• 
• 

Three Part Question

In [patients with suspected cardiac chest pain] does [measurement of N-terminal-pro-brain natriuretic peptide] enable [exclusion of acute coronary syndromes]?

Clinical Scenario

Recognising the limitations of troponin testing in suspected acute coronary syndromes you investigate possible alternatives within the literature. Having reviewed the evidence for brain natriuretic peptide (BNP), you discover that N-terminal-pro-BNP may be superior. You wonder if there is any evidence for its effectiveness.

Search Strategy

OVID Medline 1966 - 2005 July Week 4
OVID Embase 1980 - 2005 Week 32
The Cochrane Library 2005 Issue 2
Medline and Embase:
(exp Myocardial Infarction/ OR exp Coronary Thrombosis/ OR exp Angina, Unstable/ OR (myocard$ adj (infarct$ OR ischaem$ OR ischem$)).mp. OR (acute coronary syndrome OR ACS OR MI OR AMI).mp.) AND (exp Natriuretic Peptide, Brain/ OR $BNP.mp. OR ((brain OR B) adj natriuretic peptide$).mp.) limit to human and English language
Cochrane:
(Myocardial Infarction [MeSH] OR Angina, Unstable [MeSH] OR (myocard* NEAR (infarct* OR ischaem* OR ischem*)) OR acute coronary syndrome OR ACS OR MI OR AMI) AND (Natriuretic Peptide, Brain [MeSH] OR BNP OR ((brain OR B) NEAR (natriuretic peptide*)))

Search Outcome

Medline: 294 papers were identified, of which eight were relevant.
Embase: 354 papers were identified, of which eight were relevant.
Cochrane: 37 papers were identified, of which one was relevant.
Altogether 8 relevant papers were identified.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Campbell et al 2001 Australia	201 consecutive patients presenting to the Emergency Department with suspected ACS (89% of whom had chest pain). Blood sent at presentation. 36 patients with confirmed ACS had repeat NT-pro BNP an average of 5 hours later.	Prospective observational cohort	NT-pro BNP levels in patients with heart failure	Raised in 64 of 67 (96%) of patients with past history of heart failure and 36 of 38 (95%) of patients with radiological evidence of heart failure; Elevated in 59 of 132 (45%) of patients without either of these.	Diverse symptomatology on presentation. May have been better applied in a specific population with chest pain. High study prevalence of cardiac pathology (87.8%). Only a small proportion of patients had a repeat NT-pro BNP sent, making this analysis less meaningful.
			NT-pro BNP levels in patients with ACS	Raised in 48 of 97 (49.5%).
			NT-pro BNP levels in patients with any cardiac origin (including ACS, dysrhythmia and heart failure) v. non-cardiac origin	Calculated: Sensitivity 51%, specificity 82%, PPV 95.4%, NPV 18.9%, LR+ 2.85, LR- 0.60. Thus in a population with a prevalence of 15%, expected post-test probability following -ve test would be 9.8%.
			Time course of NT-pro BNP rise	Levels rose by 40% or more in approximately 1/3 of patients.
Omland et al 2002 Norway	All patients had been enrolled in the TIMI-11B trial (enoxaparin v. heparin in NSTEACS). 53 patients with death (n=22) or non-fatal MI (n=31) by 43 days after enrolment and 53 age- and sex-matched controls. All included patients had blood taken 12-24 hours after admission.	Nested case-control study	NT-proBNP levels in cases with death	Significantly higher in cases than controls (P=0.032 after adjustment for relevant variables).	ACS already ruled in at enrolment Small numbers, no sample size calculation. Unusual design for a diagnostic study. Case-control studies are subject to a series of considerable potential biases. Late sample time. Unmatched cases and controls: 8 cases had heart failure v. 3 controls.
			NT-proBNP levels in cases with non-fatal MI	No significant differences between cases and controls. (Of note, TnI levels and ST depression on the ECG were associated with non-fatal MI).
James et al 2003 Sweden	6809 (87.3%) patients enrolled into the GUSTO-IV trial (patients with NSTEACS randomised to abciximab or placebo). Blood taken at randomisation (mean 9.5 (5.0 to 16.5) hours from symptom onset). Follow up at 30 days and 1 year.	Nested prospective observational cohort	Mortality according to NT-proBNP	Significantly increased mortality with increasing quartiles of NT-pro BNP at 48 hours (P=0.001) and 1 year (P<0.001). Mortality 0.4% in lowest decile and 27.1% in highest decile. As dichotomous variable (cut-off 669ng/L), OR 1.97 (95% CI 1.48-2.61). NT-proBNP an independent predictor of mortality on multivariable logistic regression analysis.	ACS had already been ruled in at the time of enrollment (either by ECG or gold standard cardiac marker elevation).
			Incidence of MI at follow-up	Significantly increased risk of MI at 30 days with increasing quartiles of NT-proBNP (P<0.001). NT-proBNP not an independent predictor on multivariable logistic regression analysis.
			Combination of NT-proBNP and CCl for prediction of mortality	Lowest NT-proBNP quartile and highest CCl quartile: 0.3% 1-year mortality. Highest NT-proBNP quartile and lowest CCl quartile: Mortality 25.7%.
Bazzino et al 2004 Argentina	1483 consecutive patients with resting chest pain within 24 hours, admitted to CCU. Excluded if eligible for thrombolysis. Admitted median 3.2 hours following symptom onset. Blood sent median 3.0 hours following admission. Follow-up at 180 days.	Multicentre prospective observational cohort	Death in hospital	NT-proBNP an independent predictor (P<0.01) following logistic regression analysis	Patients eligible for thrombolysis excluded (no reason given). Enrolled following admission to CCU - ACS already ruled in. No robust criteria for diagnosis of NSTEACS.
			MI in hospital	NT-proBNP not an independent predictor following logistic regression analysis (but TnT was, P=0.042).
			Death by 180 days	NT-proBNP an independent predictor (P<0.001) following logistic regression analysis (of note, TnT not an independent predictor of death at 180 days)
			MI by 180 days	NT-proBNP not an independent predictor (but TnT was, P=0.041).
Gill et al 2004 New Zealand	24 patients admitted to CCU with STEMI within 6 hours of symptom onset. Blood taken at 3.9+/-0.3 hours from symptom onset and serially thereafter.	Prospective observational cohort	NT-BNP v. other natriuretic peptide levels	Higher than BNP and ANP (P<0.0001), but lower than NT-ANP levels. At time 0, identical percentages of patients had BNP, NT-BNP, ANP or NT-ANP above the normal range (29%).	Small numbers MI had already been ruled in at enrolment. No clinical follow-up data. Baseline characteristics not reported. Correlation with LV function not examined.
			Time to peak levels	NT-BNP peaked at 12-24 hours. Troponin, CK-MB and myoglobin all peaked earlier (1-10 hours).
			NT-BNP levels at 8, 12 and 24 hours	Levels above normal range raised in 95% (8 hours), 96% (12 hours) and 96% (24 hours) of patients.
Jernberg et al 2004 Sweden	775 patients with suspected ACS and no STE admitted to CCU, enrolled in the FAST study. Blood taken on admission and at 6 hours. Follow-up at median 40 months.	Nested prospective diagnostic cohort study	Mortality according to NT-proBNP levels	Compared with lowest quartile, patients in 2nd, 3rd & 4th quartiles had RR (95% CI) of 4.2 (1.6-11.1), 10.7 (4.2-26.8) and 26.6 (10.8-65.5), respectively	Only 53% of patients had ACS, but NT-proBNP not evaluated as a diagnostic test for ACS. The investigators aimed to study an unselected population but only patients admitted to CCU were included. LV function not assessed or accounted for.
			NT-proBNP according to diagnosis	Significantly higher NT-pro BNP in unstable angina/angina, AMI and other cardiac causes v. non-cardiac or unknown cause (All P<0.001).
			Association between mortality and NT-proBNP following multivariate analysis	NT-proBNP remained an independent predictor of mortality
Heeschen et al 2004 Germany	1791 eligible (of 3232) patients enrolled in the PRISM trial (tirofiban v. heparin in UA)	Prospective observational cohort	Death or MI within 48 hours	No significant difference between NT-pro-BNP quartiles	Baseline blood samples only available for 55.4% of the study population Nested study – ACS already ruled in at enrolment Insufficient data presented to allow calculation of clinically meaningful values, such as specificity, NPV, PPV and LR
			Death or MI at 7 and 30 days	Significantly higher in 3rd (P=0.008, P=0.002 respectively) and 4th quartiles (P=0.006 and P<0.001 respectively)
			Death or MI (dichotomous; NT-pro-BNP>240ng/l	At 48 hours sensitivity 78.9%; 7 days 76.8%; 30 days 77.2%. Predictive value maintained after adjustment for baseline characteristics, CRP and TnI (all P<0.001)
			Combination of NT-pro-BNP and TnT	NT-proBNP identified patients at risk despite low TnT (crude event rates 15.1% v. 5.6%, OR 3.0, P=0.004).
			Value of repeat NT-pro-BNP at 72 hours	Initial NT-pro-BNP +ve: Cardiac events at 30 days in 17.2% (repeat NT-pro-BNP +ve) and 0.6% (repeat NT-pro-BNP –ve), P<0.001
Galvani et al 2004 Italy	1756 (89.1%) of patients enrolled in the EMAI study (investigating biomarkers in ACS). Eligible patients had resting chest pain plus ischaemic ECG changes. Blood sent at a median of 3.0 (1.8 to 6) hours after symptom onset.	Prospective observational cohort	30-day mortality according to NT-proBNP levels	Increased directly across quartiles of NT-proBNP (P<0.0001). Association persisted following stratification based on TnT levels , especially if TnT <0.1ug/L (P<0.0001). NT-proBNP an independent predictor of mortality following logistic regression analysis. At a cut-off of 437ng/L, sensitivity 78.9%, specificity 56.7%	Patients already had ACS ruled in by ischaemic ECG changes. Application in an undifferentiated group would be more relevant.
			Recurrent ischaemic events by 30 days	Increased according to NT-proBNP quartiles (P=0.005). NT-proBNP not an independent predictor.

Comment(s)

The natriuretic peptide BNP is predominantly secreted by the ventricles in response to ventricular wall stress, initially in the form of a prohormone, proBNP. Following cleavage by the enzyme furin, this prohormone is split into BNP and NT-proBNP. It is unknown whether NT-proBNP has biological effects but it may have a longer half life in the circulation than BNP (Hall, 2004). Gill et al reported a greater rise in NT-proBNP after AMI than BNP or ANP, tentatively suggesting that NT-proBNP may be superior to BNP in this situation. Four studies have shown NT-proBNP to be an independent predictor of mortality in patients with ACS. In addition, Heeschen et al found that raised NT-proBNP predicted death or MI at 48 hours with a sensitivity of 78.9%. The same study reported that NT-proBNP may identify a population of patients who are at risk despite a negative TnI. Bassan et al reported that NT-proBNP had a superior sensitivity to TnI on admission for detection of NSTEMI (70.8% v. 50.7%). On the other hand, four studies reported that NT-proBNP was not an independent predictor of recurrent MI. In addition, the optimum time for testing NT-proBNP is yet to be established. Gill et al found that levels peaked at 12-24 hours, later than troponin. Campbell et al found that levels rose significantly in a third of patients when re-tested 5 hours after admission. While this does not preclude NT-proBNP from consideration for use as a diagnostic test in the Emergency Department, it may be that serial measurements enhance its predictive value, as shown by Heeschen et al. Three additional studies were identified that did not answer the three-part question but are relevant to the discussion. Talwar et al demonstrated raised NT-proBNP in UA, while Schnabel et al reported that baseline NT-proBNP levels predicted future cardiovascular events in patients with stable and unstable angina. Finally, Omland et al found that NT-proBNP is a powerful indicator of long-term mortality in 609 patients with ACS. Only one study, by Campbell et al, investigated the application of the test in an Emergency Department population. However, broad inclusion criteria limited the value of this study (3% of patients had tachydysrhythmia, 4% had heart failure as their primary diagnosis). In addition, the study prevalence of cardiac diagnoses was 87.8%, compared with approximately 15% in a typical undifferentiated Emergency Department population. As such, although the authors reported an unacceptably high NPV of 18.9%, the test may be expected to perform slightly better in the real-life clinical situation. NT-proBNP shows promise as a cardiac marker and may be superior to BNP, although further evidence is required. A large prospective observational study of Emergency Department patients with undifferentiated chest pain is necessary to further evaluate its utility. Serial estimations may enhance its value. In addition, measurement of other promising cardiac markers and incorporation into a multimarker strategy may yield complimentary information, enhance sensitivity and minimise the false negative diagnosis rate (NPV).

Editor Comment

Abbreviations: UA: Unstable angina; TnT: troponin T; OR: Odds ratio; NPV: Negative predictive value; PPV: Positive predictive value; LR: Likelihood ratio; MI: Myocardial infarction; NSTEACS: Non-ST elevation acute coronary syndrome; CI: Confidence interval; CCl: Creatinine clearance; BNP: brain natriuretic peptide; AMI: acute myocardial infarction; NSTEMI: non-ST elevation myocardial infarction.

Clinical Bottom Line

NT-proBNP is a promising cardiac marker for use in the Emergency Department. Incorporation into a multimarker strategy and serial estimations would be expected to enhance its value.

Level of Evidence

Level 2 - Studies considered were neither 1 or 3.

References

1. Campbell DJ; Munir V; Hennessy OF; Dent AW Plasma amino-terminal pro-brain natriuretic peptide levels in subjects presenting to the Emergency Department with suspected acute coronary syndrome: possibl erole in selecting patients for follow up Internal Medicine Journal 2001; 31: 211-219
2. Omland T; de Lemos JA; Morrow DA; Antman EM; Cannon CP; Hall C; Braunwald E Prognostic value of N-terminal pro-atrial and pro-brain natriuretic peptide in patients with acute coronary syndromes American Journal of Cardiology 2002; 89: 463-465
3. Omland T; Persson A; Ng L; O'Brien R; Karlsson T; Herlitz J; Hartford M; Caidahl K N-terminal pro-B-type natriuretic peptide and long-term mortality in acute coronary syndromes Circulation 2002; 106: 2913-2918
4. James SK; Lindahl B; Siegbahn A; Stridsberg M; Venge P; Armstrong P; Barnathan ES; Califf R; Topol EJ; Simoons ML; Wallentin L N-terminal pro-brain natriuretic peptide and other risk markers for the separate prediction of mortality and subsequent myocardial infarction in patients with unstable coronary disease Circulation 2003; 108: 275-281
5. Bazzino O; Fuselli JJ; Botto F; de Arenaza DP; Bahit C; Dadone J; for the PACS group of investigators Relative value of N-terminal probrain natriuretic peptide, TIMI risk score, ACC/AHA prognostic classification and other risk markers in patients with non-ST elevation acute coronary syndromes European Heart Journal 2004; 25: 859-866
6. Gill D; Seidler T; Troughton RW; Yandle G; Frampton CM; Richards M; Lainchbury JG; Nicholls G Vigorous response in plasma N-terminal pro-brain natriuretic peptide (NT-BNP) to acute myocardial infarction Clinical Science 2004; 106: 135-139
7. Jernberg T; Stridsberg M; Venge P; Lindahl B N-terminal pro brain natriuretic peptide on admission for early risk stratification of patients with chest pain and no ST-semgent elevation Journal of the American College of Cardiology 2004; 6: 319-325
8. Heeschen C; Hamm CW; Mitrovic V; Lantelme NH; White HD; for the Platelet Inhibition in Ischemic Syndrome Management (PRISM) Investigators N-terminal pro-B-type natriuretic peptide levels for dynamic risk stratification of patients with acute coronary syndromes Circulation 2004; 110: 3206-3212
9. Schnabel R; Rupprecht HJ; Lackner KJ; Lubos E; Bickel C; Meyer J; Munzel T; Cambien F; Tiret L; Blankenberg; for the AtheroGene Investigators Analysis of N-terminal-pro-brain natriuretic peptide and C-reactive protein for risk stratification in stable and unstable coronary artery disease: results from the AtheroGene study European Heart Journal 2005; 26: 241-249
10. Galvani M; Ottani F; Oltrona L; Ardissino D; Gensini GF; Maggioni AP; Mannucci PM; Mininni N; Prando MD; Tubaro M; Vernocchi A; Vecchio C; on behalf of ANMCO N-terminal pro-brain natriuretic peptide on admission has prognostic value across the whole spectrum of acute coronary syndromes Circulation 2004; 110: 128-134
11. Lindahl B; Lindback J; Jernberg T; Johnston N; Stridsberg M; Venge P; Wallentin L Serial analyses of N-terminal pro-B-type natriuretic peptide in patients with non-ST-segment elevation acute coronary syndromes Journal of the American College of Cardiology 2005; 45(4): 533-541
12. Talwar S; Squire IB; Downie PF; Davies JE; Ng LL; Baxter GF. Plasma N terminal pro-brain natriuretic peptide and cardiotrophin 1 are raised in unstable angina Heart 2000; 84: 421-424