Author + information
- Received August 24, 2009
- Revision received January 7, 2010
- Accepted January 18, 2010
- Published online July 6, 2010.
- Onkar S. Dhillon, BSc, MBChB*,* (, )
- Sohail Q. Khan, BSc (Hons), MBChB, MD*,
- Hafid K. Narayan, MBChB*,
- Kelvin H. Ng, MBChB*,
- Joachim Struck, PhD†,
- Paulene A. Quinn, MPhil*,
- Nils G. Morgenthaler, MD†,
- Iain B. Squire, MD*,
- Joan E. Davies, PhD*,
- Andreas Bergmann, PhD† and
- Leong L. Ng, MD*,* ()
- ↵*Reprint requests and correspondence:
Dr. Onkar S. Dhillon or Prof. Dr. Leong L. Ng, Department of Cardiovascular Sciences, Clinical Sciences Building, Leicester Royal Infirmary, Leicester LE2 7LX, United Kingdom
Objectives The purpose of this study was to assess the prognostic value of admission and discharge mid-regional pro-adrenomedullin (sAM) levels in non–ST-elevation myocardial infarction (MI) and identify values to aid clinical decision making. N-terminal pro–B-type natriuretic peptide and GRACE (Global Registry of Acute Coronary Events) score were used as comparators.
Background sAM is a stable precursor of adrenomedullin.
Methods We measured plasma sAM on admission and discharge in 745 non–ST-elevation MI patients (514 men, median age 70.0 ± 12.7 years). The primary end point was a composite of death, heart failure, hospitalization, and recurrent acute MI over mean follow-up of 760 days (range 150 to 2,837 days), with each event assessed individually as secondary end points.
Results During follow-up, 120 (16.1%) patients died, and there were 65 (8.7%) hospitalizations for heart failure and 77 (10.3%) recurrent acute MIs. Both admission and discharge levels were increased (median 0.81 nmol/l [range 0.06 to 5.75 nmol/l] and 0.76 nmol/l [range 0.25 to 6.95 nmol/l], respectively) compared with established normal ranges. Multivariate adjusted Cox regression models revealed that both were associated with the primary end point (hazard ratio: 9.75 on admission and 7.54 on discharge; both p < 0.001). Admission sAM was particularly associated with early (<30 days) mortality (c-statistic = 0.90, p < 0.001), and when compared with N-terminal pro–B-type natriuretic peptide and GRACE score, it was the only independent predictor of this end point. Admission sAM >1.11 nmol/l identified those at highest risk of death (p < 0.001). Patients with above-median admission sAM may benefit from revascularization.
Conclusions sAM level is prognostic for death or heart failure. Admission levels are a strong predictor of early mortality and, when >1.11 nmol/l, complements the GRACE score to improve risk stratification.
- Global Registry of Acute Coronary Events score
- myocardial infarction
- N-terminal pro–B-type natriuretic peptide
Non–ST-segment elevation myocardial infarction (NSTEMI) now has a higher annual prevalence than ST-segment elevation myocardial infarction (STEMI) (1), and although in-hospital mortality rates are lower, they gradually increase and in the long term are higher than in STEMI patients (2). In this heterogeneous disorder, identifying where an individual lies on the spectrum of risk is crucial to guide appropriate management (3). For example, risk stratification may be useful when balancing the risk and potential benefit of invasive treatment in an aging population. Promising risk stratification tools and biomarkers have emerged such as the GRACE (Global Registry of Acute Coronary Events) score (4) and N-terminal pro–B-type natriuretic peptide (NT-proBNP) (5), but neither is fully discriminatory. A multimarker strategy may become the most effective tool for risk prediction (6), and new markers should offer new information over available clinical and biochemical factors.
Adrenomedullin (AM) is a 52-amino-acid peptide that is elevated in heart failure (HF) (7) and post–acute coronary syndrome (ACS) (8). Discovered in human phaechromocytoma, it is also present in heart, brain, lung, kidney, and gastrointestinal organs and elicits its potent vasodilatory activity (9) through an increase in cyclic adenosine monophosphate levels. AM increases cardiac output and induces a diuresis and natriuresis (10), and the elevated levels found in HF and myocardial infarction (MI) are also predictive of adverse events (11). Despite this clear role in cardiovascular homeostasis and pathogenesis, further investigation of plasma AM levels is hampered by its short half-life (22 min), rapid clearance, and the presence of a binding protein (12). AM is derived from a larger (185 amino acid) precursor molecule (13). A more stable mid-regional fragment composed of amino acids 45 to 92 of pre-pro-AM has been identified, which is secreted in equimolar amounts (12), and therefore acts as a surrogate for adrenomedullin (sAM) levels. Our center (14) previously showed that this mid-regional fragment has independent prognostic value but in a cohort of primarily STEMI patients with levels measured solely on discharge. In this study, we addressed the use of mid-regional pro-AM (sAM) in NSTEMI and compared the prognostic value of admission and pre-discharge levels for major adverse cardiac events (MACE) using NT-proBNP and the GRACE score as benchmark comparators. Furthermore, we identified a cutoff level that could aid clinical decision making.
We studied 745 NSTEMI patients admitted to University Hospitals of Leicester NHS trust and primarily recruited between August 2005 and April 2007. Attempts were made to enroll all consecutive NSTEMI patients, but approximately 20 to 40 declined. The study complied with the Declaration of Helsinki and was approved by the local ethics committee; written informed consent was obtained from all patients. AMI was diagnosed if a patient had a cardiac troponin I level above the 99th centile for our population with at least one of the following: chest pain lasting >20 min or diagnostic serial electrocardiographic changes consisting of new pathological Q waves or ST-segment and T-wave changes (15). Exclusion criteria were known malignancy, renal replacement therapy, or surgery in the previous month. Demographic, clinical, biochemical, and echocardiographic data were obtained. The estimated glomerular filtration rate (eGFR) was calculated from the simplified formula derived from the Modification of Diet in Renal Disease study (16). All patients received standard medical treatment, and revascularization was at the discretion of the attending physician.
Blood samples were drawn after 15 min of bed rest into tubes containing ethylenediamine tetraacetic acid and aprotinin. Admission levels refer to those obtained after diagnosis of NSTEMI but within 36 h of onset of symptoms, and discharge levels were taken 3 to 5 days post-admission. Plasma was stored at −80°C until assayed in a single batch for determination of plasma sAM and NT-proBNP with both clinicians and patients blinded to the results.
Transthoracic echocardiography was performed in 501 (67.2%) patients during the index admission but after day 3 using either a Sonos 5500 or IE 33 instrument (Philips Medical Systems, Reigate, United Kingdom). A 16-segment left ventricular wall motion index score was determined based on the American Society of Echocardiography method. In suitable patients, the left ventricular ejection fraction was calculated using the biplane method of disks formula. Impaired left ventricular systolic function was defined as either a left ventricular ejection fraction <40% or a ventricular wall motion index score >1.8.
Based on an international observational database of ACS patients, GRACE scores can be calculated on initial presentation to primarily predict in-hospital mortality and reinfarction rates (4). When calculated before discharge, it includes age, heart rate, systolic blood pressure, serum creatinine, congestive heart failure, ST-segment depression, increased cardiac enzymes, history of MI or coronary artery bypass graft, and whether in-hospital revascularization was performed, and it shows superior estimation of 6-month mortality and reinfarction rates. We calculated both scores, but only used scores on initial presentation for comparison with events at 30 days.
Our NT-proBNP assay was based on a noncompetitive assay, as previously published (5). The lower limit of detection was 0.3 pmol/l. Pre-discharge levels of NT-proBNP were used unless otherwise specified.
sAM was detected using a novel commercial assay (BRAHMS AG, Hennigsdorf, Germany), as previously described (17). Briefly, tubes were coated with a purified sheep polyclonal antibody raised against a peptide representing amino acids 83 to 94 of pre-pro-AM. A purified sheep methylacridinium N-hydroxysuccinimide ester–labeled polyclonal antibody directed against amino acids 68 to 86 of pre-pro-AM was used as a tracer. The immunoassay was performed by incubating 10 μl of samples/standards and 200 μl of tracer in coated tubes for 2 h at room temperature and measuring the bound chemiluminescence using an LB952T luminometer (Berthold, Bad Wildbad, Germany). The lower detection limit of the assay is 0.08 nmol/l. The intra-assay (interassay) coefficients of variation at 0.5 and 5 nmol/l were 3% (8.5%) and 3.5% (6.5%), respectively.
The primary composite end point was MACE, including all-cause mortality, HF hospitalization, or recurrent AMI, all evaluated individually as secondary end points within 30 days, 6 months, and 2 years. This contrasts with the previous study (14) from our center that reported death or readmission with HF as the primary outcome measure because NT-proBNP, not GRACE score, was used as the primary comparator and has a recognized association with death and HF readmission (14). However, GRACE score estimates the risk of death or recurrent MI (4) and to enable nonbiased comparison with this marker as well, readmission with AMI was added to the primary end point. Hospitalization for HF was defined as a hospital readmission for which HF was the primary reason, requiring treatment with high-dose diuretics, inotropes, or intravenous nitrate. Recurrent AMI could occur on index admission or on rehospitalization and was diagnosed if the cardiac troponin I level was above the 99th centile for our population with at least 1 of the following: chest pain lasting >20 min or diagnostic serial electrocardiographic changes consisting of new pathological Q waves or ST-segment and T-wave changes (15). End points were obtained by reviewing the local hospital databases and the Office of National Statistics Registry and by phone calls to patients. We achieved 100% follow-up.
Statistical analyses were performed using SPSS version 14 (SPSS Inc., Chicago, Illinois). Assuming an event rate of 15% and that the covariates predict up to 30% of the variance of the biomarker, a sample size of 600 patients would be powered (89.6% at p < 0.05) to detect a hazard ratio (HR) of the biomarker of 1.5. We recruited >600 patients to cover for loss to follow-up. sAM and NT-proBNP levels were log10-transformed, and therefore, HRs refer to a 10-fold increase in the levels of these markers. GRACE scores were used as the original scores. Nonparametric tests were used against non-Gaussian data (Mann-Whitney Utest, Kruskal-Wallis test, and Spearman [rs] correlations). Correction for multiple comparisons was done with Hochberg's GT2 test, and all significance levels quoted are 2 tailed. Independent predictors of sAM levels were assessed using univariate general linear models. To assess the prognostic value of the peptides, 2 multivariate Cox regression models were constructed. First, a base model was generated that included variables that were significantly (p < 0.20) associated with any of the study end points on univariate analysis (age, sex, history of angina, AMI, hypertension or diabetes, Killip class, eGFR, log glucose level, and log troponin I [with addition of echocardiographic data in a separate model]). Either admission sAM or discharge sAM levels were added to this model to evaluate the relative prognostic value of each with all variables entered simultaneously. A second comparative Cox model was used to assess the relative prognostic power of sAM, NT-proBNP, and GRACE score. This model excluded age and eGFR because both are included in calculating GRACE scores. Receiver-operator characteristic (ROC) curves with c-statistics were calculated. Kaplan-Meier survival analysis and Mantel-Cox log-rank tests were performed. The additional prognostic value of sAM taken at admission or discharge to the GRACE score was evaluated by reclassification analysis with calculation of net reclassification improvement (NRI) as described by Pencina et al. (18).
We sought interactions between the plasma levels of sAM and revascularization in Cox models to investigate the usefulness of this biomarker in directing intervention. As these were observational data, models were corrected for the probability of receiving revascularization (the propensity score, which was calculated as the probability of revascularization using the following predictive variables: age, sex, history of MI, angina, hypertension, diabetes, Killip class, eGFR, site of MI, drug therapies).
The characteristics of the study population are shown in Table 1,of which 109 (14.6%) were from the Indian subcontinent. Of the 745 patients, 631 had sAM levels taken within 36 h of symptoms, 684 had levels obtained before discharge, and 700 had NT-proBNP levels. During a follow-up of a mean of 760 days (range 150 to 1,059 days), 120 (16.1%) patients died, 65 (8.7%) were readmitted with HF, and 77 (10.3%) had a reinfarction.
sAM levels in NSTEMI patients
Both admission and discharge sAM levels were increased (median levels of 0.81 nmol/l [range 0.06 to 5.75 nmol/l] and 0.76 nmol/l [range 0.25 to 6.95 nmol/l], respectively) compared with a disease-free cohort (721 people of whom 417 were male, mean age 61.6 years; median 0.47 nmol/l, range 0.27 to 1.18 nmol/l). Independent predictors of admission sAM showed the following significant predictors: age, eGFR, history of diabetes, Killip class >1, and loop diuretic therapy (all p < 0.001). Independent predictors of discharge sAM included age, eGFR, Killip class >1 (all p < 0.001), loop diuretic therapy (p = 0.026), and history of MI (p < 0.02). Beta-blockers were not a significant predictor of sAM levels.
sAM levels were also higher in those who reached the primary end point compared with the event-free survivors (admission sAM 1.11 nmol/l [range 0.06 to 5.75 nmol/l] vs. 0.72 nmol/l [range 0.09 to 2.75 nmol/l], p < 0.001; discharge sAM 1.09 nmol/l [range 0.33 to 6.95 nmol/l] vs. 0.70 nmol/l [range 0.25 to 6.66 nmol/l]; p < 0.001) (Fig. 1).Both showed a comparable relationship to clinical and other variables (Table 2).
NT-proBNP levels in NSTEMI patients
NT-proBNP levels ranged from 0.12 to 24,016 pmol/l, with a mean of 1,318 pmol/l, and showed a typical distribution across variables (Table 2).
Prognostic value of early versus pre-discharge sAM levels
For the primary end point, multivariate Cox analysis showed that admission sAM levels were predictive of MACE (HR: 9.75, p < 0.001) along with age and, in the subset of patients with echocardiographic data, impaired left ventricular function (HR: 1.95, p = 0.001). Discharge sAM was also predictive (HR: 7.54, p < 0.001) along with age and Killip class >1 (HR: 1.49, p = 0.021).
Examination of the relationship with the secondary end points revealed a difference between the prognostic value of sAM taken at the 2 time points (Table 3).High early levels of sAM have a stronger relationship with death than pre-discharge levels (HR: 16.68 vs. 7.70; both p < 0.001). Using Kaplan-Meier analysis, Figure 2shows that when the cohort is separated into 3 groups by admission sAM level, it allowed identification of patients at low, medium, and high risk of death (log-rank test: 75.44, p < 0.001). There is early separation of the survival plot; this represents no deaths in patients in the lowest tertile by 30 days, 1 in the second, and 19 in the highest. Multivariate Cox analysis for death at 30 days revealed that a 1-SD increase in log-transformed admission sAM was associated with a 117-fold increase in hazard (95% confidence interval [CI]: 8.35 to 1,636.18, p < 0.001) compared with discharge sAM (HR: 27.89, 95% CI: 3.77 to 206.53, p = 0.001). Admission sAM was associated with HF readmission (HR: 8.39, p = 0.018) along with Killip class >1 (HR: 2.77, p = 0.003) and impaired left ventricular function (HR: 2.00, p = 0.028), but this was weaker than for discharge sAM (HR: 15.99, p < 0.001). This leads to a statistically similar ability to predict the primary composite end point overall. No association with recurrent MI was observed.
The difference in levels between the 2 sampling times was also evaluated, which revealed a univariate association with death (HR: 2.48, p = 0.002) but no other end points. This relationship was lost on multivariate analysis.
Comparison of prognostic value of admission sAM, GRACE score, and NT-proBNP
Primary Composite End Point: MACE
Using the comparative Cox regression model, all 3 markers were significant independent predictors of the primary end point: NT-proBNP (HR: 1.54, 95% CI: 1.07 to 2.24, p = 0.027), GRACE score (HR: 1.01, 95% CI: 1.01 to 1.02, p = 0.001), and admission sAM (HR: 3.44, 95% CI: 1.25 to 9.50, p = 0.017). Results are shown for the model containing admission sAM but were similar for discharge sAM. ROC curve c-statistics illustrate that overall discriminatory ability for the primary end point is similar among prognostic markers for admission sAM (HR: 0.72, 95% CI: 0.67 to 0.77), NT-proBNP (HR: 0.73, 95% CI: 0.69 to 0.77), and GRACE score (HR: 0.71, 95% CI: 0.66 to 0.76), all at p < 0.001.
Secondary End Points
Table 4shows the relationship between admission sAM, NT-proBNP, and GRACE score and mortality in the long (mean 760 days), medium (6 months), and short term (30 days) by Cox regression analysis. For long-term mortality, NT-proBNP was displaced by admission sAM and GRACE score with no other significant covariates. The GRACE score c-statistic of 0.80 (95% CI: 0.73 to 0.86) could be improved to 0.84 (95% CI: 0.78 to 0.90) when combined with admission sAM (c-statistic 0.82, 95% CI: 0.77 to 0.88), all p < 0.001. Admission sAM is a stronger predictor of mortality at 6 months than the GRACE score in this cohort (HR: 2.45, p = 0.002 and HR: 1.01, p = 0.045, respectively), and it was the only marker to retain independent prediction of death at 30 days (HR: 66.89, p = 0.004; c-statistic: 0.90 [95% CI: 0.85 to 0.95], p < 0.001) (even when both the GRACE scores on admission and pre-discharge were assessed).
To evaluate the additional contribution of admission sAM levels to the GRACE score (calculated on presentation) for prediction of early mortality at 30 days, we performed reclassification analysis (18) with calculation of NRI by the following method. Individuals were classified into risk tertiles using binary logistic regression with the GRACE score as the covariate and then reclassified with admission sAM as a further covariate in the regression model. The change in risk tertiles for the end point of mortality at 1 month is shown in Table 5.Admission sAM improved the classification in 168 subjects who did not experience the end point and made it worse in 36, while improving classification in 2 subjects who experienced the end point, resulting in a significant NRI of 34.1% (range 18.6% to 49.5%), p = 0.00001 (discharge sAM demonstrated an NRI of 23.1% [range 3.6% to 42.6%], p = 0.002). For prediction of early mortality, prognostic markers that can be performed on admission are advantageous; therefore, for the purpose of comparison, we assessed the additional contribution of admissionlevels of NT-proBNP to the GRACE score. Admission NT-proBNP level failed to significantly improve the overall classification of individuals into risk tertiles with a calculated NRI of 1.1% (range −18.9% to 21.1%), p = 0.917.
Significant independent predictors of HF were Killip class >1 (HR: 2.49; 95% CI: 1.29 to 4.80, p = 0.006) and NT-proBNP (HR: 2.02, 95% CI: 1.05 to 3.88, p = 0.036) but not admission sAM (p = 0.215) or impaired left ventricular function (p = 0.074). However, discharge sAM level is a strong predictor of this outcome (HR: 10.11, 95% CI: 2.36 to 43.31, p = 0.002) that displaces NT-proBNP (p = 0.072) when substituted into the model, with Killip class remaining significant (HR: 2.60, p = 0.004). ROC curve analysis showed a c-statistic of 0.74 for both NT-proBNP and discharge sAM, which increased to 0.76 with both markers combined. Because the GRACE score is not designed to predict readmission with HF, it was not evaluated against this end point. Neither biomarkers nor the GRACE score predicted recurrent MI as an individual end point in multivariate analysis.
Complementary use of admission sAM and GRACE score
We assessed whether the strong association of admission sAM with mortality could be used to improve risk stratification by the GRACE score. ROC analyses provided the optimum cutoff value for admission sAM for prediction of death at 1.11 nmol/l (67% sensitivity, 22% false-positive rate) and predicted optimum survival stratification on Kaplan-Meier analysis (log-rank test: 85.19, p < 0.001). Of the 16 patients who died by day 30, 15 (94%) had levels above this value. A total of 564 patients had both GRACE score and admission sAM data. This cohort was separated into centiles by the GRACE score to avoid loss of prognostic value due to oversimplification. The improvement in risk stratification using knowledge of sAM levels was most evident in the top 5 centiles of the GRACE score (log-rank test: 15.71, p < 0.001 pooled comparison) (Fig. 3).For example, for those with GRACE scores in the top centile, an admission sAM level of >1.11 nmol/l identified those with 46.5% mortality rate during the follow-up period and, if <1.11 nmol/l, isolated those with a risk comparable to having a GRACE score in the 6th centile in this cohort (15.4% vs. 16.1%).
Interaction of admission sAM levels and revascularization
The data show that the revascularization by day 30 (Table 1) rates were reduced in high-risk patients at 10.9% in patients with admission sAM levels in the top tertile compared with 28.6% and 30.9%, respectively, for the middle and lowest tertiles of admission sAM (confirming the risk-averse management that is prevalent in many other registries). A propensity score for predicting revascularization was calculated from variables listed in the statistical analysis section. ROC curve analysis of the propensity score for prediction of actual revascularization showed reasonable discrimination (c-statistic: 0.71, p < 0.0005). Cox models for the prediction of MACE incorporating factors in the base model, with the addition of revascularization, admission sAM, and an interaction term (revascularization × admission sAM) and the propensity score for revascularization, yielded the following HR for significant predictors: admission sAM: 3.69 (p = 0.02). The interaction term revascularization*admission sAM was significant (p = 0.02). Kaplan-Meier plots of the effect of revascularization on MACE of patients stratified by the admission sAM median level are presented in Figure 4.Revascularization was associated with reduced MACE in patients whose admission sAM levels were above the median (p = 0.05) but not significantly in those with admission sAM below the median.
Advancing from a previous study from our center (14) that identified the prognostic value of sAM in a primarily STEMI population, this is the first report demonstrating the biomarker's substantial prognostic value in a large cohort of patients with NSTEMI, and it also addresses the optimum time to measure sAM. Levels taken within 36 h of symptoms displayed similar overall prediction for MACE as those taken before discharge but with an important caveat; admission levels are strongly associated with increased risk of death, particularly early mortality, where they are superior to NT-proBNP and the GRACE score, whereas pre-discharge levels demonstrate a stronger association with HF readmission that is comparable to NT-proBNP.
Because AM has shown numerous cardioprotective effects (19,20), it is initially counterintuitive that high levels are associated with adverse events. It has been argued that AM counterbalances other negative processes, which is supported by studies showing AM infusions enhancing left ventricular function (20) with similar beneficial actions seen in patients with HF (21). This paradox mirrors that already seen with B-type natriuretic peptide (BNP) as does the variation in sAM levels across clinical factors, such as age and sex. Furthermore, a significant correlation between the biomarkers (8,14) was found, with them sharing 43% of the same information, and overall one could extrapolate that release of AM and BNP may share the same stimuli. Consequently, multicollinearity may explain why NT-proBNP is displaced from Cox analysis by sAM for both death and HF readmission.
AM has pleiotropic effects both against the arterial response to injury (22) and on left ventricular remodeling (20). An unstable atheromatous plaque underpins the etiology of an ACS, and one could speculate that the link between higher initial levels and mortality may reflect a greater burden of deleterious arterial processes that are being counterbalanced by increased AM production. Levels at discharge showed a stronger relationship with HF readmission because they are likely to relate to the remodeling process that is under way at that time. The potent association with adverse events clearly illustrates the significance of AM as an independent neurohormonal pathway and, in view of its beneficial effects, a potential therapeutic pathway.
AM levels peak 24 to 48 h after onset (8), and the importance of timing of blood sampling for prognostication is illustrated by previous studies in STEMI patients treated by primary percutaneous coronary intervention (23). Katayama et al. (23) concurred with our current findings by also demonstrating AM as a stronger prognostic marker of early mortality than NT-proBNP when levels were measured 24 h after onset.
International guidelines (3) recognize that risk stratification plays a central role in the optimum management of non–ST-segment elevation ACS, which has been compounded by the results of recent studies. For example, Mehta et al. (24) did not find early intervention beneficial compared with a delayed approach unless patients had GRACE scores in the top tertile. The present study demonstrates that using a simple cutoff value of sAM of 1.11 nmol/l can markedly improve risk stratification over GRACE scoring alone.
We also examined the effect of revascularization within 30 days on outcomes, after stratification by sAM levels and adjusting Cox models for propensity scores of likelihood of revascularization. These preliminary data suggest that the admission sAM level may help to identify patients who will benefit most from revascularization. The use of sAM to select patients for invasive or potentially hazardous treatments may be a step toward a personalized management approach, but further prospective, randomized studies are needed to confirm this.
Our findings were based on a single-center population, with 2 admitting hospitals, and would need to be verified in other populations. The rate of early revascularization in our NSTEMI population was low, 23.5% by day 30, and may not reflect a more contemporary invasive approach of revascularization within 72 h of presentation. Correction of hazard models using propensity scores for revascularization may not completely remove bias from unmeasured factors.
This is the first report confirming activation of the AM system in a cohort of patients with NSTEMI. sAM is a stable precursor released in equimolar amounts that represents a powerful new biomarker of risk of adverse events, death, or HR rehospitalization beyond established clinical, biochemical, or echocardiographic markers, NT-proBNP, and GRACE score. Admission sAM level is a particularly strong predictor of early mortality and, when >1.11 nmol/l, complemented the GRACE score to improve risk stratification.
Drs. Dhillon, Khan, and Narayan were supported by British Heart Foundation Junior Research Fellowships(grant numbers FS/03/028/15486, FS/03/028/15486, and FS/09/040, respectively). Dr. Ng was supported by the Leicester National Institute for Health Research Cardiovascular Biomedical Research Unitand has submitted patent applications on behalf of the University of Leicester on biomarkers of cardiovascular disease. Drs. Struck and Morgenthaler are employees of BRAHMS AG, a mid-sized company based in Hennigsdorf, Germany that commercializes immunoassays and has developed the Midregional Pro-Adrenomedullin assay, for which it owns the patent rights. Dr. Bergmann holds ownership in and is a member of the board of directors of BRAHMS AG.
- Abbreviations and Acronyms
- acute coronary syndrome
- acute myocardial infarction
- B-type natriuretic peptide
- estimated glomerular filtration rate
- heart failure
- hazard ratio
- major adverse cardiac event(s)
- myocardial infarction
- net reclassification improvement
- non–ST-segment elevation myocardial infarction
- N-terminal pro–B-type natriuretic peptide
- receiver-operator characteristic
- mid-regional pro-adrenomedullin abbreviated to surrogate for adrenomedullin
- ST-segment elevation myocardial infarction
- Received August 24, 2009.
- Revision received January 7, 2010.
- Accepted January 18, 2010.
- American College of Cardiology Foundation
- Hasdai D.,
- Behar S.,
- Wallentin L.,
- et al.
- Terkelsen C.J.,
- Lassen J.F.,
- Norgaard B.L.,
- et al.
- Anderson J.L.,
- Adams C.D.,
- Antman E.M.,
- et al.
- Omland T.,
- Persson A.,
- Ng L.,
- et al.
- Sabatine M.S.,
- Morrow D.A.,
- de Lemos J.A.,
- et al.
- Miyao Y.,
- Nishikimi T.,
- Goto Y.,
- et al.
- Vari R.C.,
- Adkins S.D.,
- Samson W.K.
- Pousset F.,
- Masson F.,
- Chavirovskaia O.,
- et al.
- Khan S.Q.,
- O'Brien R.J.,
- Struck J.,
- et al.
- Thygesen K.,
- Alpert J.S.,
- White H.D.
- Smilde T.D.,
- van Veldhuisen D.J.,
- Navis G.,
- Voors A.A.,
- Hillege H.L.
- Morgenthaler N.G.,
- Struck J.,
- Alonso C.,
- Bergmann A.
- Niu P.,
- Shindo T.,
- Iwata H.,
- et al.
- Nagaya N.,
- Satoh T.,
- Nishikimi T.,
- et al.
- Kato J.,
- Tsuruda T.,
- Kita T.,
- Kitamura K.,
- Eto T.