NT–probrain natriuretic peptide predicts complexityand severity of the coronary lesions in patients withnon–ST-elevation acute coronary syndromesJose Luis Navarro Estrada, MD, FACC, MSc,a Fernando Rubinstein, MD, MSc,a Maria Cecilia Bahit, MD,a

Florencia Rolandi, MD,a Diego Perez de Arenaza, MD,a Jose M. Gabay, MD,a Jose Alvarez, MD,b

Ricardo Sarmiento, MD,c Carlos Rojas Matas, MD,a Carlos Sztejfman, MD,d Alejandro Tettamanzi, MD,e

Raul de Miguel, Biochem,a and Luis Guzman, MD, FACC,f on behalf of PACS Investigators Buenos Aires,Argentina; and Jacksonville, FL

Background NT–probrain natriuretic peptide (NT-proBNP) has been associated with left ventricular (LV) dysfunctionand adverse outcome in patients with non–ST-elevation acute coronary syndromes (NSTEACS). However, the underlyingpathophysiological mechanisms responsible for this association have not been well established. We sought to explore therelation between NT-proBNP levels and extension of coronary artery disease (CAD) and the presence of more complexand severe coronary lesions.

Methods This prospective, multicenter angiographic substudy included 585 patients admitted with NSTEACS. Blindedmeasurements of NT-proBNP and troponin T were performed at a median time of 3 hours after admission and analyzedcentrally. Angiograms were read at a core laboratory by 2 independent readers blinded to patient data. Complex coronarylesion was defined as the presence of at least one of the following: thrombus (+), TIMI flow b2, or ulcerated plaque.

Results NT–probrain natriuretic peptide levels increased proportionally as LV function decreased. The levels of NT-proBNP were directly related to the extent of the CAD. This association was maintained when we analyzed patients withnormal LV function (n = 257). Patients with complex coronary lesions or those with at least one of its individual componenthad higher levels of NT-proBNP compared with those without complex coronary lesions. After adjusting for clinical andelectrocardiographic variables and other biomarkers, positive troponin (OR 2.20, 95% CI 1.50-3.22, P b .0001) andsupramedian NT-proBNP levels (OR 1.72, 95% CI 1.19-2.47, P = .003) independently contributed to the prediction ofcomplex coronary lesions.

Conclusion In this study of patients with NSTEACS, NT-proBNP levels progressively increase with the severity ofCAD and degree of LV dysfunction. Increased levels of NT-proBNP independently predict the presence of more complexcoronary lesions. (Am Heart J 2006;151:1093.e1-1093.e7.)

B-type natriuretic peptide (BNP) is a cardiac neuro-

hormone released from ventricular myocardium in

response to increased wall stress and myocardial

tension.1 The prognostic value of BNP levels in patients

with heart failure has been demonstrated in several

studies.2,3 In patients with acute myocardial infarction

(MI), elevated plasma levels of BNP were associated with

From the aHospital Italiano de Buenos Aires, Ciudad de Buenos Aires, Buenos Aires,

Argentina, bHospital Britanico, Ciudad de Buenos Aires, Argentina, cHospital Frances,

Ciudad de Buenos Aires, Argentina, dSanatorio Mitre, eCEMIC, Buenos Aires, Argentina,

and fUniversity of Florida, Jacksonville, FL.

Submitted September 9, 2005; accepted December 6, 2005.

Reprint requests: Jose Luis Navarro Estrada, MD, FACC, MSc, Hospital Italiano de

Buenos Aires, Gascon 450, 1181 Buenos Aires, Argentina.

E-mail: jose.navarroestrada@hospitalitaliano.org.ar

0002-8703/$ - see front matter

n 2006, Mosby Inc. All rights reserved.

doi:10.1016/j.ahj.2005.12.020

increased mortality, independently of left ventricular

(LV) function. More recently, the important prognostic

information of BNP has extended across the spectrum of

patients with non–ST-elevation acute coronary syn-

dromes (NSTEACS). In this setting, elevated levels of

BNP and its inactive form NT–probrain natriuretic

peptide (NT-proBNP) were independently associated

with adverse prognosis, including increased mortality.4 - 9

However, the underlying pathophysiological mecha-

nisms responsible for the association between BNP and

NT-proBNP and outcome in patients with NSTEACS have

not been well established. We undertook the present

study to explore the hypothesis that the prognostic

information of NT-proBNP levels in these patients could

be linked to transient ischemia and, therefore, would

be associated to the extension of CAD and the presence

of more complex and severe coronary lesions.

American Heart Journal

May 20061093.e2 Navarro Estrada et al

MethodsStudy patients

This was a prospectively designed substudy of patients

undergoing coronary angiography in a prospective, multicenter

cohort study (PACS) developed in 11 coronary care units of

university-affiliated hospitals. The original study evaluated

the prognostic value of different biochemical markers,

isolated or in combination, in the risk stratification of patients

with NSTEACS.9

The PACS study included a total of 1483 consecutive

patients admitted to the coronary intensive care unit with

diagnosis of NSTEACS.

Only 9 centers with cardiac catheterization capabilities were

included in the angiographic substudy.

After informed consent had been obtained, serum samples

for a single measurement of NT-proBNP, Troponin T (TnT),

high-sensitivity C-reactive protein (hsCRP) and myoglobin were

drawn. These were stored at �708C until they were analyzed at

a core laboratory. The investigators were blinded to the test

results until the study was completed.

All patients were followed up from admission until day 180 or

death, whichever occurred first.

Study organizationThe ethics committee of each center approved the study

protocol, and patients provided informed consent. All major

events were adjudicated by a clinical event committee that did

not know the results of the serum biomarker determinations or

the identities of patients and centers. All case report forms

were reviewed automatically and manually to ensure the

consistency and quality of the data.

Coronary angiography and lesion morphologyThe angiographic procedures were described in a previous

report.10 In brief, coronary angiography was performed follow-

ing a standard technique. All angiograms were evaluated

prospectively in an independent angiographic core laboratory,

following a prespecified protocol. Two expert interventional

cardiologists reviewed the angiograms with no knowledge of the

biomarker levels and patient outcome. If they disagreed, a third

expert examined the angiogram to determine the characteristics

of the lesions. Angiographic severe lesions were defined as the

presence of stenosis N50% in vessels with a diameter equal or

N 2.0 mm. Extension of coronary disease was classified

according to the standard way into 1-, 2-, and 3-vessel disease and

or left main disease. Angiographic characteristics were analyzed

in each of the lesions to define their complexity. bcomplex

coronary lesionQ was defined as the presence of at least one of

the following: (1) presence of intracoronary thrombus, (2)

abnormal TIMI flow, and (3) ulcerated plaque. Intraluminal

thrombus was defined as the presence of a filling defect visible in

multiple projections with at least 3 edges surrounded by

contrast.11 Flow distal to the lesion was classified according to

the classic TIMI criteria. Lesion was divided as with bnormal

flowQ when there was a TIMI 3 grade flow and babnormal flowQas TIMI V2 flow. In the case of a lesion with TIMI 0 flow, it was

left to the reader interpretation as to whether the occlusion

appeared to be recent or chronic. However, the angiographic

features of a recent occlusion were previously determined by

consensus. Only those lesions with TIMI flow 0 classified as a

recent occlusion were considered as criteria for a complex

coronary lesion. Ulcerated plaque was defined as the presence

of a small crater with a narrow neck that extended beyond

the lumen of the vessel or the presence of spontaneous intimal

flaps at the site of coronary obstruction.12,13 Irregular contours

were not considered criteria for complex coronary lesion

characteristic in this study. For the evaluation of the morpho-

logical lesion characteristics, the 2 readers agreed in 499 cases

(84%). The LV function was measured by contrast ventriculo-

gram and classified as normal (if LV ejection fraction [LVEF] is

z50%), mild (LVEF 40%-49%), moderate (LVEF 30-39%), and

severe (LVEF b30%) LV dysfunction.

Laboratory measurementsPlasma samples for the determination of the cardiac markers

were collected at a median time of 3 hours after admission.

NT–probrain natriuretic peptide was measured with a sand-

wich immunoassay on a Modular E170 (Roche Diagnostics,

Baden-Wurttemberg, Germany). The analytical range extends

from 5 to 35000 pg/mL. The total coefficient of variation was

3.3% (n = 28) at a level of 252.6 pg/mL and 3.7% (n = 25) at a

level of 6130.8 pg/mL of NT-proBNP. Troponin T levels were

measured with a third-generation immunoassay method

(Elecsys 2010, Roche Diagnostics). The reference limit recom-

mended by the manufacturer was b0.03 ng/mL. High-sensitivity

CRP was measured by immunonephelometry with intensifying

particles (Dade Behring, Inc., Deerfield, IL) with a reference

limit of N3 mg/L. Myoglobin was measured by immunoassay

(Elecsys 2010, Roche Diagnostics) with a reference limit of

N72 ng/mL for male and N51 ng/mL for female.

Statistical methodsTo assess the association between the variable of interest,

NT-proBNP, with both the presence of complex coronary

lesions, and the extension of coronary artery disease (CAD), an

exploratory analysis was performed to define z1 meaningful

limits of NT-proBNP for clinical evaluation. Plasma concentra-

tion levels of NT-proBNP followed a highly right skewed

distribution; therefore, this was evaluated as a continuous and

as a semicontinuous variable with multiple categories based on

percentiles (20 groups). After looking at the distribution of the

outcome (presence of complex coronary lesions) across the

categories of NT-proBNP, and to facilitate interpretability, it

was decided to use the median as the cutoff point because it

gave optimal discrimination with regard to the presence of

complex coronary lesions (Figure 1). This finding was

confirmed by receiver operating characteristic curve analysis

because the median showed the widest difference between

positive and negative likelihood ratios.

For the descriptive analysis and the bivariate comparisons

between NT-proBNP, other clinical characteristics and the

outcomes of interest, medians and interquartile range, and the

corresponding non parametric test statistic analysis were

reported. When the distribution of the values of NT-proBNP

was explored as a continuous variable, Wilcoxon rank sum

test or Kruskal-Wallis test was used for N2 categories.

Pearson m2 or a m2 for trends was used for the comparison

of the dichotomous NT-proBNP using the median as the

cutoff point and other clinical variables.

Table I. Baseline characteristics by NT-proBNP levels

NT-proBNP__bbbbV278.7 pg/mL

(n = 293)

NT-proBNPNNNN278.7 pg/mL

(n = 292) P

Age N75 y 41 (14.5) 104 (35.4) .000Male sex 223 (76.6) 205 (69.7) .60HTA 178 (61.2) 204 (69.4) .037Smoker 106 (36.4) 61 (24.8) .002Diabetes 43 (14.8) 61 (20.7) .59Hyperlipidemia 178 (61.2) 169 (57.5) .364Prior MI 51 (17.4) 78 (26,7) .007Prior CABG 23 (7.9) 41 (13.9) .019Prior PTCA 35 (12.0) 41 (13.9) .490Ischemic changes

(ECG)152 (52.2) 206 (70.1) .000

Myoglobin-positive 83 (28.5) 131 (44.6) .000hsCRP N3 mg/dL 153 (52.6) 199 (67.7) .000TnT N0.03 ng/mL 108 (36.9) 211 (72.1) .000

Data are shown as n (%) for dichotomous variables. HTA, Hypertension; CABG,coronary artery bypass graft; PTCA, percutaneous transluminal coronary angioplasty;ECG, electrocardiogram.

Figure 1

Complex coronary lesion stratified by NT-proBNP levels.

Figure 2

Left ventricular function and NT-proBNP levels. Results are expressedas median and interquartile range.

Figure 3

Coronary artery disease extension and NT-proBNP levels. Results areexpressed as median and interquartile range.

American Heart Journal

Volume 151, Number 5Navarro Estrada et al 1093.e3

To assess the independent effect of NT-proBNP level on the

risk of complex coronary lesions adjusting for potential

confounders, a stepwise interactive multiple logistic regression

analysis was performed. A model including those covariates

showing a P value of .10 on the univariate screen analysis one

at a time was built. Then, only those with a significant Wald

test or likelihood ratio test or those showing a confounding

effect on the association between NT-proBNP and the outcome

were retained in the final model. Global fit of this explanatory

model was evaluated with the Hosmer-Lemeshow test.

ResultsOf the 1483 patients from the PACS study, 1252 pa-

tients were enrolled in 9 centers that participated in the

angiographic substudy. Of those, 633 patients (50.5%),

underwent coronary angiography at a median time of

46 hours (25th to 75th percentile 20-70 hours) from

admission. Of the 633 angiograms, 48 (7.5%) were

excluded from the analysis because of technical reasons

or unacceptable quality.

A total of 585 patients had both core laboratory angio-

graphic data and NT-proBNP levels on admission. The

median (25th-75th percentile) NT-proBNP levels were

278.7 pg/mL (109.8, 758.5) pg/mL, respectively (range

12.6-35000 pg/mL).

The baseline clinical characteristics of the patients

are shown in Table I. Patients with elevated levels of

NT-proBNP were older, more commonly smokers, and

had more likely history of MI. At presentation, patients

with elevated NT-proBNP more frequently had

Figure 4

Supramedian NT-proBNP levels according to LV function and extension of CAD.

Table II. Median and 25th to 75th percentiles of NT-proBNPaccording to the presence of complex coronary lesions and theirdifferent components

Lesionmorphologyabnormality

NT-proBNP Median (25th-75th) (pg/mL)

Absent Present P

Complex coronarylesion

189 (86-523) 386 (164-874) .000

Thrombus 257 (103-759) 358 (186-790) .009Abnormal

TIMI flow235 (96-639) 443 (203-871) .000

Ulceratedplaque

309 (131-815) 331 (154-802) .686

American Heart Journal

May 20061093.e4 Navarro Estrada et al

ischemic ST-T changes and elevation of TnT, hsCRP,

and myoglobin.

Patients with baseline NT-proBNP levels N278.7 pg/mL

were at significantly higher risk of death at 180 days

(6.8% vs 2.1%, P b .0001), and the composite of death or

MI (11.9% vs 6.5%, P = .0025).

Left ventricular function, CAD extension, andNT-proBNP levels

Left ventriculograms were performed in 506 patients

(87%). Among them, 257 patients (51%) had normal

LV function; 136 (27%), 81(16%), and 32 (6%) patients

had mild, moderate, and severe LV dysfunction, respec-

tively. NT–probrain natriuretic peptide levels increased

significantly as LV function decreased (Figure 2).

Ninety-nine patients (17%) showed non significant

lesions or completely normal coronary arteries;

187 (32%), 155 (26%), and 142 (25%) had 1-vessel,

2-vessel, and 3-vessel and/or left main disease,

respectively. As seen in Figure 3, the levels of

NT-proBNP were directly related to the extent of the

CAD, and the difference between each category was

highly significant ( P = .0001).

The combination of the extension of CAD and the

severity of LV dysfunction provided the strongest

association with high NT-proBNP levels. Among patients

in every category of LV function, NT-proBNP levels

increased as the extension of CAD progressed (Figure 4).

This association between NT-proBNP levels and num-

ber of vessels was maintained in patients with normal LV

function (n = 257). The levels of NT-proBNP increased

significantly as the number of vessels affected increased)

(median [interquartile range ] in patients with normal

coronaries 110.6 pg/mL [172.3 ], 1 vessel 162.0 pg/mL

[322.5]; 2-vessels 198.7 pg/mL [261.4]; and 3-vessels or

left main disease 315.4 pg/mL [885.8] [ P b .0001].

Lesion morphology and NT-proBNP levelsComplex coronary lesions were observed in 264

(45.1%) patients with NSTEACS. Coronary thrombi were

seen in 104 (17.8%), abnormal TIMI flow in 133 (22.7%),

and ulcerated plaque in 142 (24.3%) patients. Patients

with complex coronary lesions or those with at least one

of its individual components had higher levels of NT-

proBNP compared with those without complex coro-

nary lesions. (Table II).

As seen in Figure 1, the proportion of complex

coronary lesions rose gradually as NT-proBNP centiles

increased; however, the predictive capacity of baseline

measurement of NT-proBNP showed a threshold at the

Figure 5

Stepwise logistic regression model for the prediction of complex coronary lesion. NT-proBNP 278, Supramedian NT-proBNP level.

American Heart Journal

Volume 151, Number 5Navarro Estrada et al 1093.e5

median value. This threshold was confirmed by receiver

operating characteristic curve analysis.

NT-proBNP as independent predictor of complexcoronary lesions

A multivariable logistic regression analysis was per-

formed to identify predictors of complex coronary

lesions. NT–probrain natriuretic peptide, troponin

N0.03 ng/mL, ST-segment depression z0.5 mV,

hsCRP N3 mg/L, positive myoglobin, severe angina

defined as N2 episodes of chest pain within 24 hours of

admission, and diabetes were found to be predictive in

the univariate analysis and, therefore, were introduced

in the regression model. NT–probrain natriuretic

peptide entered into the model as a dichotomous

variable with a cutoff at the median value (N278.7 vs

V278.7 pg/mL). Positive troponin and NT-proBNP

N278.7 pg/mL independently contributed to the pre-

diction of complex coronary lesions, and ST-segment

changes had only borderline significance (Figure 5).

When NT-proBNP was entered in the multivariate

analysis as a continuous variable, the predictors

remained the same (log NT-proBNP: OR 1.18, P = .018).

Goodness of fit was adequate, with a P = .785 in the

Hosmer-Lemeshow test (df = 8).

DiscussionThe results of this study demonstrate that elevated

NT-proBNP levels in patients with NSTEACS are associ-

ated with lower LV ejection fraction and more extensive

CAD. Furthermore, using a previously defined and

prospectively evaluated criteria for bunstableQ angio-

graphic lesions (complex coronary lesions), the data

from this study demonstrate that NT-proBNP level

independently predicts the presence of more complex

coronary lesions.

Several previous studies have demonstrated the

relation between elevations of BNP4,5 and NT-proBNP

levels7-9 and adverse outcomes in patients with

NSTEACS. Because BNP is released in response to

increased myocardial wall stress, it was proposed that

in patients with acute coronary syndromes with overt

heart failure or decreased LV function, elevated levels

of the biomarker reflect myocardial dysfunction and

are associated with high risk of adverse events. The

present results found a significant association between

NT-proBNP and ventricular function, confirming the

findings by Omland et al,7 which showed that supra-

median NT-proBNP values were associated with lower

LV function.

A previous work has demonstrated that patients with

elevated BNP were more likely to have multivessel

CAD.5 Furthermore, a recent study has shown that

patients with NSTEACS with elevated BNP have tighter

culprit artery diameter stenosis, a higher corrected TIMI

frame count, and a more frequently left anterior

descending involvement and proximal culprit lesion

location.14 Another study assessed NT-proBNP levels in

American Heart Journal

May 20061093.e6 Navarro Estrada et al

patients with the wide spectrum of CAD and controls

and showed that NT-proBNP concentrations increased

progressively with the increase of CAD severity.15 The

results of our study confirmed these findings; among

patients with NSTEACS, elevated NT-proBNP levels had

a very close relation with the number of vessels

involved.5 More importantly, the present study has also

shown that even in patients with normal LV function,

the level of NT-proBNP maintains the same relation with

the extension of CAD. Besides, LV ejection fraction was

measured within the first hours (median 46 hours) of

the onset of the coronary event, and the ventriculograms

were blindly analyzed in a core laboratory, giving an

accurate measurement of LVEF in the acute phase of

acute coronary syndrome. This finding suggests that a

mechanism other than LV dysfunction might be re-

sponsible for the adverse outcome in patients with

NSTEACS and high levels of NT-proBNP.

One of the more important and novel findings of the

present study was the strong and independent associa-

tion between NT-proBNP levels and the presence of

more complex coronary lesions. This association

persisted even when other well-recognized clinical

predictors like cardiac TnT and ischemic ST changes

were included in the multivariate model. The relation

between complex coronary lesions and the magnitude

of the ischemic insult and long-term prognosis is well

known.16 In contrast, other biomarkers such as

C-reactive protein does not predict either the extension

or the complexity of coronary disease in patients with

NSTEACS.10 Several observations have suggested that

BNP increases with severity and extent of ischemia.

Levels of BNP increase transiently after uncomplicated

coronary angioplasty in the absence of demonstrable

changes in LV function or LV end-diastolic pressure.17,18

Likewise, in patients with proven CAD, BNP levels

increased during dynamic exercise, compared with

patients without CAD19,20; and in another study, BNP

increased during thallium exercise testing and the

severity of myocardial perfusion defect correlated with

greater plasma levels of BNP.17,18

The pathophysiological mechanism behind the asso-

ciation between ischemia and elevation of BNP is not

well defined. However, recent reports may help to

clarify this association. Toth et al21 have found evidence

that tissue hypoxia alone trigger release of BNP in

absence of LV dysfunction. These results were con-

firmed by a recent physiological study showing that

ventricular BNP gene expression is up-regulated by

myocardial hypoxia resulting in augmented plasma

concentrations of BNP and NT-proBNP.22 Thus, the

available evidence supports a plausible relation between

severity of CAD and the presence of complex coronary

lesions that would predispose to a more severe and

extensive myocardial ischemia and, thus, provoking a

rise in NT-proBNP independently of LV function.

Study limitationsThe NT-proBNP threshold used in the present study

has not been validated before. However, given a

stepwise relation between increasing levels of

NT-proBNP, the presence of complex coronary lesions

with a robust predictive capacity, and the absence of

available data, the NT-proBNP result at the median (50th

percentile) appears to be a convenient decision limit.

In addition, these findings were confirmed when

NT-proBNP was included as a continuous variable.

ConclusionIn this large series of consecutive patients with

NSTEACS, NT-proBNP levels progressively increases

with the severity of CAD and degree of LV dysfunction.

Increased levels of NT-proBNP independently predict

the presence of more complex coronary lesions. These

findings suggest that, in addition to LV dysfunction,

greater severity and extent of myocardial ischemia may

explain, at least in part, the underlying mechanism of

the adverse clinical outcome associated with NT-proBNP

in patients with NSTEACS.

References1. Wiese S, Breyer T, Dragu A, et al. Gene expression of brain

natriuretic peptide in isolated atrial and ventricular human myo-cardium: influence of angiotensin 2 and diastolic fiber length.Circulation 2000;102:3074 -9.

2. Cowie MR, Mendex GF. BNP and congestive heart failure. ProgCardiovasc Dis 2002;44:293 -321.

3. McCullough PA, Maisel AS, BNP Multinational Study Investigators.B-type natriuretic peptide and clinical judgement in emergencydiagnosis of heart failure. Circulation 2002;106:416 -22.

4. de Lemos JA, Morrow DA, Bentley JH, et al. The prognostic value ofB-type natriuretic peptide in patients with acute coronary syndromesN Engl J Med 2001;345:1014 -21.

5. Morrow DA, de Lemos JA, Sabatine MS, et al. Evaluation ofB-type natriuretic peptide for risk assessment in unstable angina/non–ST-elevation myocardial infarction. J Am Coll Cardiol2003;41:1264 -72.

6. Jernberg T, Stridsberg M, Venge P, et al. N-terminalpro brain natriuretic peptide on admission for early riskstratification of patients with chest pain and noST-segment elevation. J Am Coll Cardiol2002;40:437 -45.

7. Omland T, Persson A, Ng L, et al. N-terminal pro–B-type natriureticpeptide and long-term mortality in acute coronary syndromes.Circulation 2002;106:2913 -8.

8. James SK, Lindahl B, Siegbahn A, et al. N–terminal pro-brainnatriuretic peptide and other risk markers for the separate predic-tion of mortality and subsequent myocardial infarction inpatients with unstable coronary artery disease. Circulation 2003;108:275 -81.

9. Bazzino O, Fuselli JJ, Botto F, et al. Relative value of N-terminalprobrain natriuretic peptide, TIMI risk score, ACC/AHAprognostic classification and other risk markers in patients withnon–ST elevation acute coronary syndromes. Eur Heart J2004;25:859 -66.

American Heart Journal

Volume 151, Number 5Navarro Estrada et al 1093.e7

10. Navarro Estrada JL, Gabay JM, Alvarez J, et al. Relation ofC-reactive protein to extent and complexity of coronarynarrowing in patients with non–ST elevation acute coronarysyndromes. A prospective cohort study. Coron Artery Dis2004;15:477 -84.

11. Alderman EL, Stadius M. The angiographic definitions of the BypassAngioplasty Revascularization Investigation study (BARI). CoronArtery Dis 1992;3:1189 -207.

12. Ellis SG, Vandormael MG, Cowley MJ. Coronary morphologic andclinical determinants of procedural outcome with angioplasty formultivessel coronary disease: implications for patients selection.(Multivessel Angioplasty Prognosis Study Group). Circulation 1990;82:1193 -202.

13. ACC/AHA guidelines for the management of patients with unstableangina and non–ST-segment elevation myocardial infarction.Braunwald y col. J Am Coll Cardiol 2000;36:970 -1062.

14. Sadanandan S, Cannon CP, Chekuri K. Association of elevatedB-type natriuretic peptide levels with angiographic findings amongpatients with unstable angina and non–ST elevation acute coronarysyndromes. JACC 2004;44:564 -8.

15. Ndrepepa G, Braun S, Mehilli J, et al. Plasma levels of N-terminalPro-BNP in patients with coronary artery disease and relation to

clinical presentation, angiographic severity, and left ventricularejection fraction. Am J Cardiol 2005;95:553 -7.

16. Ambrose JA, Israel DH. Angiography in unstable angina. Am JCardiol 1991;68:78B -84B.

17. Morrow DA, Braunwald E. Future of biomarkers in acute coronarysyndromes—moving toward a multimarker strategy. Circulation2003;108:250 -2.

18. de Lemos JA, Morrow DA. B-type natriuretic peptide measurementin acute coronary syndromes: ready for clinical application?Circulation 2002;106:2868-70.

19. Foote RS, Pearlman JD, Siegel AH, et al. Detection of exercise-induced ischemia by changes in B-type natriuretic peptides. J AmColl Cardiol 2004;44:1980-7.

20. Sabatine MS, Morrow DA, de Lemos JA, et al. Acute changes incirculating natriuretic peptide levels in relation to myocardialischemia. J Am Coll Cardiol 2004;44:1988 -95.

21. Toth M, Vuorinen KH, Vuolteenaho O, et al. Hipoxia stimulatesrelease of ANP and BNP from perfused rat ventricular myocardium.Am J Physiol 1994;266:H1572 -80.

22. Goetze JP, Christoffersen C, Perko M. Increased cardiac BNPexpression associated with myocardial ischemia. FASEB J 2003;17:1105-7.