a b s t r a c t

Objective: The main objective of this study was to determine a predictive cutoff value for plasma N-terminal pro-B-type natriuretic peptide (NT-proBNP) that could successfully predict the long-term (4-year) survival of patients with acute heart failure at the time of admission to the emergency department (ED). To our best knowledge, our study is the first research done to identify a predictive cutoff value for admission NT-proBNP to the prescriptive 4-year survival of patients admitted to ED with Acute HF diagnosis.

Methods: Nt-proBNP levels were measured in plasma obtained from 99 patients with dyspnea and left ventricular dysfunction upon admission to the ED. The end point was survival from the time of inclusion through 4 years.

Results: The mean age of the patients in this study was 71.1 +- 10.3 years; 50 of these patients were female. During the 4-year follow-up period, 76 patients died; survivors were significantly younger than non-survivors (64.26 +- 11.42 years vs72.83 +- 11.07 years, P = .002). The optimal NT-proBNP cutoff point for predicting 4- year survival at the time of admission was 2300 pg/mL, which had 85.9% sensitivity and 39.1% specificity (95% confidence interval, area under the curve: 0.639, P = .044).

Conclusion: Elevated NT-proBNP levels at the time of admission are a strong and independent predictor of all- cause mortality in patients with acute HF 4 years after admission. Furthermore, the optimal cutoff level of NT- proBNP used to predict 4-year survival had high sensitivity. However, especially in the case of long-term survival, additional prospective, large, and multicenter studies are required to confirm our results.

(C) 2013

Introduction

The prevalence and incidence of congestive heart failure are increasing worldwide. Due to its high mortality, early diagnosis and treatment of HF, as well as the identification of high-risk groups are very important [1,2]. Natriuretic peptides (B-type natriuretic peptide [BNP] and N-terminal pro-BNP [NT-proBNP]) are produced in both ventricles of the heart (predominantly in the left ventricle) and released into the circulation in response to increased cardiac wall stretching, volume overload or pressure overload. They have been shown to be useful in diagnosing and identifying the severity of HF in dyspneic patients upon

? The authors declare that they have no commercial associations or sources of support that might pose a conflict of interest.

* Corresponding author. Bitli State Hospital, Besminare Str. No:165, Bitlis, Turkey. Tel.: +90 4342468520; fax: +90 4342468425.

E-mail address: [email protected] (Y. Velibey).

admission to the emergency department (ED). Therefore, serial evaluations of the natriuretic peptides may be useful in determining the effectiveness of treatment of HF patients [3-15]. The objective of this study was to determine a cutoff value of plasma NT-proBNP that could successfully predict the long-term (4-year) survival of patients with acute HF at the time of admission to the ED. To the best of our knowledge, our work is the first research done to determine a cutoff value for NT-proBNP at time of admission to the ED that can predict 4- year survival of patients admitted to ED with a diagnosis of acute HF.

Materials and methods

Study design and patient selection

This prospective, observational, cross-sectional, single-center study was conducted in the academic ED of a tertiary care hospital.

0735-6757/$ – see front matter (C) 2013 http://dx.doi.org/10.1016/j.ajem.2013.08.033

Y. Velibey et al. / American Journal of Emergency Medicine 31 (2013) 1634–1637 1635

Upon approval from the local ethics committees, this study was carried out according to the principles of the Declaration of Helsinki, and written informed consent was obtained from all patients. Patients 21 years or older who were admitted to the ED with the primary complaint of shortness of breath were included in the study. Patients with significant heart valve disease, acute coronary syndrome (ST elevation or non-ST-elevation myocardial infarction and unstable angina pectoris), life-threatening cardiac arrhythmias, end-stage renal disease requiring hemodialysis or continuous ambulant perito- neal dialysis, post-traumatic shortness of breath, chronic obstructive lung disease, or chronic diseases such as cancer (which are known to increase NT-proBNP levels) at the time of admission were excluded. For patients who presented to the ED more than once, only data from their First visit was included in the study; 350 successive patients who came to the ED for acute dyspnea between August 2007 and September 2008 were included in the study. Blood samples were collected for the measurement of NT-proBNP. Of these 350 patients, a total of 251 (71.7%) were excluded (233 patients [66.5%] due to NT- proBNP levels below the age-based cut-off level, 17 [4.8%] due to diagnoses other than HF and failed to communicate with 1 [0.2%] patient diagnosed with HF). As a result, the complete follow-up was available for 99 patients, and all of the patients were followed up to May 2013.

NT-proBNP assay

At the time of enrollment, a 20-mL peripheral blood sample was collected into separate tubes for the measurement of NT-proBNP, cardiac markers, Biochemical parameters, and complete blood count using direct venipuncture of the antecubital vein. Samples were transported to our laboratory immediately. Plasma was separated from blood cells by centrifugation at 5000g for 10 minutes. Troponin I was measured using an Olympus AU640 auto analyzer (Olympus, Kobe, Japan) using standard methods. Plasma NT-proBNP (expressed as pg/mL) was determined using a sandwich enzyme immuno-assay method (Dade Dimension RxL autoanalyzer; Dade Behring, Deerfield, IL) with b 0.001% cross-reactivity with bioactive BNP. In the constituent studies mentioned in this report, this assay had an inter-run coefficient of variation ranging between 0.9% and 5.5%.

Echocardiographic measurements

A standard transthoracic Echocardiographic examination was performed on each patient by a single experienced consultant cardiologist who was blinded to the results of the NT-proBNP assays. This examination was performed with the available ultrasound equipment (M3S Probe, Vivid 7, GE Medical System, Horten, Norway) within the first 24 hours of ED or coronary care unit admission. Patients were defined as having LV diastolic dysfunction if they had normal EF (EFN 50%) with an increased E/e? ratio (N 15), decreased Tissue Doppler imaging-derived early diastolic myocardial velocities (b 8 cm/s septal, b 10 cm/s lateral, or b 9 cm/s average), an increased A pulmonary-A mitral duration (N 30 ms), a restrictive LV filling pattern (E/A N 2), or any combination of these parameters. Left ventricular EF (LVEF) was calculated by Simpson’s rule from Apical 4-chamber measurements, and patients were defined as having LV systolic dysfunction if their LVEF was b 50%.

Study protocol and follow-up evaluation

Patients with acute HF were defined as having elevated NT-proBNP levels if their levels exceeded age-specific cutoffs (21-50 years old: N 450 pg/mL, 50-75 years old: N 900 pg/mL, N 75 years-old: N 1800 pg/ mL), and serum Creatinine levels were 2.5 mg/dL or less. After enrollment into the study, venous blood samples for the NT-proBNP assay were obtained from eligible patients at presentation to the ED.

Patients were excluded if their NT-proBNP levels were below the age- based cutoff value or if their final diagnosis was not HF. All of the patients included in this study were followed up to May 2013 and were treated according to the American and European Society of Cardiology guidelines for the diagnosis and treatment of acute and chronic HF. The end point was survival from the time of inclusion through 4 years. survival status and dates of death were obtained from patients or their relatives by telephone interviews on May 2013.

Statistical analysis

Based on their survival during the follow-up, patients were divided into 2 groups (survival and non-survival). A receiver-operating characteristic (ROC) curve analysis was performed to identify the optimal cutoff value of NT-proBNP for predicting 4-year survival. The area under the ROC curve was calculated and used to evaluate diagnostic accuracy. The cumulative survival rate was calculated by the Kaplan-Meier method and differences in survival between the groups were compared using the Mantel-Cox log rank test. Efficiency level was studied by ROC curve and the result was checked by Kappa inter-rater correlation test. For all tests, P b .05 (2-sided) was considered to be statistically significant. All analyses were performed using SPSS version 21.0 statistical software (SPSS Inc, Chicago, IL).

Results

The mean age of the 99 patients included in this study was 71.1 +-

10.3 years, and 50 of these patients were women. The median level of NT-proBNP for all 99 patients was 6293.7 (IQR, 922.2-68978.4) pg/ mL. At the end of 4-year follow-up, 23 patients were still alive and we observed that survivors were significantly younger than non- survivors (64.26 +- 11.42 vs 72.83 +- 11.07 years, P = .002). During the review of 4-year survival cases, the admission NT-proBNP level was significantly able to predict survival (area under the curve [AUC]: 0.728, P b .005). The optimal NT-proBNP cutoff value for predicting 4- year survival was determined by ROC analysis to be 2300 pg/mL which had 85.9% sensitivity and 39.1% specificity (95% confidence interval [CI], AUC: 0.639, P = .044) (Fig. 1). There was a significant compatibility between the identified cutoff value and the possibility of

Fig. 1. Receiver-operating characteristic curves of NT-proBNP levels as a predictor of 4- year mortality.

1636 Y. Velibey et al. / American Journal of Emergency Medicine 31 (2013) 1634–1637

Table 1

survival time (month) of patients according to the determined cutoff values for predicting 30-day and 4-year survival

NT-proBNP	Mean				P
	Estimate	SE	95 % CI
			LB	UB
N 9200 pg/mL	11.61	3.08	5.58	17.64	.000
2300-9200 pg/mL	23.85	3.73	16.55	31.16	.000
b2300 pg/mL	38.63	5.42	28.00	49.25	.000
Overall	22.50	2.48	17.64	27.36	.000

Fig. 2. Kaplan-Meier survival curves for 4 years by NT-proBNP levels above and below the optimal cutoff value of 2300 pg/mL.

death during the 4-year period (? = 0.258, P = .010). Fig. 2 shows Kaplan-Meier survival curves for NT-proBNP levels according to this cutoff value. Patients with NT-proBNP levels above the cutoff value had significantly higher mortality than did those with NT-proBNP levels below the cutoff at 4-year (log-rank test, P = .001).

In previous study of the same group of patients, 79% lived during the first 30 days of follow-up. The optimal cutoff point for predicting 30-day survival was identified as 9152.4 pg/mL, which had 71.4% sensitivity and 81.3% specificity (95% CI, AUC: 0.726; P = .002). Fig. 3 shows Kaplan-Meier survival curves for NT-proBNP levels according cutoff values of 2300 and 9200 pg/mL. The predicted survival time for patients whose plasma NT-proBNP level was under 2300 pg/mL was

38.63 (28.00-49.25) months; between 2300 and 9200 pg/mL, 23.85

(16.55-31.16) months; and greater than 9200 pg/mL, 11.61 (5.58-

17.64) months–a significant increase (P b .005) (Table 1).

Fig. 3. Kaplan-Meier survival curves by NT-proBNP levels above and below the Optimal cutoff values of 2300 and 9200 pg/mL.

LB, lower board; UB, upper board.

The average survival time of the patients who died during 4-year follow-up was 132 days (range, 1-1400 days). Almost all the mortalities are a result of either arrhytmic complications or pump failure. Admission NT-proBNP levels were significantly lower in survivors compared to 4-year nonsurvivors (median 5051.50 [1166- 34383] pg/mL vs 6519 [IQR 1573-54876] pg/mL, P = .044, respec- tively). During follow-up, the independent predictors of mortality were determined to be NT-proBNP (N 2300 pg/mL) and admission oxygen saturation (P = .044, and P b .007, respectively (Table 2). There was no statistical difference between LVEF values of deceased and living patients (median 45.55 +- 10.33% vs 41.26 +- 11.61%, P = 0.249). Other traditional serum markers that are used to predict mortality in patients with HF (sodium, potassium, calcium, glucose, creatine kinase, creatine kinase-MB fraction, white blood cell, and hemoglo- bin) did not significantly associate with mortality in our population. In addition, New York Heart Association functional capacity, heart rate, and systolic and diastolic blood pressure did not predict mortality.

Discussion

The importance of identifying plasma NT-proBNP and BNP levels for the prognosis of acute HF at the time of admission was recently well described [4,5,14-20]. Although there are several studies reporting admission BNP cutoff values which can predict long-term survival, these studies typically report only the 1-Year survival time predictive cutoff values. The most familiar studies among these is the B-type Natriuretic Peptide for Acute Shortness of Breath Evaluation (BASEL) study. In the BASEL study, Boldanova et al found that the optimal BNP cutoff point for predicting 1-year survival on admission was 1275 pg/mL, with a sensitivity of 39% and a specificity of 72% [17]. Throughout our research, we have not seen studies related to longer predictive cutoff values.

However, despite the sufficient number of studies using admission BNP as a prognostic cutoff level in the literature, there are also a few studies that report admission NT-proBNP cutoff values that are predictive of long-term mortality. Our study demonstrates that NT- proBNP levels at the time of admission are predictive of long-term survival in patients complaining of dyspnea who were subsequently diagnosed with acute HF. In patients with shortness of breath, the age- adjusted values of NT-proBNP demonstrate clinical benefits for more than one outcome and risk classification in long-term disease. Analysis of the area under the ROC curve of the current data suggests that NT-proBNP cutoff values of 2300 pg/mL provide prognostic information regarding 4-year survival in patients with acute HF. There are several previous studies indicating cutoff values for plasma NT- proBNP at the time of admission for the prediction of long-term

Table 2

Multivariate logistic regression analysis for the prediction of death during follow-up

Variables	OR	95% CI	P
NT-proBNP (2300 pg/mL)	5.129	1.623-16.202	.005
Oxygen (O2) saturation	0.779	0.626-0.969	.025

Y. Velibey et al. / American Journal of Emergency Medicine 31 (2013) 1634–1637 1637

(especially 1-year) survival; there is no clinical consensus on a single optimal cut-off value. Gegenhuber et al examined 252 patients who were diagnosed with acute HF by emergency physicians; the optimal admission NT-proBNP cutoff point for predicting 1-year survival in this study was 2060 pg/mL [16]. In another study we determined that 3700 pg/ml is the cutoff value which had 83.0% sensitivity and 52.2% specificity (95% CI, AUC: 0.644; P = .014), predicting 1-year survival time as a result of the former study conducted on the same participants by Golcuk and colleagues. Patients who were alive after one year had lower median NT-proBNP levels (median: 4977.4 [922.2-47066.8] pg/mL [P = .014]) at the time of admission relative to those of the non-surviving patient group (median: 8331.8 [IQR, 988.6- 69868.3] pg/mL) [21]. Cumulatively, these studies suggest that patients presenting to the ED with dyspnea and who are diagnosed with acute HF have an increased risk of death, suggesting that aggressive treatment is appropriate for this patient population.

We also aimed to compare our observed results with the results of previous studies on this subject, but to the best of our knowledge, our study is the first research done looking at determining a predictive cutoff value of admission NT-proBNP to the prescriptive 4-year survival of patients admitted to ED with a diagnosis of acute HF.

Study limitations

The results of our study provide additional insight into the role of NT-proBNP measurements in the diagnostic assessment and subse- quent management of patients with shortness of breath in the ED setting. However, there are some potential limitations in this study. First, our cross-sectional study has a relatively small sample size and the study population consisted of patients at only one center (thus, the outcomes may not be applicable to other centers). Therefore, future prospective studies with larger sample sizes will be needed to validate our results. Second, the levels of NT-proBNP prior to admission to the ED and after discharge from the ED or hospital remain unknown, and our analyses were based on a single baseline determination. Hence, we could not address the potential importance of the change in NT-proBNP levels over time. Third, the effects of unmeasured confounding variables including body mass index, duration of HF, or complex interactions between covariates, such as effects of socioeconomic factors, race/ethnicity, cause of HF, and smoking status cannot be excluded.

Conclusion

In this study, we found that NT-proBNP levels upon admission are a strong and independent predictor of survival in patients with HF at 4-year follow-up. Furthermore, NT-proBNP levels were found to be a useful and incremental tool for clinical risk stratification of patients with acute HF and provided prognostic information beyond that provided by traditional Cardiovascular risk factors. We also demon- strate the cutoff value of admission NT-proBNP levels for the long- term estimation of survival in acute HF; these cutoff value have high sensitivity. However, especially in the case of long-term survival, additional multi-center, large and prospective studies are required in order to confirm our results.

Acknowledgments

The authors would like to thank Mr. Ertan Koc (biostatistics expert) for his great assistance in data management and statistical analysis.

References

1. Redfield MM. Heart failure-an epidemic of uncertain proportions. N Engl J Med 2002;347:1442-4.
2. Peacock WF, Nowak R, Christenson R, et al. Short-term mortality risk in emergency department acute heart failure. Acad Emerg Med 2011;18:947-58.
3. Levin ER, Gardner DG, Samson WK. Natriuretic peptides. N Engl J Med 1998;339: 321-8.
4. Trinquart L, Ray P, Riou B, et al. Natriuretic peptide testing in EDs for managing acute dyspnea: a meta-analysis. Am J Emerg Med 2011;29:757-67.
5. Chen HH, Burnett JCJ. The natriuretic peptides in heart failure: diagnostic and therapeutic potentials. Proc Assoc Am Physicians 1999;111:406-16.
6. Carpenter CR, Keim SM, Worster A, Rosen P, BEEM (Best Evidence in Emergency Medicine). Brain natriuretic peptide in the evaluation of emergency department dyspnea: is there a role? J Emerg Med 2012;42:197-205.
7. McDonagh TA, Holmer S, Raymond I, et al. NT-proBNP and the diagnosis of heart failure: a Pooled analysis of three European epidemiological studies. Eur J Heart Fail 2004;6:269-73.
8. Chenevier-Gobeaux C, Claessens YE, Voyer S, Desmoulins D, Ekindjian OG. Influence of renal function on N-terminal pro-brain natriuretic peptide (NT- proBNP) in patients admitted for dyspnoea in the Emergency Department: comparison with Brain natriuretic peptide . Clin Chim Acta 2005;361: 167-75.
9. Januzzi JL, van Kimmenade R, Lainchbury J, et al. NT-proBNP testing for diagnosis and short-term prognosis in acute destabilized heart failure: an international pooled analysis of 1256 patients: the International Collaborative of NT-proBNP Study. Eur Heart J 2006;27:330-7.
10. Hammerer-Lercher A, Neubauer E, Muller S, et al. Head-to-head comparison of terminal pro-brain natriuretic peptide, brain natriuretic peptide and N-terminal pro-atrial natriuretic peptide in diagnosing left ventricular dysfunction. Clin Chim Acta 2001;310:193-7.
11. Januzzi Jr JL, Camargo CA, Anwaruddin S, et al. The N-terminal Pro-BNP investigation of dyspnea in the emergency department (PRIDE) study. Am J Cardiol 2005;95:948-54.
12. Januzzi JL, Bayes-Genis A. Evolution of amino-terminal pro-B type natriuretic peptide testing in heart failure. Drug News Perspect 2009;22:267-73.
13. Maisel AS, Koon J, Krishnaswamy P, et al. Utility of B-natriuretic peptide as a rapid, Point-of-care test for Screening patients undergoing echocardiography to determine left ventricular dysfunction. Am Heart J 2001;141:367-74.
14. Masson S, Vago T, Baldi G, et al. Comparative measurement of N-terminal pro- brain natriuretic peptide and brain natriuretic peptide in ambulatory patients with heart failure Clin Chem Lab Med 2002;40:761-3.
15. Bayes-Genis A, Santalo-Bel M, Zapico-Muniz E, et al. N-terminal probrain natriuretic peptide (NT-proBNP) in the emergency diagnosis and in-hospital monitoring of patients with dyspnoea and ventricular dysfunction. Eur J Heart Fail 2004;6:301-8.
16. Gegenhuber A, Mueller T, Dieplinger B, et al. B-type natriuretic peptide and amino terminal proBNP predict one-year mortality in short of breath patients independently of the baseline diagnosis of acute destabilized heart failure. Clin Chim Acta 2006;370:174-9.
17. Boldanova T, Noveanu M, Breidthardt T, Potocki M, et al. Impact of history of heart failure on Diagnostic and prognostic value of BNP: Results from the B-type Natriuretic Peptide for Acute Shortness of Breath Evaluation (BASEL) Study. Int J Cardiol 2010;142:265-72.
18. Noveanu M, Breidthardt T, Potocki M, Reichlin T, et al. Direct comparison of serial B-type natriuretic peptide and NT-proBNP levels for prediction of short- and long- term outcome in acute decompensated heart failure. Crit Care 2011;15(1):R1.
19. Januzzi Jr JL, Chen-Tournoux AA, Moe G. Amino-terminal pro-B-type natriuretic peptide testing for the diagnosis or exclusion of heart failure in patients with acute symptoms. Am J Cardiol 2008;101:29-38.
20. Rutten JH, Steyerberg EW, Boomsma F, van Saase JL, et al. N-terminal pro-brain natriuretic peptide testing in the emergency department: beneficial effects on hospitalization, costs, and outcome. Am Heart J 2008;156:71-7.
21. Golcuk Y, Golcuk B, Velibey Y, Oray D, et al. Predictive cutoff point of admission N- terminal pro-B-type natriuretic peptide testing in the ED for prognosis of patients with acute heart failure. Am J Emerg Med 2013;31:1191-5.