B iomarkers in the A ssessment of C ongestive H eart Failure - - PowerPoint PPT Presentation

Biomarkers in the Assessment of Congestive Heart Failure

Mid-Regional pro-Adrenomedullin (MR-proADM) vs BNP &

NT-proBNP as Prognosticator in Heart Failure Patients: Results of the BACH Multinational Trial

PI: Alan Maisel, San Diego / USA Co-PI: Stefan D Anker, Berlin / GER (presenting)

Disclosure Stefan D Anker: Consulting: BRAHMS Aktiengesellschaft Research Support: BRAHMS Aktiengesellschaft, Biosite/Inverness

Struck J et al., Peptides. 2004; Morgenthaler NG et al., Clin Chem. 2005

Pro-Adrenomedullin 24 – 71

ELRMSSSYPTGLADVKAGPAQTLIRPQDMKGASRSPEDSSPDAARIRV

NH2-

O

//

• C

\ OH

Adrenomedullin:

• Peptide hormone consisting of 52 amino acids
• vasodilator, important for microcirculation & endothelial (dys)function
• quantification of ADM could be helpful in cardiac diseases (CHF, MI,

ACS,...) as well as in others (like infections & kidney disease)

• ADM measurement is not suitable for clinical routine diagnosis assessment

due to its ex vivo instability (immediate binding to receptors, 22min half-life time)

• mid regional pro-ADM (MR-proADM) is a stable and reliable surrogate

marker for ADM release

1 24 71 74 125 164

N-terminal peptide

MR-proADM

Adrenomedullin PAMP

MR-proADM is a reliable surrogate marker of Adrenomedullin

Methods

QUESTIONS:

• What is the preferred (best) clinical tool for prognostication in HF?
• Does measuring of MR-proADM, better predict prognosis than BNP or NT-proBNP?
• 1° endpoint: Is MR-proADM superior in predicting 90-day mortality compared to BNP?

PATIENTS & FOLLOW-UP

• Patients included who presented to ED with SOB not from trauma, or obvious MI & not on

dialysis – then MD assessment for heart failure

• Follow-up for 90 days for survival, main outcome “All cause mortality within 90 days”
• 15 enrolling centers (US, Europe, NZL), recruiting 1641 patients.
• N=1641, Age 64±17 yrs, Female 48%, History of HF 36%, prior AMI 19%, Diabetes 29%,

BMI 29±8 kg/m2

Admitted

443

Discharged

628

Deceased

2

Admitted

477

Discharged

89

Transfers

2

Non-AHF

1073

AHF

568

Enrolled

Adjudicated Diagnosis ED Disposition 1641

Primary Endpoint = Successful - MR-proADM & mortality in AHF

AUC 30 days 90 days MR-proADM 0.739 0.674 NT-proBNP 0.641 0.664 BNP 0.555 0.606

Predictor (univariate) Chi2 Statistic p c index log MR-proADM 31.0 <0.001 0.669 log BNP 7.1 0.008 0.596 log NT-proBNP 17.1 <0.001 0.654

Results of Cox regression with continuous predictors:

MR-proADM is superior to BNP and NT-proBNP for predicting 90-day mortality (Cox regression).

Chi2 Statistic p adding MR-proADM to BNP 23.9 <0.0001 adding MR-proADM to NT-proBNP 15.3 <0.0001 adding BNP to MR-proADM 0.0 0.906 adding NT-proBNP to MR-proADM 1.1 0.291

MR-proADM adds significantly to BNP or NT- proBNP, however neither BNP nor NT-proBNP add to MR-proADM.

Prognostic Accuracy: MR-proADM 73.5% vs BNP 60.8%

(p<0.001)

vs NT-proBNP 63.6% (p<0.001)

Survival in AHF - Troponin adjustment & MR-proADM Quartiles

Days

20 40 60 80

Cumulative Survival

0.75 0.80 0.85 0.90 0.95 1.00

3rd 4th 2nd 1st

Risk is greatest in the highest quartile

• f MR-proADM

Days

20 40 60 80

Cumulative Survival

0.75 0.80 0.85 0.90 0.95 1.00

4th 1st-3rd

MR-proADM < 2.07 nmol/L

MR-proADM

≥ 2.07 nmol/L HR 3.3– P<0.001

(95% CI 2.0-5.4)

Troponin values were available in 511 of 568 HF patients – in 107 (20.9%) patients they were elevated (i.e. above local normal range). Shown are results of 3-marker models. Conclusion: MR-proADM provides independent prognostic utility, but BNP & NT-proBNP do not.

Predictor (multivariable) HR 95% CI p log MR-proADM 8.5 2.7-26.5 <0.001 log BNP 0.9 0.5-1.9 0.812 Elevated Tn 2.6 1.5-4.5 <0.001 Predictor (multivariable) HR 95% CI p log MR-proADM 7.5 2.1-26.4 <0.001 log NT-proBNP 1.1 0.6-2.2 0.295 Elevated Tn 2.6 1.5-4.4 <0.001

Analysis of MR-proADM quartiles:

Survival in all patients with SOB - Utility of MR-proADM

Cox Regression Analysis:

MR-proADM performs well in all SOB patients.

Predictor (univariate) Chi2 Statistic p c index log MR-proADM 129.5 <0.0001 0.755 log BNP 60.1 <0.0001 0.691 log NT-proBNP 83.7 <0.0001 0.721 Chi2 Statistic p adding MR-proADM to BNP 69.4 <0.0001 adding MR-proADM to NT-proBNP 46.6 <0.0001 adding BNP to MR-proADM 0.1 0.731 adding NT-proBNP to MR-proADM 1.5 0.229

MR-proADM is superior to BNP and NT-proBNP.

1 - Specificity

0.0 0.2 0.4 0.6 0.8 1.0

Sensitivity

0.0 0.2 0.4 0.6 0.8 1.0

Measurement AUC 95% CI p MR-proADM 0.761 0.719-0.803 <0.0001 BNP 0.698 0.654-0.742 <0.0001 NT-proBNP 0.729 0.685-0.773 <0.0001

Measurement AUC 95% CI p MR-proADM 0.762 0.721-0.804 <0.0001 BNP 0.699 0.655-0.743 <0.0001 NT-proBNP 0.729 0.685-0.773 <0.0001

Survival in patients without AHF - Utility of MR-proADM Elevated MR-proADM is strongly prognostic in patients with and without AHF – even more so in non-AHF than in AHF (interaction p=0.005).

MR-proADM 0.674 1.985 nmol/l AHF patients AUC (90 days)

• ptimal cut point

from ROC

<0.001 3.6-8.9

5.7

high > 1.985 0.027 1.1-2.7

1.7

low < 1.985

AHF

<0.001 5.1-14.4

8.6

high > 1.985 reference

1

low < 1.985

Non-AHF

p 95% CI

HR

MR-proADM Diagnosis

Days

20 40 60 80

Cumulative Survival

0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00

MR-proADM < 1.985 nmol/l MR-proADM ≥ 1.985 nmol/L

AHF Non-AHF AHF Non-AHF

• MR-proADM is a strong prognosticator in patients with AHF and

in patients presenting with SOB.

• MR-proADM is superior to BNP or NT-proBNP for predicting

90-day mortality, both in AHF as well as in all ED pts with SOB.

• All these results are unaffected by adjustment for Troponin.
• MR-proADM is particularly strong in predicting short-term

prognosis within 4 weeks after assessment.

• MR-proADM can significantly improve risk stratification over

BNP or NT-proBNP.

• Assessment of MR-proADM can help to identify patients who

should “move to the front of the line” of medical care.

Summary