Incremental Value of N-terminal Pro–Brain Natriuretic Peptide Over Left Ventricle Ejection Fraction and Aerobic Capacity for Estimating Prognosis in Heart Failure Patients

FAILING HEART—MEDICAL ASPECTS

Incremental Value of N-terminal Pro–Brain Natriuretic Peptide Over Left Ventricle Ejection Fraction and Aerobic Capacity for Estimating Prognosis in Heart Failure Patients Manolis S. Kallistratos, MD, Athanasios Dritsas, MD, FESC, Ioannis D. Laoutaris, PhD, FESC, and Dennis V. Cokkinos, MD, FESC, FACC Background: N-terminal pro– brain natriuretic peptide (NT-proBNP) plasma levels have been associated with indices of left ventricular (LV) function and aerobic capacity in heart failure. Methods: We prospectively followed-up 149 patients with impaired left ventricular function for 30 ⫾ 10 months. During this period, 22 patients died and 5 underwent heart transplantation. Blood samples for NT-proBNP assessment were taken at baseline and before cardiopulmonary exercise to estimate peak oxygen consumption (VO2). LV cavity diameter, left atrial size and LV ejection fraction (LVEF) were measured by echocardiography. Results: NT-proBNP plasma levels ⬎1,164 pg/ml showed 85% sensitivity and 82% specificity for detecting VO2⬍14 ml/kg/min (area under the curve [AUC] ⫽ 90%, p ⬍ 0.001). Patients above this cutoff showed a 13.6-fold greater hazard ratio compared with those with values below this cutoff (p ⬍ 0.001). NT-proBNP plasma levels of ⬎760 pg/ml showed 77% sensitivity and 69% specificity for detecting LVEF ⬍28% (AUC ⫽ 77%, p ⬍ 0.001). Patients with values above this cutoff showed a 15.85-fold greater hazard ratio compared to those with values below this cutoff (p ⬍ 0.001). The addition of NT-proBNP to an assessment model that includes peak VO2, LVEF and New York Heart Association (NYHA) classification can significantly improve predictive ability. Conclusions: Assessment of NT-proBNP should be performed to detect candidates for heart transplantation because of the useful prognostic information that it can provide. J Heart Lung Transplant 2008;27: 1251– 6. Copyright © 2008 by the International Society for Heart and Lung Transplantation.

The selection of patients for heart transplantation (HT) is notoriously difficult and traditionally involves clinical assessment. Widely accepted markers of the severity of chronic heart failure (CHF) include left ventricular ejection fraction (LVEF) and peak VO2. In the V-HeFT-I study, which enrolled patients with mild–moderate heart failure, LVEF was dichotomized at a mean value of 28%. Patients with LVEF below this value had an annual mortality of 22%, compared with 13% for patients with LVEF above the mean.1 However, echocardiographic determination of LVEF is obtained only in 70% to 85% of patients2,3 in everyday practice. Several studies have suggested that, in patients with heart failure, maximum oxygen consumption (VO2max) is a good short-term predictor of mortality and that its deterioration freFrom the First Cardiology Department, Onassis Cardiac Surgery Center, Athens, Greece. Submitted December 14, 2007; revised February 23, 2008; accepted July 20, 2008. Reprint requests: Manolis Kallistratos, MD, First Cardiology Department, Onassis Cardiac Surgery Center, 356 Sygrou Avenue, 17674 Athens, Greece. Telephone: ⫹30-210-94-93-000. Fax: ⫹30-210-94-93336. E-mail: [email protected] Copyright © 2008 by the International Society for Heart and Lung Transplantation. 1053-2498/08/$–see front matter. doi:10.1016/ j.healun.2008.07.030

quently precedes clinical decompensation.4 – 6 Patients with a peak VO2 ⱕ14 ml/kg/min benefit in terms of prognosis from transplantation.7–9 However, patients with severe CHF often cannot achieve a true peak VO2due to leg fatigue, angina or general debilitation. Brain natriuretic peptide (BNP) has been employed as a marker of heart failure over the last 10 years. It increases in proportion to the severity of chronic heart failure,10 can independently predict morbidity and mortality in asymptomatic LV systolic dysfunction11 and in mild–moderate CHF,12–14 and can predict low-functional-class patients and candidates for heart transplantation (HT).15 In addition, BNP has been shown to be a strong, independent predictor of sudden death in patients with CHF.16 The aim of our study was to evaluate the incremental prognostic value of N-terminal pro– brain natriuretic peptide (NT-proBNP) to differentiate patients with LVEF ⬍28% and peak VO2 ⬍14 ml/kg/min,17,18 as well as the occurrence of death or HT at follow-up, and its correlation with functional capacity as assessed by the New York Heart Association (NYHA) class. METHODS All patients included in this study were recruited from a single center (Onassis Cardiac Surgery Center, Athens, Greece) during the period from September 2003 to 1251

1252

Kallistratos et al.

March 2007. We prospectively studied 160 patients (130 men and 30 women) diagnosed with LV dysfunction due to coronary artery disease, dilated cardiomyopathy and valvular heart disease. Patients were either post-operative cases or being evaluated for cardiac surgery during entry into our heart failure program. For the patients with dilated cardiomyopathy and valvular heart disease, coronary artery disease was ruled-out on the basis of coronary arteriography. Patients were on medical therapy dictated by their physicians. The mean (⫾ SD) age of patients was 59 ⫾ 12 years. All had impaired LV systolic function characterized by LVEF ⬍50% (measured by 2-dimensional echocardiography). They were classified according to NYHA functional class. The main exclusion criteria were age ⬍18 or ⬎80 years, treadmill exercise duration ⬍2 minutes, systolic blood pressure ⬍90 mm Hg or ⬎160/100 mm Hg, primary pulmonary disease and peripheral vascular or degenerative joint disease that could restrict ability to exercise. The primary end-point at follow-up was cardiac death or heart transplantation. Our hospital’s institutional committee approved the study and written informed consent was obtained from all participants. Cardiopulmonary Exercise Testing During treadmill exercise using the Dargie protocol,19 peak VO2, expired volume (VE)/VO2 slope and VE/VCO2 slope were assessed using a standard gas-analysis technique with Med Graphics CPX/MAX (Medical Graphics Corp., St. Paul, MN). A 12-lead electrocardiogram (ECG) was continuously utilized to exclude significant myocardial ischemia or arrhythmias. Blood pressure was recorded every minute by a cuff sphygmomanometer. Peak VO2was obtained as the highest value in the terminal phase of exercise. Laboratory Analysis After patients had been seated for 30 minutes, baseline blood samples were obtained from an antecubital vein and collected into ethylene-diamine tetraacetic acid (EDTA)-containing tubes. The samples were then centrifuged and plasma was stored in aliquots at ⫺20°C within 30 minutes. Plasma NT-proBNP was determined by sandwich immunoassay (Elecsys 1010; Roche Diagnostics). The analytical range extends from 20 to 35,000 pg/ml. This automated system has been shown to have the smallest coefficient of analytic variation, just below 2%.20 Statistical Analysis Data were analyzed using the STATA statistical program. Results were expressed as mean values (SD) or as median values (interquartile range). Qualitative variables are expressed as absolute and relative frequen-

The Journal of Heart and Lung Transplantation November 2008

cies. For comparability of uni- and multivariate analyses, results apply to the data set without missing values. Correlations between NT-proBNP and the other variables investigated were assessed using Spearman’s nonparametric correlation coefficient. Peak VO2 ⬍10, peak VO2 ⬍14 and LVEF ⬍28% were estimated via NT-proBNP using receiver operating characteristic (ROC) curves. Sensitivity and specificity were calculated for optimal cutoffs. The area under the curve (AUC) was also calculated. Cox proportional hazards models were used to determine variables associated with overall survival. Because of the high correlation between the main predictors (i.e., NT-proBNP, peak VO2, LVEF and NYHA class) in the multiple regression models, adjustment was made only for gender and age because model diagnostics with two or more correlated predictors in the models indicate that the regression estimates were highly collinear. The assumption of proportional hazards was evaluated by testing for interaction with a continuous time variable. The accuracy of the different models as predictors of survival was evaluated by concordance (c)-statistics (equivalent to the area under the ROC curve). Harrell’s c-statistic can range from 0.5 (prediction no better than chance) to 1.0 (perfect prediction). Each model was considered to have diagnostic accuracy when the c-statistic was ⬎0.70 and excellent diagnostic accuracy in when cstatistic was ⬎0.80. The actual survival curve for each group was calculated using the Kaplan–Meier method. All reported p-values are 2-tailed. Statistical significance was set at p ⬍ 0.05 and analyses were conducted using STATA statistical software (version 6.0). RESULTS From a total of 160 patients, 11 dropped out due to poor attendance. Thus, a total of 149 subjects, 122 men and 27 women, were included in the analysis. During the follow-up period (median 30 months, range 3 to 50), 5 subjects underwent heart transplantation and 22 died. Patients’ demographic and clinical characteristics are shown in Table 1. Cardiopulmonary Exercise Testing Mean peak VO2 achieved was 17.6 ⫾ 5 ml/kg/min. NTproBNP levels correlated significantly with peak VO2 (r ⫽ ⫺0.71, p ⬍ 0.001). Resting NT-proBNP plasma levels ⬎1,164 pg/ml showed 85% sensitivity and 82% specificity for detecting peak VO2 ⬍14 ml/kg/min (AUC ⫽ 90%, standard error [SE] ⫽ 0.03, p ⬍ 0.001) (Figure 1). NT-proBNP plasma levels ⬎1,460 pg/ml showed 91% sensitivity and 81% specificity for detecting peak VO2 ⬍10 ml/kg/min (AUC ⫽ 88%, SE ⫽ 0.03, p ⬍ 0.001) (Figure 2).

The Journal of Heart and Lung Transplantation Volume 27, Number 11

Kallistratos et al.

0.75 0.25

20 (13%) 70 (47%) 59 (40%)

Sensitivity 0.50

149 59 ⫾ 13 46 (30.9%) 59 (39.6%) 44 (29.5%)

131 (88%) 128 (85%) 91 (61%) 108 (72%) 33.6 ⫾ 9.3 64 ⫾ 12 52 ⫾ 14 47 ⫾ 8 17.6 ⫾ 5 38 ⫾ 11 35 ⫾ 8 604 (1,072 ⫾ 1,302) 22 (14.8%) 5 (3.4%)

LVEF, left ventricular ejection function; LVEDD, left ventricle end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LA, left atrial diameter.

Echocardiography Mean LVEF was 33.6 ⫾ 9.3%. LVEF correlated significantly with NT-proBNP (r ⫽ ⫺0.59, p ⬍ 0.001). NTproBNP plasma levels ⬎760 pg/ml showed 77% sensitivity and 69% specificity for detecting LVEF ⬍28% (AUC ⫽ 77%, SE ⫽ 0.04, p ⬍ 0.001) (Figure 3).

0.00

0.25

0.50 1 - Specificity

0.75

1.00

Area under ROC curve = 0.8847

Figure 2. ROC curve of NT-proBNP for the prediction of PVO2 ⬍10 ml/kg/min.

plasma levels of ⬎1,164 pg/ml (cutoff that predict peak VO2 ⬍14 ml/kg/min) had a 13.6-fold greater hazard ratio for death or heart transplantation (HT) and a median survival range of 25 months (interquartile range 17 to 33 months) compared to those with values below this cutoff (p ⬍ 0.001) (median survival range 33 months, interquartile range 25 to 39 months). Patients with NT-proBNP plasma levels of ⬎1,460 pg/ml (cutoff that predict peak VO2 ⬍10 ml/kg/min) showed a 7.6-fold greater hazard ratio for death or HT and a median survival range of 25.5 months (interquartile range 16.5 to 33 months) compared to those with values below this cutoff (p ⬍ 0.001) (median survival range 31 months, interquartile range 17 to 33 months). Patients with NT-proBNP plasma levels of ⬎760 pg/ml (cutoff that predict LVEF ⬍28%) showed a 15.8-fold greater hazard ratio for death or HT and a median survival range of 29 months (interquartile range 23 to 38 months) compared to those with values bellow this cutoff (p ⬍ 0.005) (median survival range 30 months, interquartile range 24.5 to 38 months).

0.00

0.25

0.50 1 - Specificity

0.75

1.00

Area under ROC curve = 0.8973

Figure 1. ROC curve of NT-proBNP for the prediction of PVO2 ⬍14 ml/kg/min.

0.00

0.00

0.25

0.25

Sensitivity 0.50

Sensitivity 0.50

0.75

0.75

1.00

1.00

NT-proBNP Plasma Levels NT-proBNP median interquartile range was 604 pg/ml (mean 1,072 ⫾ 1,302 pg/ml). Patients with NT-proBNP

0.00

Characteristic N Age, years (mean ⫾ SD) NYHA Class I NYHA Class II NYHA Class III HF etiology Valvular heart disease Coronary artery disease Dilated cardiomyopathy Medication Angiotensin-converting enzyme inhibitors Diuretics Digitalis Beta-blockers LVEF (%) LVEDD (mm) LVESD (mm) LA (mm) VO2 peak (ml/kg/min) VE/VO2 slope VE/VCO2 slope NT-proBNP median interquartile range (mean) Death, N (%) Heart transplantation, N (%)

1.00

Table 1. Patient Clinical and Demographic Characteristics

1253

0.00

0.25

0.50 1 - Specificity

0.75

1.00

Area under ROC curve = 0.7743

Figure 3. ROC curve of NT-proBNP for the prediction of LVEF ⬍28%.

1254

Kallistratos et al.

The Journal of Heart and Lung Transplantation November 2008

Table 2. Crude and Adjusted Hazard Ratios for Predictors of Mortality/Heart Transplantation HR (95% CI) crude 0.83 (0.76–0.9) 5.53 (2.32–13.19) 3.61 (1.69–7.7) 1.05 (1.03–1.07) 6.22 (2.88–13.45) 11.77 (4.45–31.18) 12.63 (3.8–42) 0.93 (0.89–0.97) 3.0 (1.41–6.42) 2.25 (1.31–3.87)

PVO2 PVO2 ⱕ10 ml/kg/min PVO2 ⱕ14 ml/kg/min NT-proBNP (100 units) NT-proBNP ⱖ1,460 pg/ml NT-proBNP ⱖ1,164 pg/ml NT-proBNP ⱖ760 pg/ml LVEF (%) LVEF ⱕ28% NYHA class

p ⬍0.001 ⬍0.001 0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.001 0.005 0.003

c-statistic 0.723 0.593 0.625 0.800 0.703 0.765 0.744 0.676 0.618 0.647

HR (95% CI) adjusteda 0.82 (0.76–0.9) 5.37 (2.25–12.83) 4.51 (2.06–9.85) 1.07 (1.04–1.09) 7.58 (3.45–16.66) 13.61 (5.07–36.55) 15.85 (4.63–54.24) 0.92 (0.88–0.97) 3.09 (1.44–6.66) 2.60 (1.5–4.49)

p ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.004 0.001

c-statistic 0.735 0.641 0.657 0.763 0.731 0.780 0.778 0.690 0.639 0.670

HR, hazard ratio; CI, confidence interval. a Adjusted for gender and age.

Crude and adjusted hazard ratios for the predictors of mortality/HT are shown in Table 2. Plasma NT-proBNP levels also associated strongly with NYHA (r ⫽ 0.64, p ⬍ 0.001). Survival The 3-, 6-, 12- and 24-month survival rates without death or HT were 99%, 97%, 87% and 74%, respectively. Unadjusted Kaplan–Meier survival estimates over the follow-up period for death or HT, according to the various cutoffs for NT-proBNP, are presented in Figure 4. Crude and adjusted hazard ratios for death or HT indicated that NT-proBNP, peak VO2, LVEF and NYHA class are significant predictors (Table 2). NT-proBNP had the better diagnostic accuracy, as indicated by the adjusted models, with the c-statistic ranging from 0.78 (using the cutoff of 1,164) to 0.778 (using the cutoff of 760). Peak VO2 diagnostic accuracy was shown to be satisfactory with the c-statistic ⫽ 0.735, yet lower when cutoffs of 10 and 14 were used. The diagnostic accuracy of LVEF and NYHA class improved after adjusting for A

Survival Probability 0.00 0.250.50 0.75 1.00

Survival Probability 0.00 0.25 0.50 0.75 1.00

10

20 30 Time(months)

40

Table 3. Improvement in Prediction Ability Provided by c-Statistic by Addition of NT-proBNP Plasma Levels

50

10

20 30 Time(months)

BNP<760

40

50

BNP≥760

D

Survival Probability 0.00 0.25 0.50 0.75 1.00

Survival Probability 0.00 0.250.50 0.75 1.00

C

0

DISCUSSION Peak VO2 derived from treadmill cardiopulmonary exercise testing is an objective estimator of severity of heart failure. In our study NT-proBNP correlated strongly with VO2 peak. Moreover, we found that NT-proBNP levels can be used to identify CHF patients traditionally considered candidates for HT (peak VO2 ⬍14 ml/kg/ min, peak VO2 ⬍10 ml/kg/min, LVEF ⬍28%17,18). Recent studies are consistent with our findings with regard to the ability of BNP to predict functional capacity and ventilatory response to exercise,21–23 but also for discriminating those patients with a high likelihood of being candidates for HT.15,24 In accordance with other studies, this investigation has confirmed the poor prognosis associated not only with advanced heart failure but also with high NT-

B

0 0

gender and age but there was no practical difference in the discrimination provided by the c-statistic. The addition of NT-proBNP into the model with peak VO2, LVEF, and NYHA, when adjusted for gender and age, significantly improved the predictive ability of the c-statistic (Table 3).

10

20 30 Time(months)

BNP<1164

40 BNP≥1164

50

0

10

20 30 Time(months)

BNP<1461

40

50

BNP≥1461

Figure 4. Overall Kaplan–Meier estimates (A), using 760 pg/ml as a cutoff for BNP (B), using 1,164 pg/ml as a cutoff for BNP (C), and using 1,461 pg/ml as a cutoff for BNP (D).

Model 1 PVO2 NT-proBNP (100 units) Model 2 LVEF NT-proBNP (100 units) Model 3 NYHA class NT-proBNP (100 units)

HR (95% CI) adjusteda

p

c-statistic

0.89 (0.81–0.98) 1.05 (1.03–1.08)

0.017 ⬍0.001

0.773

0.96 (0.91–1.01) 1.06 (1.03–1.08)

0.090 ⬍0.001

0.782

1.41 (0.75–2.64) 1.06 (1.03–1.08)

0.289 ⬍0.001

0.756

Model 1: adjusted for gender, age, PVO2, NT-proBNP; Model 2: adjusted for gender, age, LVEF, NT-proBNP; Model 3: adjusted for gender, age, NYHA, NT-proBNP.

The Journal of Heart and Lung Transplantation Volume 27, Number 11

proBNP plasma levels.25 The 1-year mortality rate of 13% is in keeping with mortality rates reported in a recent trial of ␤-adrenoreceptor antagonists in advanced heart failure (i.e., the COPERNICUS study),26 wherein annual mortality rates were 18.5% in the placebo group and 11.4% in the carvedilol-treated group. Cox proportional hazards regression analyses showed NT-proBNP to be a significant predictor of long-term prognosis. We found that NT-proBNP has the predictive ability to detect patients who satisfy the criteria as candidates for HT with various cutoff levels. The hazard ratios of elevated NT-proBNP concentrations of 760 pg/ ml, 1,164 pg/ml and 1,460 pg/ml (plasma levels that predict patients with LVEF ⬍28%, peak VO2 ⬍14 ml/kg/ min and peak VO2 ⬍10 ml/kg/min—indicative criteria for HT17,18) were calculated to be approximately 15.8, 13.6 and 7.6, respectively, in the multivariate model. In our study population we included some patients with mild symptoms due to a structural disorder of the heart and a relatively well-preserved EF, who would be considered Stage B patients according to the 2001 ACC/AHA guidelines for CHF; however, all patients were considered to be in need of CHF medication to enjoy a good quality of life. The 2001 ACC/AHA guidelines for CHF proposed a new classification of CHF that includes “asymptomatic CHF patients” as Stage A or B. However, patients classified as Stage B will progress to symptomatic heart failure in the future and preventive/ front-loaded measures may be required to improve their prognosis and quality of life with regard to symptoms, function, exercise tolerance and prognosis. In our study, the use of various cutoffs of NT-pro-BNP offers significant prognostic and predictive information. Thus, NT-proBNP added prognostic information not only to LVEF but also to peak VO2 and NYHA, thus increasing the total predictive value of the multivariate model. Study Limitations In the present study, blood samples were obtained prior to cardiopulmonary exercise testing before any changes in therapy were made. Adjustments to therapy may well influence both outcome and NT-proBNP concentration. It remains unclear whether the relative prognostic value of the natriuretic peptides would have differed from the present results if we had obtained them serially after therapy adjustment, as recently suggested by the STARS-BNP multicenter study.27 Furthermore, the large number of patients with what could be considered moderate heart failure has already been discussed. In conclusion, NT-proBNP plasma levels correlated with both LVEF and aerobic capacity can predict low functional cardiopulmonary exercise capacity in pa-

Kallistratos et al.

1255

tients with impaired LV function and can be useful in the detection of HT candidates. A single measurement of NT-proBNP in patients with CHF can help to identify patients at highest risk of death. REFERENCES 1. Cohn JN, Archibald DG, Francis GS, et al. Veterans Administration Cooperative Study on Vasodilator Therapy of Heart Failure: influence of prerandomization variables on the reduction of mortality by treatment with hydralazine and isosorbide dinitrate. Circulation 1987;75(suppl IV):IV-49 –54. 2. McDonagh TA, Morrison CE, Lawrence A, et al. Symptomatic and asymptomatic left-ventricular systolic dysfunction in an urban population. Lancet 1997;350:829 –33. 3. Nieminen MC, Brutsaert D, Dickstein K, et al. EuroHeart Failure Survey II (EHFS II): a survey on hospitalized acute heart failure patients: description of population. Eur Heart J 2006;27:2725–36. 4. Likoff MJ, Chandler SL, Kay HR. Clinical determinants of mortality in chronic congestive heart failure secondary to idiopathic dilated or to ischemic cardiomyopathy. Am J Cardiol 1987;59:634 – 8. 5. Szlachcic J, Massie BM, Kramer BL, Topic N, Tubau J. Correlates and prognostic implication of exercise capacity in chronic congestive heart failure. Am J Cardiol 1985;55:1037– 42. 6. Pilote L, Silberberg J, Lisbona R, Sniderman A. Prognosis in patients with low left ventricular ejection fraction after myocardial infarction: importance of exercise capacity. Circulation 1989; 80:1636 – 41. 7. Mancini DM, Eisen H, Kussmaul W, et al. Value of peak exercise consumption for optimal timing of cardiac transplantation in ambulatory patients with heart failure. Circulation 1991;83:778 – 86. 8. Mancini D, LeJemtel T, Aaronson K. Peak VO2: a simple yet enduring standard. Circulation 2000;101:1080 –2. 9. Pardaens K, Van CJ, Vanhaecke J, et al. Peak oxygen uptake better predicts outcome than submaximal respiratory data in heart transplant candidates. Circulation 2000;101:1152–7. 10. Tsutamoto T, Wada A, Maeda K, et al. Attenuation of compensation of endogenous cardiac natriuretic peptide system in chronic heart failure—prognostic role of brain natriuretic peptide concentration in patients with chronic symptomatic left ventricular dysfunction. Circulation 1997;96:509 –16. 11. Tsutamoto T, Wada A, Maeda K, et al. Plasma brain natriuretic peptide level as a biochemical marker of morbidity and mortality in patients with asymptomatic or minimally symptomatic left ventricular dysfunction— comparison with plasma angiotensin II and endothelin-1. Eur Heart J 1999;20:1799 – 807. 12. Hüllsmann M, Berger R, Sturm B, et al. Prediction of outcome by neurohumoral activation, the six-minute walk test and the Minnesota Living with Heart Failure Questionnaire in an outpatient cohort with congestive heart failure. Eur Heart J 2002;23:886 –91. 13. Richards AM, Doughty R, Nicholls MG, et al. Plasma N-terminal pro-brain natriuretic peptide and adrenomedullin: prognostic utility and prediction of benefit from carvedilol in chronic ischemic left ventricular dysfunction. J Am Coll Cardiol 2001;37: 1781–7. 14. Koglin J, Pehlivanli S, Schwaiblmair M, et al. Role of brain natriuretic peptide in risk stratification of patients with congestive heart failure. J Am Coll Cardiol 2001;38:1934 – 41. 15. Kallistratos MS, Dritsas A, Laoutaris ID, Cokkinos DV. N-terminal prohormone brain natriuretic peptide as a marker for detecting low functional class patients and candidates for cardiac transplantation: linear correlation with exercise tolerance. J Heart Lung Transplant 2007;26:516 –21.

1256

Kallistratos et al.

16. Berger R, Hülsmann M, Strecker K, et al. B-type natriuretic peptide predicts sudden death in patients with chronic heart failure. Circulation 2002;105:2392–7. 17. Hunt SA, Abraham WT, Chin MH, et al. ACC/AHA 2005 guideline update for the diagnosis and management of chronic heart failure in the adult: a report of the American College of Cardiology/ American Heart Association Task force on practice guidelines. Circulation 2005;112:e154 –235. 18. Costanzo MR, Augustine S, Bourge R, et al. Selection and treatment of candidates for heart transplantation. Circulation 1995; 92:3593– 612. 19. Riley M, Northridge DB, Henderson E, et al. The use of an exponential protocol for bicycle and treadmill exercise testing in patients with chronic cardiac failure. Eur Heart J 1992;13:1363–7. 20. Wu AHB, Smith A, Wieczorek S, et al. Biological variation for N-terminal Pro and B-type natriuretic peptides and implications for therapeutic monitoring of patients with congestive heart failure. Am J Cardiol 2003;92:628 –31. 21. Passino C, Polletti R, Bramanti F, et al. Neuro-hormonal activation predicts ventilatory response to exercise and functional capacity in patients with heart failure. Eur J Heart Fail 2006;8:46 –53.

The Journal of Heart and Lung Transplantation November 2008

22. Williams SG, Ng LL, O’Brien RJ, et al. Is plasma N-BNP a good indicator of the functional reserve of failing hearts? The FRESHBNP study. Eur J Heart Fail 2004;6:891–900. 23. Williams SG, Ng LL, O’Brien RJ, et al. Complementary roles of simple variables, NYHA and N-BNP, in indicating aerobic capacity and severity of heart failure. Int J Cardiol 2005;102:279 – 86. 24. Rothenburger M, Wichter T, Schmid C, et al. Aminoterminal pro type B natriuretic peptide as a predictive and prognostic marker in patients with chronic heart failure. J Heart Lung Transplant 2004;23:1189 –97. 25. Perna ER, Macin SM, Cimbaro Canella JP, et al. Importance of early combined N-terminal pro-brain natriuretic peptide and cardiac troponin T measurements for long-term risk stratification of patients with decompensated heart failure. J Heart Lung Transplant 2006;25:1230 – 40. 26. Packer M, Coats AJ, Fowler MB, et al. Effect of carvedilol on survival in severe chronic heart failure. N Engl J Med 2001;344: 1651– 8. 27. Jourdain P, Jondeau G, Funk F, et al. Plasma brain natriuretic peptide-guided therapy to improve outcome in heart failure: the STARS-BNP Multicenter Study. Am Coll Cardiol 2007;49:1733–9.