I, level of evidence C
Class I, level of evidence A
• Multiple other biomarkers, including those reflecting inflammation, oxidative stress,
neurohormonal disarray and myocardial and matrix remodeling (e.g. ST2, galectin-3,
copeptin, adrenomedullin) have been investigated for their diagnostic and prognostic
value in acute HF
• In patients with chronic HF, measurement of other clinically available tests, such
as biomarkers of myocardial injury or fibrosis, may be considered for additive risk
stratification
No definite evidence to recommend
Class II b, level of evidence B
• Renal function should be considered in order to evaluate the patient's suitability
for particular HF therapies
Abbreviations: ACC, American College of Cardiology; ACS, acute coronary syndromes;
AHA, American Heart Association; BNP, B-type natriuretic peptide; ESC, European Society
of Cardiology; HF, heart failure; NP, natriuretic peptides; NT-proBNP, N-terminal
pro-B-type natriuretic peptide; sST2, soluble suppression of tumorigenicity 2.
2016 European Society of Cardiology Clinical Guidelines of HF
[5 ]
The plasma concentration of NPs can be used as an initial diagnostic test, especially
in the nonacute setting when echocardiography is not immediately available. Elevated
NPs help to establish an initial working diagnosis, identifying those who require
further cardiac investigation.
The upper limit of normality in the nonacute setting for BNP is 35 pg/mL and for NT-proBNP
it is 125 pg/mL. The presence of elevated levels of NPs (BNP > 35 pg/mL and/or NT-proBNP > 125
pg/mL) is used for the algorithm diagnosis of HF in patients with or without reduced
ejection fraction. A normal electrocardiogram (ECG) and/or plasma concentrations of
BNP < 35 pg/mL and/or NT-proBNP < 125 pg/mL make a diagnosis of HF unlikely.
Cardiac biomarkers (NPs and troponins) can be used to identify patients at higher
risk of cardiotoxicity and may be helpful in the monitoring of the use and dosage
of cardiotoxic/cytotoxic agents.
2017 ACC/AHA/Task Force on Clinical Practice Guidelines and the Heart Failure Society
of America
During a HF hospitalization, a predischarge NP levels can be useful in establishing
a postdischarge prognosis (Class IIa, Level of evidence B).
Routine determination of BNP or NT-proBNP as preventive tool in patients at risk of
HF is controversial (Class IIa).
Comment on Recommendations in the Clinical Guidelines
Clinical guidelines make an important common point about the use of NPs for the diagnosis
of acute HF and for the ruling out of HF in the emergency room. However, the evidence
for the latter recommendation is derived from clinical trials and observational registries
with moderate sample size.
In a study of 278 patients, Berger et al[10 ] demonstrated that the addition of NT-proBNP-guided therapy management in HF patients
improved mortality and adverse outcomes rates compared with only CV risk factors-guided
management. The ADHERE (Acute Decompensated Heart Failure National Registry) Registry
analyzed 48,629 patients with acute decompensated HF and analyzed BNP quartiles at
admission, and found that an elevated admission BNP levels were a significant predictor
of in-hospital mortality in acute decompensated HF with either reduced or preserved
systolic function, independent of other clinical and laboratory variables.[11 ] In the Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients
with Heart Failure (OPTIMIZE-HF) trial,[12 ] discharge BNP levels were the best 1-year predictor of mortality and/or rehospitalization
among older patients hospitalized with HF.
NPs have also demonstrated value for predicting new-onset HF and new admissions with
this condition. Brouwers et al[13 ] analyzed the role of 13 biomarkers in 8,569 HF-free participants in the Prevention
of Vascular and Renal Endstage Disease (PREVEND) study, and the best model for new-onset
HF included the combination of NT-proBNP, troponin T, and urinary albumin excretion,
especially for those with reduced ejection fraction. The authors suggest that routine
biomarker testing should be limited to the use of NPs and troponin T in patients with
increased CV risk.
The cut-off to determine abnormal levels of NPs and diagnose HF is important. Roberts
et al[14 ] conducted a systematic review and meta-analyses of two reports that involved 15,263
test results and observed that the lower recommended thresholds of 100 ng/L for BNP
and 300 ng/L for NT-proBNP, the NPs have sensitivities of 0.95 (95% confidence interval
[CI] 0.93–0.96) and 0.99 (95% CI 0.97–1.00), and negative predictive values of 0.94
(95% CI 0.90–0.96) and 0.98 (95% CI 0.89–1.00), respectively, for a diagnosis of acute
HF. Importantly, the authors did not observe a significant difference between BNP
and NT-proBNP. Although this is the main study to support the recommendations in clinical
guidelines, the cut-off proposed is for acute patients in the emergency room but this
has been generalized to all HF patients, including outpatients. Indeed, the diagnostic
thresholds vary depending on clinical characteristics due to the nature of the dynamic
process of HF and the intra-/interpatient variability in biomarkers measurements.
For example, NP levels may be disproportionally low in obese patients, potentially
leading to underdiagnosis in patients with high body mass index.[15 ]
Another important issue in HF clinical guidelines is related to the prediction of
cardiac toxicity with cancer therapies. The use of NT-proBNP levels to assess structural
cardiac involvement should be considered but it is not a substitute for cardiac evaluation
with echocardiography. For that reason, clinical guidelines only suggest the use of
biomarkers in cancer therapies with a weak recommendation.[16 ]
Troponin for Diagnosis-Making Process in Chronic and Acute Heart Failure (Formal Clinical
Guidelines Recommendations in [Table 1 ])
2016 European Society of Cardiology Clinical Guidelines of HF
[5 ]
Elevated concentrations of circulating cardiac troponins are detected in the vast
majority of patients with acute HF, often without obvious myocardial ischemia or an
acute coronary event, suggesting ongoing myocardial stress.
Comment to the Recommendations in Clinical Guidelines
In the Atherosclerosis Risk in Communities (ARIC) study, the association of troponin
T with incident coronary heart disease and subjects with high-sensitivity troponin
T (cTnT) levels in the highest category had significantly increased risk for HF (hazard
ratio [HR] 5.95; 95% CI 4.47–7.92); of note, even minimally elevated cTnT (≥0.003 g/L)
was associated with increased risk for mortality and HF (p < 0.05). Similar results regarding the troponin T levels related to mortality and
poor prognosis were observed in the ADHERE study.[11 ] Based on these results, clinical guidelines recommend routine ischemic assessment
using high-sensitivity troponins at admission to rule-out concomitant coronary disease
and as a biomarker of prognosis.
Renal Function in Complete Management of HF Patients (Formal Clinical Guidelines Recommendations
in [Table 1 ])
2016 European Society of Cardiology Clinical Guidelines of HF
[5 ]
HF and chronic kidney disease (CKD) frequently coexist, sharing many risk factors
(diabetes, hypertension, and hyperlipidemia) and interacting to worsen prognosis.
Comment to the Clinical Guidelines Recommendations
About one-third of HF patients have a concomitant mild or moderate CKD and about a
quarter develop worsening renal function during their hospitalization for HF.[17 ] The worsening of renal function over the time in patients with HF has been associated
with a reduction in survival and an increase of hospital admissions. Some of the deterioration
may be related to diuretic therapy or with use of drugs such as the angiotensin-converting
enzyme inhibitors, angiotensin receptor blockers, or aldosterone antagonist use. The
worsening of heart pump function is also related to renal dysfunction[17 ]
[18 ]
Other Biomarkers (Formal Clinical Guidelines Recommendations in [Table 1 ])
Comment to the Recommendations in Clinical Guidelines
In recent years, neurohormonal biomarkers and other biomarkers related to the inflammation
and with remodeling/fibrosis have emerged. Despite this, clinical trials and observational
registries are usually based only on a single measurement at baseline or study entry,
and extrapolating their usefulness for treatment, monitoring, or as a prognosis markers
should be taken with caution.
Neurohormonal activation has an important role in the progression and worsening of
patients with HF, for example, components of the renin angiotensin system (renin,
angiotensin II, aldosterone), sympathetic nervous system (norepinephrine, chromogranin
A, mid-regional pro-adrenomedullin), arginine vasopressin system (arginine vasopressin,
copeptin), and endothelins (ET-1, big proET-1).[5 ] In the Biomarkers in Acute Heart Failure (BACH) trial, Maisel et al[15 ] analyzed the role of copeptin and other biomarkers in 557 patients with HF. Copeptin
was highly prognostic for 90-day adverse events in patients with acute HF, adding
prognostic value to clinical predictors, serum sodium, and NPs; however, the potential
variability and the influence in plasma levels due to the different laboratory tests
and patient clinical stability limits the generalization of these results.
Related to cardiac remodeling and fibrosis, the measurement of soluble suppression
of tumorigenicity 2 (sST2) is correlated to LV hypertrophy, fibrosis, and remodeling
via interaction with interleukin (IL)-33, a protein with antifibrotic and antiremodeling.[19 ] In one study, Ky et al[20 ] determined whether plasma sST2 levels predict adverse outcomes in 1,141 patients
with chronic HF and observed those with the highest sST2 tertile (sST2 > 36.3 ng/mL)
had a markedly increased risk of adverse outcomes compared with the lowest tertile
(sST2 ≤22.3 ng/mL), even after multivariable adjustment (HR 1.9; 95% CI, 1.3–2.9;
p = 0.002).
Galectin-3 (Gal-3) is another important biomarker related to cardiac fibrosis. In
the DEAL-HF study, Lok et al[21 ] analyzed the role of Gal-3 levels in cardiac remodeling in HF and concluded that
this biomarker was a significant predictor of mortality risk, even after adjustment
for age and sex. Indeed, its prognostic value was independent of severity of HF and
renal dysfunction, as assessed by NT-proBNP and estimated glomerular filtration rate
(eGFR), respectively.
The use of these potential novel biomarkers for daily clinical practice is controversial.
One limitation is the small sample size of many (often heterogeneous) studies, as
well as the potential variability of the biomarker over time and the status of the
HF patient at the moment of the measurement. Indeed, the predictive ability that offers
the inclusion of biomarkers for prognosis of HF is modest. Also, the exclusion of
the influence of other biomarkers that could act as confounding factors is not always
controlled. For that reason, recommendations for routine use of other nonstandardized
biomarkers (copeptin, sST2, or Gal-3) are weak and limited to specific clinical situations.
For example, increased Gal-3 was a proportional predictor of CV death and all-cause
mortality, also after adjustment for NT-proBNP, in subjects without previously diagnosed
CVD.[22 ]
Biomarkers in Atrial Fibrillation
Biomarkers in Atrial Fibrillation
Atrial fibrillation (AF) is associated with high morbidity and mortality, largely
attributable to an increased risk of stroke and thromboembolism. Oral anticoagulation
treatment is highly effective in reducing the risk of stroke by 64% and all-cause
mortality by 26% in comparison with placebo or control.[23 ]
Biomarkers such as markers of inflammation, coagulation activity, CV stress, myocardial
injury, and cardiac and renal dysfunction have all shown high association with clinical
events and have been proposed to refine risk assessment in AF patients.[24 ] Cardiac biomarkers, such as troponin and NPs, also significantly improve risk stratification
in addition to current clinical risk stratification models.
The use of biomarkers in AF helps in the understanding of the pathophysiology of this
prevalent disease and could refine stroke and major bleeding risk in AF patients.[25 ] Indeed, most of the clinical guidelines recommendations are only weak, based on
current evidence and expert consensus documents. Any biomarker, whether blood, urine,
or imaging (cardiac, cerebral, or otherwise), will always improve on risk prediction
based on clinical factors, but this needs to be balanced against the practical usefulness,
cost, and daily applicability for everyday clinical practice.[26 ]
Pathophysiology of Atrial Fibrillation
2016 European Society of Cardiology Clinical Guidelines of Atrial Fibrillation
[27 ]
Activation of fibroblasts, enhanced connective tissue deposition, and fibrosis are
the hallmarks of structural remodeling in the atria. Atrial fatty infiltration, inflammatory
infiltrates, myocyte hypertrophy, necrosis, and amyloidosis are found in AF patients
with concomitant conditions predisposing to AF.
Although biomarkers such as NPs are elevated in AF patients, there is insufficient
data to suggest that blood-based parameters are independent markers for AF.
2014 AHA/ACC/HRS Guideline for the Management of Patients with Atrial Fibrillation
[28 ]
Antithrombotic Therapy for Atrial Fibrillation: American College of Chest Physicians
Guidelines and Expert Panel Report- ACCP 2018
[26 ]
Comment on Recommendations in the Clinical Guidelines
Different biomarkers have been proposed to predict patients at risk of AF development.
An inflammatory state is related to AF burden and atrial remodeling. IL-6 and C-reactive
protein (CRP) are robust and established markers of inflammation and have been most
frequently investigated in CVD and AF.
Aviles et al[29 ] analyzed the influence of CRP levels in 5,806 patients enrolled in the Cardiovascular
Health Study with 7 years of follow-up, and concluded that elevated baseline CRP levels
predicted an increased risk for developing future AF. Similarly, Patton et al[30 ] analyzed the role of NT-proBNP in the same registry and observed that this biomarker
was also an independent predictor of AF, even after adjustment for an extensive number
of covariates; also NT-proBNP levels were also strongly associated with prevalent
AF. Recently, Chua et al[31 ] showed in 648 patients that elevated levels of BNP (odds ratio [OR] 1.29, 95% CI
1.11–1.63; p = 0.002) and elevated levels of fibroblast growth factor-23 (OR 1.67, 95% CI 1.36–2.34;
p = 0.001) were significantly associated with prevalent AF.
However, many of the published studies generally analyzed the effect of one (bio)marker
in the time and the pathophysiology of AF development is a continuum.
Stroke Risk Assessment in Atrial Fibrillation Patients (Formal Clinical Guidelines
Recommendations in [Table 2 ])
Table 2
Main current clinical guidelines recommendations for biomarkers use in atrial fibrillation
ESC (27)
AHA/ACC (64)
American College of Chest Physicians Guidelines-2018 ACCP (26)
Atrial fibrillation
• Measurement of cardiac troponin and NT-proBNP may provide additional prognostic
information in selected AF patients.
• Biomarkers such as Hs-troponin and NPs may be considered to refine stroke and bleeding
risk
Class IIb, level of evidence B
• Several bleeding risk scores have been developed, mainly in patients on VKAs. These
include HAS-BLED, ORBIT, and more recently, the ABC bleeding score, which also makes
use of selected biomarkers
• The assessment of kidney function by serum creatinine or creatinine clearance is
recommended in all AF patients
Class I, level of evidence A
• Renal function and hepatic function should be evaluated before initiation of a NOAC
and should be reevaluated at least annually
Class I, level of evidence: B
• Many biomarkers are also predictive of stroke, bleeding, death, heart failure, and
hospitalization in AF patients
• Current studies do not inform whether biomarkers will discriminate/identify low
risk in lower/intermediate risk patients who are not anticoagulated
• The addition of biomarkers to bleeding risk scores would all improve the c-indexes
(to approx. 0.65) over scores based on clinical risk factors alone but marginally.
The performance of biomarker-based scores in real world clinical practice (outside
highly selected trial cohorts) has also been disappointing
• CKD is an independent predictor of risk of stroke/thromboembolism. The clinical
relevance of renal function is not only for risk prediction, but also for choice of
anticoagulation and other therapies
Abbreviations: ACC, American College of Cardiology; AF, atrial fibrillation; AHA,
American Heart Association; CKD, chronic kidney disease; ESC, European Society of
Cardiology; NT-proBNP, N-terminal pro-B-type natriuretic peptide; VKAs, vitamin K
antagonists.
2016 European Society of Cardiology Clinical Guidelines of Atrial Fibrillation
[27 ]
Biomarker-based risk scores may, in the future, be helpful to better stratify patients
(e.g., those at a truly low risk of stroke).
Antithrombotic Therapy for Atrial Fibrillation: American College of Chest Physicians
Guidelines and Expert Panel Report-ACCP 2018
[26 ]
The importance of biomarkers probably lies in the “very low risk” strata of clinical
scores (e.g., CHA2 DS2 -VASc = 0–1 group) where they may influence the decision to anticoagulate. For these
reasons, the clinical application of biomarkers in management of AF is unlikely to
be significant.
Current studies do not inform whether biomarkers will discriminate/identify low risk
in lower/intermediate risk patients who are not anticoagulated.
Comment to the Clinical Guidelines Recommendations
Stroke and thromboembolic risk in AF patients is assessed using CHA2 DS2 -VASc risk score.[26 ]
[32 ]
[33 ] This score has only clinical and demographic variables and has modest ability to
predict thromboembolic events and mortality.[34 ]
[35 ] Biomarkers were first proposed to refine clinical risk stratification > 10 years
ago.[36 ] Lip et al[36 ] analyzed the role of von Willebrand factor (vWF) (as plasma biomarker of endothelial
damage/dysfunction and associated with stroke risk) to refine stroke prevention in
AF patients. The authors observed that the addition of plasma vWF levels refined stroke
risk stratification in CHADS2 score, particularly in AF patients at moderate risk with the corresponding c-indexes
for ischemic stroke and vascular events of 0.691 (95% CI, 0.600–0.772) and 0.697 (95%
CI, 0.626–0.763), respectively. Indeed, high vWF levels were independently associated
with a higher risk of vascular events.
In recent years, there has been renewed interest in these biomarkers which have been
part of ancillary analyses to large randomized controlled trials. These biomarkers
are related to myocardial stress (troponin or NT-proBNP), renal function (cystatin
C), the prothrombotic state in AF patients (D-dimer, vWF, soluble thrombomodulin,
or soluble E-selectin), or the inflammatory state (CRP or IL-6).[37 ] Recently, other oxidative stress markers have also been related with adverse events
in AF patients. Oxidative stress has been shown to promote and maintain AF with a
mechanism related to upregulation of myeloperoxidase (MPO) and NOX2. Indeed, oxidative
stress is implicated in clotting activation.[38 ] The urinary excretion of F2-IsoP (a particular type of prostaglandin) have been
related with thrombosis and systemic atherosclerosis.[39 ] Pignatelli et al[40 ] analyzed the effect of oxidative biomarkers (serum NOX2 and urinary isoprostanes),
in AF patients and observed that high levels of these biomarkers were related with
adverse events and mortality. The addition of tertiles of urinary isoprostanes significantly
increased the predictive performance and net reclassification index of CHA2 DS2 -VASc for CV events and mortality.
The use of biomarkers (whether urine, blood, or imaging) adds statistical improvement
in the predictive performance of risk scores compared to schemes using clinical variables,
approximately 0.65 to 0.70. The use of biomarkers should balance the assay availability,
laboratory variability, costs and added complexity, and lower practicality for everyday
use. Also, many biomarker studies are based on anticoagulated highly selected clinical
trial cohorts.[26 ] For that reason, the minimal improvement and complexity of use makes the applicability
in daily clinical practice doubtful.
Related to myocardial damage, minor elevations of troponin below the 99th percentile
of the upper reference limit (URL) (to diagnosis of myocardial infarction [MI]) have
been observed in patients with AF, in most cases due to rapid and irregular ventricular
response in AF patients. Van den Bos et al[41 ] conducted a prospective study with consecutive AF patients, excluding patients with
MI, and observed that minor elevations in troponin I during hospital admission had
an independent association with mortality (HR 2.35; 95% CI 1.17–4.73] and major adverse
cardiovascular events (MACEs) (HR 2.48; 95% CI 1.33–4.63).
Roldán et al[42 ] analyzed the role of cTnT in 930 AF patients with stable oral anticoagulation. In
this study, median (interquartile range) values of hsTnT were 8.86 (4.24–15.21) pg/mL,
and 291 (31%) patients had hsTnT levels above 13 pg/mL, that is the 99th percentile
of hsTnT and the cut-point proposed for the diagnosis of MI. The authors found that
high levels of troponin (above 8.04 pg/mL) were significantly associated with long-term
adverse CV events, even after adjusting for the CHADS2 score (HR 2.21; 95% CI 1.46–3.35, p < 0.001).
Abnormal levels of troponins are related not only with mortality and MI in AF patients
but also to stroke and thromboembolic risk. The role of cardiac biomarkers was also
analyzed in a biomarker substudy of 6,189 AF patients from Randomized Evaluation of
Long Term Anticoagulant Therapy (RE-LY),[43 ] where troponin I was significantly and independently associated with increased risk
of stroke or systemic embolism. The annual rates of stroke or systemic embolism were
lowest, being 0.84%, in the group with undetectable troponin I, which is in comparison
with 2.09% (HR, 1.99; 95% CI, 1.17–3.39) in the highest troponin I group (p = 0.0040). The addition of troponin I and NT-proBNP significantly improved the predictive
performance of CHADS2 for stroke outcomes. Note that the improved c-statistics remained < 0.7 despite the
addition of two biomarkers, in this highly selected anticoagulated clinical trial
cohort. In a similar study, the Apixaban for the Prevention of Stroke in Subjects
with Atrial Fibrillation (ARISTOTLE) troponin substudy analyzed 14,892 AF patients
and also observed that troponin T levels were independently associated with an increased
risk of stroke, cardiac death, and major bleeding and improved risk stratification
beyond the CHA2 DS2 -VASc risk score.[44 ]
Other important biomarkers analyzed are NPs. The relationship between NPs levels and
adverse outcomes in AF patients was assessed in the RE-LY trial.[43 ] NT-proBNP levels were positively correlated with the risk of thromboembolic events
and CV mortality, with higher risk at rising levels. Indeed, the addition of NT-proBNP
to the CHADS2 and CHA2 DS2 -VASc risk stratification models resulted in significant improvements in the discrimination
performance for both outcomes as well. A “real-world” study conducted by Roldán et
al[45 ] observed that NT-proBNP provided complementary prognostic information to an established
clinical risk score (CHA2 DS2 -VASc) for the prediction of stroke/systemic embolism and all-cause mortality.
Based on adding biomarkers to ischemic risk scores in AF patients, a new thromboembolic
risk score has emerged, the ABC (age, cardiac biomarkers [NT-proBNP, high-sensitivity
cardiac troponin], and clinical history [prior stroke/transient ischemic attack])-stroke
risk score. The ABC-stroke score was derived and validated in two clinical trial cohorts
(14,701 AF patients) in which patients with AF are often highly selected and carefully
followed-up.[46 ] Compared with the widely used clinically based CHA2 DS2 -VASc score, the ABC-stroke risk score had better predictive performance for adverse
events. However, the results of the ABC score in several independent real-world cohorts
have been disappointing.[47 ]
[48 ] For example, Rivera-Caravaca et al[47 ] validated the ABC-stroke in AF patients under stable anticoagulation with acenocoumarol,
over a long-term period of follow-up (6.5 years); they found that the ABC-stroke score
did not offer significantly better predictive performance compared with the CHA2 DS2 -VASc score.
All these clinical studies analyzed the potential role of biomarkers for stroke prediction
in AF patients and could help to better understand the different mechanisms involved
in thromboembolism events in AF patients and guide the therapy in some selected patients.
However, the recent real-world studies have only shown a slight improvement in the
predictive performance with the use of biomarkers (c-statistics generally remain < 0.7
despite biomarkers), and waiting for biomarker test results could delay the initiation
of oral anticoagulation therapy ([Table 3 ]).
Table 3
Main clinical studies about biomarkers in atrial fibrillation
Study
Year
N
Population
Biomarker
c-Statistic (AUC 95% CI)
p for comparison
Commentary
Lip et al[36 ]
2006
994
Clinical trial
vWF
** Stroke
Birm score = 0.642 (0.564–0.713)
Birm + vWF = 0.679 (0.591–0.756)
CHADS2 score = 0.672 (0.582–0.754)
CHADS2 score + vWF= 0.691 (0.600–0.772)
** Vascular events
Birm score = 0.670 (0.603–0.726)
Birm + vWF = 0.716 (0.643–0.779)
CHADS2 score = 0.672 (0.605–0.737)
CHADS2 score + vWF = 0.697 (0.626–0.763)
< 0.050
< 0.050
< 0.050
< 0.050
SPAF III trial. Plasma vWF levels refined clinical risk stratification for stroke
and vascular events
García-Fernández et al[37 ]
2017
1,215
Real-world
vWF
**Stroke
CHA2 DS2 -VASc = 0.610 (0.582–0.637)
CHA2 DS2 -VASc + vWF = 0.623 (0.595–0.650)
**Cardiovascular mortality
CHA2 DS2 -VASc = 0.656 (0.628–0.683)
CHA2 DS2 -VASc + vWF = 0.670 (0.643–0.697)
0.131
0.156
Addition of vWF to CHA2 DS2 -VASc statistically improved its predictive value, but c-indexes were not significantly
different
Hijazi et al[43 ]
2012
6,189
Clinical trial
Troponin I
** Stroke
CHA2 DS2 -VASc = 0.618[a ]
CHA2 DS2 -VASc + troponin I = 0.647[a ]
0.049
RE-LY clinical trial. Cardiac biomarkers seem useful for improving risk prediction
in AF beyond currently used clinical variables
NT-proBNP
** Stroke
CHA2 DS2 -VASc = 0.618[a ]
CHA2 DS2 -VASc + NT-proBNP = 0.633[a ]
0.239
Troponin+ NT-proBNP
** Stroke
CHA2 DS2 -VASc = 0.618[a ]
CHA2 DS2 -VASc + NT-proBNP = 0.654[a ]
0.027
Hijazi et al[44 ]
2013
14,897
Clinical trial
Troponin T
** Stroke
CHA2 DS2 -VASc = 0.620 (0.592–0.647)
CHA2 DS2 -VASc + TnT = 0.635 (0.609–0.661)
**Cardiac mortality
CHA2 DS2 -VASc = 0.592 (0.567–0.616)
CHA2 DS2 -VASc + TnT = 0.699 (0.678–0.719)
0.001
0.001
Addition of hs-TnT
levels
improves risk stratification beyond the CHA2 DS2 -VASc risk score
Roldán et al[45 ]
2012
1,172
Real-world
NT-proBNP
** Stroke
CHA2 DS2 -VASc = 0.620 (0.590–0.650)
CHA2 DS2 -VASc + NT-proBNP = 0.680 (0.560–0.710)
** All-cause mortality
CHA2 DS2 -VASc = 0.660 (0.640–0.690)
CHA2 DS2 -VASc + NT-proBNP = 0.680 (0.650–0.710)
0.069
0.178
NT-proBNP provided complementary prognostic information to an established clinical
risk score (CHA2 DS2 –VASc)
Pignatelli et al[40 ]
2015
1,002
Real-world
Urinary F2-Iso P
Nox
** Cardiovascular events
CHA2 DS2 -VASc = 0.670 (0.610–0.740)
CHA2 DS2 -VASc + Oxid = 0.720 (0.670–0.780)
** Cardiovascular deaths
CHA2 DS2 -VASc = 0.680 (0.590–0.760)
CHA2 DS2 -VASc + Oxid = 0.720 (0.650–0.800)
** All-cause mortality
CHA2 DS2 -VASc = 0.650 (0.570–0.730)
CHA2 DS2 -VASc + Oxid = 0.690 (0.630–0.760)
0.0005
0.0090
0.0087
In AF patients, 8-iso-PGF2a and NOX2 levels are predictive of cardiovascular events
and total mortality
The addition of these oxidative markers increases the predictive performance of CHA2 DS2 -VASc for cardiovascular events and mortality
Hijazi et al[46 ]
2016
14,701
Clinical trial
Troponin I
NT-proBNP
(ABC-stroke)
* Stroke
ABC-Stroke 0.660 (0.580–0.740)
CHA2DS2-VASc 0.580 (0.490–0.670)
< 0.001
The ABC-stroke score
performed better than the presently used clinically based risk score (CHA2 DS2 -VASc) and may provide improved decision support
Rivera-Caravaca et al[47 ]
2017
1,125
Real-world
Troponin I
NT-proBNP
(ABC-stroke)
* Stroke
ABC-Stroke 0.662 (0.633–0.690)
CHA2DS2-VASc 0.620 (0.590–0.648)
0.116
In long-term period, ABC-stroke score does not offer
significantly better predictive performance compared with CHA2 DS2 -VASc score
Hijazi et al[51 ]
2016
14,547
Clinical trial
GDF-15, hsTnT, Hb
(ABC-bleeding)
* Major Bleeding
ABC-Bleeding 0.710 (0.680–0.730)
HAS-BLED 0.620 (0.590–0.640)
< 0.001
ABC-bleeding score performed better than HAS-BLED and ORBIT scores and should be useful
as decision support on anticoagulation treatment
Esteve-Pastor et al[48 ]
2017
1,120
Real-world
GDF-15, hsTnT, Hb
(ABC-bleeding)
* Major Bleeding
ABC-Bleeding 0.518 (0.680–0.730)
HAS-BLED 0.583 (0.554–0.612)
0.025
In “real-world” validation of the ABC-bleeding score,
HAS-BLED performed significantly better than the ABC-bleeding score
Abbreviations: AF, atrial fibrillation; AUC, area under the curve; Birm, Birmingham
score; CI, confidence interval; GDF-15, growth differentiation factor 15; Hb, hemoglobin;
hsTn, high sensitive troponin; Oxid, oxidative; vWF, von Willebrand factor.
a
p < 0.05.
Bleeding Risk Assessment in Atrial Fibrillation Patients (Formal Clinical Guidelines
Recommendations in [Table 2 ])
Antithrombotic Therapy for Atrial Fibrillation: American College of Chest Physicians
Guidelines and Expert Panel Report. ACCP 2018
[26 ]
Many biomarkers are nonspecific for a particular endpoint, and can be equally predictive
not only for stroke but bleeding, death, hospitalization, HF, etc., as well as noncardiac
conditions, for example, glaucoma (growth differentiation factor-15 [GDF-15]).
The performance of biomarker-based scores in real-world clinical practice (outside
highly selected trial cohorts) has also been disappointing, given that baseline (or
near-baseline) determination of biomarkers to predict bleeding risks after many years
is bedeviled by the changing clinical risk profile of patient's risks as well as modification
of risk factors.
Comment to the Clinical Guidelines Recommendations
In recent years, different bleeding risk scores have been proposed, some quite complex,
to assess hemorrhagic risk in AF patients. The HAS-BLED bleeding score has been proposed
as the main clinical score to assess major bleeding events, and is supported by a
systematic review from the Patient-Centered Outcomes Research Institute (PCORI) where
38 studies explored bleeding risk scores and the HAS-BLED score provides the best
prediction for bleeding risk and intracranial hemorrhage.
The HAS-BLED score has been validated in AF patients with and without anticoagulation
therapy and it is the only bleeding score for intracranial hemorrhage prediction.[49 ]
[50 ] The most common bleeding risk scores (HAEMORR2 HAGES, HAS-BLED ATRIA, ORBIT) mainly include risk factors and clinical variables that
provide only modest predictive value for predicting patients at high risk of bleeding,
and generally underperform compared to HAS-BLED in vitamin K antagonist (VKA)-treated
patients as they neglect to consider quality of anticoagulation control.[48 ]
[51 ]
[52 ] Indeed, these bleeding risk scores have been validated also in patients under venous
thromboembolism.[53 ]
Different biomarkers emerged to explain the pathophysiology of bleeding events in
AF patients and have been used to refine bleeding risk prediction in AF patients.
Roldán et al[54 ] reported the use of vWF levels to predict bleeding events and observed that high
plasma vWF levels (≥221 IU/dL) are an independent risk factor for major bleeding and
mortality in anticoagulated AF patients; but this biomarker was also predictive of
stroke.
In a substudy of the RE-LY trial, Hijazi et al analyzed the role of troponin T and
NT-proBNP for bleeding risk assessment[43 ] and reported an association between elevated troponin I levels and the risk of major
bleeding. However, there was no significant association between NT-proBNP levels and
major bleeding.[43 ]
[55 ]
Recently, a new biomarker related to bleeding in anticoagulated AF patients has been
proposed, GDF-15. This biomarker is a marker of oxidative stress and inflammation,
providing independent prognosis information on CV events beyond CV risk factors and
other biomarkers. Wallentin et al[56 ] analyzed the role of GDF-15 in a substudy of the ARISTOTLE clinical trial and observed
that this biomarker was an independent biomarker for major bleeding in AF patients.
Higher GDF-15 levels were significantly associated with a 3.5-fold higher rate of
major bleeding and the addition of GDF-15 to the HAS-BLED and ORBIT bleeding scores
significantly improved the predictive performance of the clinical factor-based scores
for major bleeding.[57 ]
However, GDF-15 is not a specific biomarker of bleeding risk and several conditions
could modify the plasma levels of this biomarker, being an oxidative stress marker.
For example, Sharma et al[58 ] analyzed the role of GDF-15 levels in HF patients enrolled in the HF-ACTION (Heart
Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trial. The
authors observed that high concentration of GDF-15 was related with a 30% greater
risk of mortality and with worse symptom burden and low functional capacity in stable
HF patients. However, GDF-15 biomarker is not only related to adverse outcomes in
CVD patients but also seems predictive of non-CVD conditions, such as glaucoma progression.[59 ] Indeed, GDF-15 acts as a molecular marker that predicts glaucomatous neurodegeneration,
and elevated levels of GDF-15 were significantly associated with worse functional
outcomes in glaucoma patients.
Based on these biomarkers tested in AF, the ABC-bleeding score Age [A], Biomarkers
(cTn-hs, hemoglobin, GDF-15 or creatinine clearance [CrCl]) [B], and clinical history
(previous bleeding) [C] was developed from the ARISTOTLE clinical trial and was externally
validated in the RE-LY trial.[51 ]
[57 ]
[60 ] The ABC-bleeding score performed better than HAS-BLED score and was proposed as
useful tool to support decision-making for anticoagulation treatment in AF patients.[51 ] The original validation of ABC-bleeding score included GFD-15, cTn-hs, and hemoglobin,
but Hijazi et al also proposed a modified ABC-bleeding score using alterative biomarkers
(cTn-hs, hemoglobin, and CrCl) instead of GDF-15; however, the use of GDF-15 has been
only validated in anticoagulated patients (with non-VKAs oral anticoagulants [NOACs]
or VKAs) but there are no data available in AF patients without anticoagulation.
An independent real-world external validation analyzed the predictive performance
of ABC-bleeding in a nonselected AF population under VKAs.[48 ] Using sensitivity, reclassification, and decision curve analyses, the authors concluded
that the HAS-BLED score performed significantly better than the ABC-bleeding score
in predicting major bleeding, gastrointestinal bleeding, and intracranial hemorrhage.
These results highlight the difference between real-world and clinical trials, since
patients from clinical trials are carefully selected with a close follow-up. Indeed,
real-world patients are older, with multiple comorbidities, and polypharmacy that
could influence the levels and variability of blood biomarkers.
In this sense, a recent study analyzed the adverse events between real-world AF cohort
and the AMADEUS clinical trial cohort and concluded that there was great heterogeneity
in both populations, which is translated into a higher risk of several adverse outcomes
in a real-world cohort, including major bleeding, ischemic stroke, and mortality.[61 ]
Biomarkers might be helpful in assessing the risk of major bleeding, providing information
in complex cases to select the best net clinical benefit therapy for the patients.
Nevertheless, the addition of biomarkers to ischemic and bleeding risk scores only
modestly improves their predictive performance, with no clinical difference when compared
to clinical risk scores.[32 ] Indeed, bleeding risk assessment is a dynamic process that needs regular evaluation
and review.[62 ] A dynamic and practical score is mandatory with modifiable variables to highlight
patients potentially at high risk of bleeding to address the reversible bleeding risk
factors.
The biological variability of blood or urine biomarkers has only had limited attention
in the clinical trials. Some biomarkers have very large diurnal variations and, for
that reason, may not be useful as single time-point measurement for patients in the
anticoagulation clinics or in the ward to decide oral anticoagulation process. Indeed,
a range of biomarker levels between normal controls and sick patients is mandatory
but depends on age, sex, and other related comorbidities.[63 ] Nowadays, it is still very difficult to find a biomarker with little variability
throughout the day, with high quality and precision to select healthy and sick patients
and high specificity for each pathology. Also in this way, a bedside simple bleeding
score for daily clinical practice is necessary and the inclusion of routine biomarkers
in risk scores implies losing simplicity and practicality. For that reason, current
clinical guidelines only provide weak recommendation to the role of biomarkers for
stroke and bleeding risk assessment in AF patients.
Renal Function in Complete Management of AF Patients (Formal Clinical Guidelines Recommendations
in [Table 2 ])
2016 European Society of Cardiology Clinical Guidelines of Atrial Fibrillation
[27 ]
AF is present in 15 to 20% of patients with CKD. The definition of CKD in most AF
trials is relatively strict. Although an estimated CrCl rate of 60 mL/min is indicative
of CKD, a number of trials in AF patients have used CrCl 50 mL/min to adapt NOAC dosage,
usually estimated using the Cockroft–Gault formula. CrCl in AF patients can deteriorate
over time.
All AF patients treated with oral anticoagulation should be considered for at least
yearly renal function evaluation to detect CKD (Class IIa, Level of Evidence B).
2019 AHA/ACC/HRS Focused Update of the 2014 AHA/ACC/HRS Guideline for the Management
of Patients with Atrial Fibrillation
[64 ]
Renal function and hepatic function should be evaluated before initiation of a NOAC
and should be reevaluated at least annually.
All 4 NOACs with FDA approval for use in patients with AF have dosing defined by renal
function (creatinine or CrCl using the Cockcroft–Gault equation).
Renal function should be regularly monitored and CrCl calculated at an interval that
depends on the individual degree of renal dysfunction and likelihood of fluctuation,
and dose adjustments should be made according to the FDA dosing guidelines.
Antithrombotic Therapy for Atrial Fibrillation: American College of Chest Physicians
Guidelines and Expert Panel Report ACCP 2018
[26 ]
AF patients with estimated glomerular filtration rate < 60 mL/min compared to those
with estimated glomerular filtration rate ≥60 mL/min have increased risk of stroke/thromboembolism
(relative risk [RR] 1.62, 95% CI, 1.40–1.87; p < 0.001), with a 0.41% (0.17–0.65%) annual rate increase for every 10 mL/min decrease
in renal function. The risk is higher in individuals requiring renal replacement therapy
(HR 1.83; 95% CI, 1.57–2.14; p < 0.001).
Comment on Recommendations in the Clinical Guidelines
CKD is highly prevalent in AF patients and both renal dysfunction and AF are independently
associated with a higher risk of mortality and thromboembolic and bleeding events.
A recent study from the prospective FANTASIIA registry[65 ] observed that renal function was independently related to CV mortality, major bleeding,
and MACE but not to thromboembolic events. Patients with severe CKD have a fivefold
risk of major bleeding compared to those with normal renal function and renal dysfunction
is one of the components of the HAS-BLED score.
Another study published by Roldán et al[66 ] showed that the addition of CKD to stroke risk scores (CHADS2 and CHA2 DS2 -VASc) did not independently improve the predictive ability of current clinical scores.
Bonde et al[67 ] demonstrated in 154,259 AF patients that CKD is associated with a higher risk of
stroke/thromboembolism across stroke risk strata in AF patients. Of note, high-risk
patients (CHA2 DS2 -VASc ≥2) with CKD had benefit with warfarin treatment for stroke prevention. In the
same way, the use of NOACs was associated with a reduced risk of stroke and major
bleeding compared to warfarin in patients with renal disease, with positive net clinical
benefit.[68 ] Dosing recommendations for patients with renal impairment differ depending on the
NOAC, whereby some of the NOACs require dose reductions based solely on renal function,
while others require consideration of additional criteria. For that reason, renal
function should be evaluated in patients on a NOAC, as worsening of renal function
may warrant change in the dose of a NOAC or change in oral anticoagulant.
Despite it seems mandatory to assess renal function at baseline, it is also important
to monitor renal function over time. Roldán et al[69 ] and FANTASIIA registry investigators[66 ] demonstrated that worsening glomerular filtrated rate (eGFR) of only ≥10 mL/min
during follow-up was significantly associated with mortality and major bleeding. Fauchier
et al[70 ] showed that worsening in eGFR is an independent predictor of ischemic stroke, thromboembolism,
and bleeding in AF patients.
For that reason, clinical guidelines recommend assessing and monitoring renal function
at least yearly in AF patients under oral anticoagulation with strong evidence.
Biomarkers in Coronary Artery Disease
Biomarkers in Coronary Artery Disease
MI is defined by clinical history, ECG, and increased levels of cardiac troponins
as biomarkers reflecting myocardial damage.[71 ] The role of biomarkers in the definition and diagnosis of ACS is essential and different
clinical guidelines reflect this point. Formal clinical guidelines recommendations
in [Table 4 ].
Table 4
Main current clinical guidelines recommendations for biomarkers use in coronary artery
disease
ESC (72,74)
AHA/ACC (73)
Fourth Universal Definition of MI (71)
Coronary artery disease
• Routine blood sampling for serum markers is indicated as soon as possible in the
acute phase but should not delay reperfusion treatment
Class I, level of recommendation A
• Measurement of a biomarker of cardiomyocyte injury, preferably hs-cardiac troponin,
is mandatory in all patients with suspected NSTEMI. It is recommended to measure cardiac
troponins with sensitive or high-sensitivity assays and obtains the results within
60 min. Additional testing after 3–6 hours is indicated if the first two troponin
measurements are not conclusive and the clinical condition is still suggestive of
ACS
Class I, level or recommendation A
• GRACE risk score provides the most accurate stratification of risk both on admission
and at discharge in ACS. This risk score includes serum creatinine values and elevated
cardiac biomarkers
Class I, level of evidence B
• Cardiac-specific troponin (troponin I or T when a contemporary assay is used) levels
should be measured at presentation and 3 to 6 hours after symptom onset in all patients
who present with symptoms consistent with ACS
Class I, level of evidence: A
• Additional troponin levels should be obtained beyond 6 hours after symptom onset
in patients with normal troponins on serial examination when electrocardiographic
changes and/or clinical presentation confer an intermediate or high index of suspicion
for ACS
Class I, level of evidence: A
• Creatine kinase myocardial isoenzyme (CK-MB) and myoglobin are not useful for diagnosis
of ACS
Class III, Level of Evidence A
• The major criteria for myocardial injury should be used when there is evidence of
elevated cardiac troponin values with at least one value above the 99th
• The clinical definition of MI denotes the presence of acute myocardial injury detected
by abnormal cardiac biomarkers in the setting of evidence of acute myocardial ischemia
Abbreviations: ACC, American College of Cardiology; ACS, acute coronary syndrome;
AHA, American Heart Association; ESC, European Society of Cardiology; Hs, high sensitive;
MI, myocardial infarction; NSTEMI, non-ST elevation myocardial infarction.
Fourth Universal Definition of Myocardial Infarction (2018)
[71 ]
2017 ESC Guidelines for the Management of Acute Myocardial Infarction in Patients
Presenting with ST-Segment Elevation
[72 ]
2014 AHA/ACC Guideline for the Management of Patients with Non–ST-Elevation Acute
Coronary Syndromes (NSTEMI)
[73 ]
If the time of symptom onset is ambiguous, the time of presentation should be considered
as the time of onset for assessing troponin values (Class I, Level of Evidence: A).
Regarding the use of biomarkers for prognosis: the presence and magnitude of troponin
elevations are useful for short- and long-term prognosis (Class I, Level of Evidence
B).
It may be reasonable to remeasure troponin once on day 3 or day 4 in patients with
a MI as an index of infarct size and dynamics of necrosis (Class IIb, Level of Evidence
B).
The use of selected newer biomarkers, especially BNP, may be reasonable to provide
additional prognostic information.
2015 ESC Guidelines for the Management of Acute Coronary Syndromes in Patients Presenting
without Persistent ST-Segment Elevation
[74 ]
A rapid rule-out and rule-in protocol at 0 and 1 hour is recommended if a hs-cTn test
with a validated 0 h/1 h algorithm is available. Additional testing after 3 to 6 hours
is indicated if the first two troponin measurements are not conclusive and the clinical
condition is still suggestive of ACS (Class I, Level of Evidence A).
Beyond diagnostic utility, cardiac troponin levels add prognostic information in terms
of short- and long-term mortality to clinical and ECG variables. The extensively validated
NPs (i.e., BNP, NT-proBNP, and mid-regional pro-A-type NP) provide prognostic information
on top of cardiac troponin.
Comment Recommendations in the Clinical Guidelines
Cardiac biomarkers have an essential role in the “rule-in” and “rule-out” of ACS in
patients with chest pain. Biomarkers complement clinical assessment and the 12-lead
ECG in the diagnosis, risk stratification, triage, and management of patients with
suspected ACS. The measurement of cardiac injury biomarkers shows the magnitude of
the damage and it is a prognostic marker.
Different studies have demonstrated the value of high-sensitive troponins (hs-cTn)
to increase the accuracy of AMI diagnosis and the benefit of hs-cTn compared with
nonsensitive assays was more pronounced in patients presenting early at emergency
room after chest pain.[75 ] Reichlin et al[76 ] analyzed the role of sensitive cardiac troponin assays in the emergency department
for ACS diagnosis and found the predictive performance of hs-cTnT and hs-cTnI was
high (c-statistics between 0.96 and 0.92, respectively) for final diagnosis of AMI.
These results were supported by a multicentric prospective study using four different
brands of hs-cTns to assess the negative predictive value of troponins to rule out
AMI in patients with chest pain and concluded that undetectable levels of hs-cTns
at presentation had a very high negative predictive value and seem to allow the simple
and rapid rule out of AMI. Recently, the algorithm 0/1 hour for rapid rule-out and
rule-in of NSTEMI using hs-cTns has been validated by the European Society of Cardiology
(ESC). Biomarkers were assessed at entry and after 1 hour of admission. Using 4,368
patients with serial cTnT measurements and 3,500 patients with serial cTnI measurements,
this algorithm was validated to be very safe and effective in triaging patients with
suspected NSETMI.[77 ] However, the results of blood biomarkers in STEMI patients in any case could lead
to a delay in revascularization.[72 ] Other biomarkers like creatinine kinase MB isoform have demonstrated less sensitivity
and less specificity for myocardial injury. In the ESC guidelines for the fourth definition
of MI, myocardial injury is defined as an elevated cTn value above the 99th percentile
URL and this definition emphasizes on the benefits of hs-cTns determination.[71 ] This high-sensitivity determination increases the ability of the biomarker to determine
small differences over time and will improve clinical practice.
The main difference between the former cTn assays and hs-cTn assays is the increase
in sensitivity, which is only apparent at values near the 99th percentile (URL).[78 ] Different clinical conditions may influence levels, including stable coronary artery
disease (CAD), chronic HF, renal dysfunction, sepsis, critically ill patients, acute
pulmonary embolism, or pulmonary arterial hypertension.
One of the differences of the recent 4th definition compared with the previous 3rd
definition of MI is related with sex reference levels of troponins for MI diagnosis.
Differences between men and women in levels of hs-cTn to the diagnosis of MI and prognostic
performance have been observed.[79 ] Significantly lower values are observed among women compared with men, and therefore
sex-specific 99th percentile URLs are recommended for hs-cT assays. Sex-specific cut-off
values have been reported to improve diagnostic and prognostic information in patients
with possible acute MI.[71 ]
[80 ]
Humphries et al[81 ] demonstrated different prognosis in women with chest pain in the emergency room;
for example, even when females have cardiac chest pain and cTn levels > 99th percentile,
they are less likely to be diagnosed with MI, less likely to undergo diagnostic cardiac
catheterization within 7 days, and less likely to use evidence-based cardiac medications,
but they have the highest 1-year MACE rate. For that reason, it is important to use
specific hs-cTn range (above the 99th percentile URL) for women to avoid underdiagnosis
of MI.
Nowadays, there are many different cardiac troponins (cTn) available, cTnT, cTnI,
and the high-sensitive hs-cTnT and hs-cTnI. Even though the results from these various
testing platform systems may yield similar clinical interpretation for diagnosis,
that is, above or below the 99th percentile of the assay with a rise and or fall of
cTn, there are considerable differences in the numerical cTn values between assays.
This variability may be due to differences in assay calibration, use of different
antibodies, differences in assay design, instrument limitations, multiple detection
technologies, and differences in the measuring, that is, some assays measure cTnI
and others cTnT.
Other important prognosis marker in patients with ACS is renal function. Patients
with AMI and severe renal disease generally have a poor prognosis. Several registries
have shown an increase of bleeding and mortality after ACS and received less aggressive
therapy than patients with normal renal function. However, bleeding risk was not increased
with the use of third-generation P2Y12 (ticagrelor and prasugrel) in patients with renal impairment[82 ] and these patients are at high risk of mortality and ischemic risk; therefore, it
is mandatory to take the most effective therapy.
Other myocardial stress biomarkers such as NT-proBNP have also been analyzed and high
values are powerful prognostic markers. Indeed, the combination with cardiac troponin
improves risk stratification in NSTEMI; however, the routine use of NT-proBNP in the
diagnostic process of ACS in a patient with chest pain is not supported by the current
evidence.[74 ]
Also, there are studies that have analyzed the influence of inflammatory biomarkers
in the pathogenesis of CAD due to the proinflammatory mediators involved in atherosclerosis
progression. One of these markers is adiponectin. Adiponectin levels are associated
not only with CAD presence, but also with CAD extension or severity. Moreover, they
can be a good predictor of CAD. Other biomarkers are anticardiolipin antibodies of
the immunoglobulin G isotype (IgG aCL). These biomarkers were suggested as risk factor
for arterial and venous thrombosis. Pastori et al[83 ] analyzed the role of IgG aCL in patients with CAD and concluded that patients with
elevated IgG aCL levels have a doubled risk of recurrent MACEs and should be suspected
in patients with juvenile CAD. However, the use of adiponectin or IgG aCL in daily
clinical practice to guide therapy management is limited.[84 ]
Biomarkers are also included in the risk score assessments of ACS. The GRACE score
includes renal function and elevated cardiac troponin levels as biomarkers. The GRACE
scores provided superior discrimination as compared to the TIMI NSTEMI score in predicting
in-hospital and 6-month mortality in NSTEMI patients, although the GRACE and TIMI
STEMI scores performed equally well in STEMI patients.[85 ] Indeed, an update of GRACE score (Grace 2.0) was also validated in 32,037 patients
from the GRACE registry (14 countries, 94 hospitals) and externally validated in the
French registry of Acute ST-elevation and non-ST-elevation MI (FAST-MI). This update
had better discrimination, is easier to use, and performed equally well acutely and
over the long-term events.[86 ] The use of GRACE risk score could help in prognosis of patients after NSTEMI and
guide antiplatelet therapy.
Biomarkers in the Prevention of Cardiovascular Disease
Biomarkers in the Prevention of Cardiovascular Disease
The relationship between classical risk factors to CVDs such as hypertension, diabetes,
smoke, or dyslipidemia is well-known. However, some patients present CVD without elevation
of traditional risk factors. In this context, the role of biomarkers for primary prevention
of CVDs has emerged, trying to assess a CV risk profile of the patients. CRP, apolipoprotein
B (apoB), homocysteine, albuminuria, or reactive oxidant species (ROS) have been proposed
as biomarkers involved in atherosclerotic process and CV disease progression. Formal
clinical guidelines recommendations in [Table 5 ].
Table 5
Main current clinical guidelines recommendations for biomarkers use in cardiovascular
disease prevention
ESC (86)
AHA/ACC (87)
Cardiovascular disease
• Routine assessment of circulating or urinary biomarkers is not recommended for refinement
of CVD risk stratification
Class III, level of evidence B
• The role of metabolomics as risk factors for CVD and to improve CV risk prediction
beyond conventional risk factors should be further assessed
• Assessment of risk-enhancing factors:
Lipids/biomarkers: associated with increased cardiovascular risk:
- Persistently elevated primary hypertriglyceridemia (≥175 mg/dL, nonfasting)
- If measured:
**Elevated high-sensitivity C-reactive protein (≥2.0 mg/L)
**Elevated Lp(a): A relative indication for its measurement is family history of
premature cardiovascular disease
**An Lp(a) ≥50 mg/dL or ≥125 nmol/L constitutes a risk-enhancing factor, especially
at higher levels of Lp(a)
**Elevated apoB (≥130 mg/dL): A relative indication for its measurement would be
triglyceride ≥200 mg/dL. A level ≥130 mg/dL corresponds to an LDL-C > 160 mg/dL and
constitutes a risk-enhancing factor
**Ankle-branchial index (< 0.9)
Class IIa, level of evidence B
• Coronary artery calcium score risk
In intermediate-risk (≥7.5% to < 20% 10-y risk) adults or selected borderline-risk
(5% to < 7.5% 10-y risk) adults in whom a coronary artery calcium score is measured
for the purpose of making a treatment decision
• If the coronary artery calcium score is zero, it is reasonable to withhold statin
therapy and reassess in 5 to 10 y, as long as higher-risk conditions are absent
• If coronary artery calcium score is 1 to 99, it is reasonable to initiate statin
therapy for patients ≥55 y of age
• If coronary artery calcium score is 100 or higher or in the 75th percentile or higher,
it is reasonable to initiate statin therapy
Class IIa, level of evidence B
Abbreviations: ACC, American College of Cardiology; AHA, American Heart Association;
CRP, C-reactive protein; CV, cardiovascular; CVD, cardiovascular disease; ESC, European
Society of Cardiology; Hs, high sensitive.
2016 European Guidelines on Cardiovascular Disease Prevention in Clinical Practice
[87 ]
CV circulating and urinary biomarkers exhibit either no or only limited value when
added to CVD risk assessment with the SCORE system. There is evidence of publication
bias in the field of novel biomarkers of CV risk, leading to inflated estimates of
strength of association and potential added value.
Not all potentially useful circulatory and urinary biomarkers have undergone state-of-the-art
assessment of their added value in CV risk prediction on top of conventional risk
factors.
Biomarkers may be useful in specific subgroups, but this has been addressed in only
a limited number of studies.
2019 ACC/AHA Guideline on the Primary Prevention of Cardiovascular Disease
[88 ]
In adults at borderline risk (5 to < 7.5% 10-year CV risk) or intermediate risk (≥7.5
to < 20% 10-year ASCVD risk), it is reasonable to use additional risk enhancing factors
to guide decisions about preventive interventions (e.g., statin therapy).
For individuals with intermediate predicted risk (≥7.5 to < 20%) or for select adults
with borderline (5 to < 7.5%) predicted risk, coronary artery calcium measurement
can be a useful tool in refining risk assessment for preventive interventions (e.g.,
statin therapy). Coronary artery calcium scoring has superior discrimination and risk
reclassification as compared with other subclinical imaging markers or biomarkers.
Comment on Recommendations in the Clinical Guidelines
Different biomarkers have been proposed to improve CV risk classification especially
in patients at intermediate risk with the common clinical factors.
The inflammatory biomarker CRP is one of the most frequently analyzed. CRP binds to
low-density lipoprotein and is present in atherosclerotic plaques, so it has been
proposed as a causal role in coronary heart disease. The Emerging Risk Factors Collaboration
group[89 ] performed a meta-analysis with individual record of 160,309 without history of CVD
and analyzed the association between CRP and mortality, as well as other adverse outcomes.
The authors observed that CRP concentration has continuous association with the risk
of coronary heart disease, ischemic stroke, vascular mortality, and death from several
cancers and lung disease. However, CRP has a role in very diverse diseases and in
some studies with small effect, thus it is difficult to certain that the final data
are not simply a result of residual confounding or selection bias due its influence
with other risk factors.
Other biomarkers such as apoB and homocysteine have also been analyzed. Wald et al[90 ] performed a meta-analysis to analyze the relationship between serum homocysteine
concentration with ischemic heart disease, deep vein thrombosis, and pulmonary embolism.
The authors observed a significant association between homocysteine and these three
diseases and lowering homocysteine concentrations by 3 µmol/L would reduce the risk
of ischemic heart disease by 16%. Based on these results, Akintoye et al[91 ] proposed a new biomarker score, the CHAN2 T3 score, a new biomarker score using five biochemical risk markers: CRP, homocysteine,
albuminuria, NT-proBNP, and troponin T. A score of ≥2 was associated with improvement
in the c-statistic of the pooled cohort equation for the estimation of CV risk with
a combination of traditional risk factors (0.748 vs. 0.734, p = 0.02). Elevated levels of plasma total homocysteine can result from genetic or
nutrient-related disturbances in the transsulfuration or remethylation pathways for
homocysteine metabolism.[92 ] The enzyme 5,10-methylenetetrahydrofolate reductase (MTHFR) catalyzes the methylation
of homocysteine to methionine. Inherited mutations in the gene that make the MTHFR
enzyme can lead to an enzyme that is not optimally active and, consequently, may lead
to elevated homocysteine levels.[93 ] However, the causal mechanism of CVD is not clear or if homocysteine is only a marker
of an increased clotting risk.[90 ]
Ma et al[94 ] identified that MTHFR polymorphism was associated with higher homocysteine levels
but not with risk of MI. For these controversial results and some methodological limitations,
the assessment of homocysteine in daily clinical practice to predict and stratify
the CV risk or routinely treatment of patients with elevated homocysteine is not recommended
in the current clinical guidelines.
Lipoprotein (a) [Lp(a)] is composed of apoB-100 covalently bound to apoA. Many studies
have concluded that Lp(a) is associated with the prevalence as well as the severity
of CAD and future CV events in the general population.[95 ]
[96 ] Indeed, Lp(a) is an independent risk factor for atherosclerotic CVD, with elevated
levels estimated to be prevalent in 20% of the population. Observational and genetic
evidence strongly support a causal relationship between high plasma concentrations
of Lp(a) and increased risk of atherosclerotic CVD-related events, such as MI and
stroke, and valvular aortic stenosis.[97 ] Zhang et al[98 ] showed that elevated Lp(a) levels were independently associated with the presence
and severity of CAD in patients with diabetes mellitus. For that reason, American
Heart Association guidelines recommend the assessment of risk factors in patients
at risk of CVD, with measuring of Lp(a) levels.
In relation to lipid control, another promising biomarker is represented by proprotein
convertase subtilisin/kexin type-9 (PCSK9). The PCSK9 inhibitors showed an important
role in cholesterol-lipid lowering in patients with ACS and some studies analyzed
the pleiotropic role of these inhibitors. Paciullo et al[99 ] reviewed the role of PCSK9 in hemostasis and thrombosis, and observed that in experimental
mice, PCSK9 knockout mice develop less venous and arterial thrombosis and show reduced
in vivo platelet activation upon arterial injury. In the same line, Pastori et al[100 ] analyzed the impact of PCSK9 levels in AF patients and observed that plasma PCSK9
levels are associated with an increase of CV events as well as positive correlation
with urinary thromboxane as a mechanism implicated in platelet activation.
Moreover, the assessment of coronary artery calcium as subclinical imaging marker
could improve risk score stratification in patients with intermediate predicted risk.
In the MESA (Multi-Ethnic Study of Atherosclerosis) trial, the coronary artery calcium
score was strongly associated with 10-year CV risk across age, sex, and racial/ethnic
groups, independent of traditional risk factors.[101 ] Indeed, a coronary artery calcium score of < 0 identifies individuals at lower risk
of CV events and death over a ≥10-year period.[101 ]
[102 ]
[103 ] Thus, the absence of coronary artery calcium could reclassify a patient downward
into a lower risk group in which preventive interventions (e.g., statins) could be
postponed.[104 ]
A new perspective in CV risk assessment involves the role of oxidative stress biomarkers.
Different studies about biomarkers related with oxidative stress and its influence
in CVD are ongoing. A panel of more than 70 biomarkers of oxidative stress, especially
related to lipid oxidation and peroxidation, has been developed but there is a lack
of validation about the role of biomarkers levels and CV risk profile of patients.[105 ] One of the biomarkers are focused in ROS-generating enzymes, NADPH oxidase, and
MPO. These enzymes have been suggested to be implicated in the atherosclerosis process
and plaque instability.[106 ] Pastori et al[107 ] performing a review of recent reports focus on NADPH oxidase and MPO measurement
and observed disappointing results in interventional studies with antioxidants in
primary and secondary prevention of CV diseases, but the use these biomarkers in the
assessment of atherosclerotic risk should be promising.
Conclusion
Ultimately, a balance is needed between simplicity and practicality for clinical decision-making.
Most biomarkers (whether blood, urine, or imaging-based) will improve on clinical
risk stratification, but awaiting biomarker results may lead to delays in the initiation
of therapy, for example, anticoagulation for stroke prevention in AF. Many biomarkers
are nonspecific, being predictive of many CV and non-CV outcomes, so would be better
as “rule-out” rather than “rule-in” assessments ([Fig. 1 ]). Derivation of some biomarkers have also been made in highly selected clinical
trial cohorts, where measurement is made at baseline but outcomes determined many
years later; given the dynamic nature of risk in the “real world” where patients get
older and develop incident risk factors, this may give a false impression of the risk
profile. Finally, some laboratory biomarkers have a diurnal variation and inter-/intravariability
(and lower limits of detection) in assays, which may be expensive, are added considerations.
Indeed, many clinical guidelines recommendations about biomarkers' use in clinical
practice reflect the consensus expert opinion, with limited and weak evidence. In
this field, the recommendations of clinical guidelines should be analyzed with caution,
using consistent, real-world, and all external validation data available to maximize
the generalization and uniformity of the recommendations. The final objective should
be to help clinicians in decision-making process in daily clinical practice, assessing
all the clinical and blood markers to offer the best therapy to patients with CVD.
A balance between simplicity and practical application, versus modest improvements
in prediction (at least statistically), is needed.
Fig. 1 Different biomarkers involved in cardiovascular diseases. CK-MB, creatine kinase
myocardial isoenzyme; CRP, C-reactive protein; GDF-15, growth differentiation factor-15;
Lp(a), lipoprotein a; MPO, myeloperoxidase; MRproADM, mid-regional pro-adrenomedullin;
NOX:Gal-3, galectin 3; NT-proBNP, N-terminal pro-B type natriuretic peptide; ROS,
reactive oxidant species; sST2, soluble suppression of tumorigenicity 2.
