Journal of the American College of Cardiology Vol. 46, No. 5, by the American College of Cardiology Foundation ISSN /05/$30.

Transcription

1 Journal of the American College of Cardiology Vol. 46, No. 5, by the American College of Cardiology Foundation ISSN /05/$30.00 Published by Elsevier Inc. doi: /j.jacc Impact of Atrial Fibrillation on the Diagnostic Performance of B-Type Natriuretic Peptide Concentration in Dyspneic Patients An Analysis From the Breathing Not Properly Multinational Study Heart Failure Cathrine Wold Knudsen, MD,* Torbjørn Omland, MD, PHD, MPH, Paul Clopton, MSC, Arne Westheim, MD, PHD,* Alan H. B. Wu, PHD, Philippe Duc, MD, James McCord, MD, Richard M. Nowak, MD, MBA, Judd E. Hollander, MD,# Alan B. Storrow, MD,** William T. Abraham, MD, FACC, Peter A. McCullough, MD, Alan Maisel, MD, FACC Oslo, Norway; San Diego and San Francisco, California; Paris, France; Detroit and Royal Oak, Michigan; Philadelphia, Pennsylvania; Cincinnati, Ohio; and Lexington, Kentucky OBJECTIVES BACKGROUND METHODS RESULTS CONCLUSIONS This study was designed to assess the diagnostic performance of B-type natriuretic peptide (BNP) in the diagnosis of acute congestive heart failure (CHF) in patients with permanent/ paroxysmal atrial fibrillation (AF) presenting with acute dyspnea. It is unknown to what extent AF affects the diagnostic performance of BNP in patients presenting with acute dyspnea. We studied 1,431 patients drawn from a cohort of patients (n 1,586) with acute dyspnea who had BNP levels measured on arrival. Patients were prospectively classified according to the presence or absence of permanent/paroxysmal AF. In total, 292 patients had permanent/paroxysmal AF. In patients without HF, permanent/ paroxysmal AF was associated with significantly higher BNP levels (p 0.001). Conversely, in patients with HF, BNP levels did not differ significantly between patients with and without AF (p 0.533). A BNP cutoff value of 100 pg/ml had a specificity of 40% and 79% for the diagnosis of acute HF in patients with and without AF, respectively. The areas under the receiver-operating characteristic curves were 0.84 (95% confidence interval 0.78 to 0.89) and 0.91 (95% confidence interval 0.89 to 0.93) for patients with and without AF, respectively. In patients without, but not in those with HF, the presence of AF is associated with higher circulating BNP levels, suggesting that a higher diagnostic threshold should be used in patients with AF. (J Am Coll Cardiol 2005;46:838 44) 2005 by the American College of Cardiology Foundation B-type natriuretic peptide (BNP) is a peptide hormone derived from atrial and ventricular cardiomyocytes (1 4). Circulating levels of BNP are increased in conditions characterized by volume overload, including cardiac (3 8) and renal failure (9,10). The BNP levels are also increased in patients with atrial fibrillation (AF) (11), even after controlling for demographic and clinical variables (12). The BNP levels appear to decrease after successful cardioversion to sinus rhythm (13 15). From the *Ullevål University Hospital, Oslo, Norway; Akershus University Hospital, University of Oslo, Oslo, Norway; University of California, San Diego, Veterans Affairs Medical Center, San Diego, California; University of California, San Francisco, San Francisco General Hospital, San Francisco, California; Hopital Bichat, Paris, France; Henry Ford Hospital, Detroit, Michigan; #University of Pennsylvania, Philadelphia, Pennsylvania; **University of Cincinnati College of Medicine, Cincinnati, Ohio; University of Kentucky College of Medicine, Lexington, Kentucky; and William Beaumont Hospital, Royal Oak, Michigan. Triage BNP devices and meters and some financial support were provided by Biosite, Inc., San Diego, California. Drs. Omland, Clopton, Wu, Duc, McCord, Nowak, Hollander, Abraham, McCullough, and Maisel have received honoraria from the manufacturer of the BNP assay used in the study. Dr. Marc Silver served as Guest Editor for this article. Manuscript received March 1, 2005; revised manuscript received May 6, 2005, accepted May 9, Currently, the principal clinical indication for BNP measurement is to diagnose heart failure (HF) in patients presenting with acute dyspnea. Using a commercially available, automated point-of-care device and a single prespecified cutoff of 100 pg/ml, we recently documented that measurement of BNP on admission provides valuable diagnostic information in this patient group (16), complementary and superior to clinical evaluation (17). Atrial fibrillation is not uncommon in patients presenting with acute dyspnea, but it is not known whether permanent/paroxysmal AF significantly affects circulating levels of BNP in dyspneic patients with and without HF. Moreover, it is unknown whether permanent/ paroxysmal AF affects the diagnostic performance of BNP in this setting and whether the conventional cutoff of 100 pg/ml provides optimal discrimination in patients with AF. To address these issues, we compared circulating levels and diagnostic performance of BNP in patients with and without AF in the Breathing Not Properly cohort. METHODS Patients. The Breathing Not Properly Multinational Study recruited patients from seven centers (5 U.S. and 2 Euro-

2 JACC Vol. 46, No. 5, 2005 September 6, 2005: Knudsen et al. BNP and AF 839 Abbreviations and Acronyms AF atrial fibrillation BNP B-type natriuretic peptide CHF congestive heart failure ECG electrocardiogram/electrocardiographic ED emergency department HF heart failure ROC receiver-operating characteristic pean teaching hospitals) from June 1999 to December In all, 1,586 patients were included. The inclusion criterion was a main complaint of dyspnea, and both patients with acute dyspnea of sudden onset and with worsening of chronic dyspnea were included. Patients with renal failure (calculated creatinine clearance 15 ml/min and those receiving dialysis therapy), acute myocardial infarction, age below 18 years, those who were unable to give informed consent, and patients whose dyspnea was clearly due to reasons other than acute HF (i.e., penetrating lung injury) were excluded from the study. The study design details and main results of the Breathing Not Properly Multinational Study have been published elsewhere (16). To determine the influence of the presence of permanent/paroxysmal AF on the diagnostic precision of BNP in the diagnosis of acute HF, this study encompasses patients in whom we had complete data with regard to the presence or absence of AF (n 1,431). Patients were classified according to permanent/paroxysmal AF and those without AF. The permanent/paroxysmal AF group of patients consists of the following patient subgroups: 1) patients with a medical history of AF and AF on the electrocardiogram (ECG) on admission (defined as permanent AF); 2) patients with a medical history of AF, but sinus rhythm on the ECG on admission (defined as paroxysmal AF); and 3) patients without a medical history of AF, but AF on the ECG on admission (defined as paroxysmal AF). The study conforms to the principles outlined in the Declaration of Helsinki, and the study protocol was approved by the institutional review boards of all participating hospitals. Written informed consent was obtained from all participants. Data collection. Research personnel in the emergency department (ED) collected baseline demographics, medical history, current clinical status, and blood test results for each patient. Electrocardiograms were interpreted by the attending ED physician and the following variables were recorded: signs of previous myocardial infarction, ST-segment deviation, left and right bundle branch block, and AF and QRS complex duration. The chest radiographs were evaluated by certified radiologists. Medical history, clinical findings, blood test results (except for BNP), ECG findings, and chest radiograph results were reported in a case report form completed by trained research personnel in collaboration with the attending physicians. It was the responsibility of the attending physicians to ensure that the case report form was in agreement with the clinical record. The attending ED physicians were, in general, certified cardiologists or senior staff emergency medicine physicians. The ED physician and radiologist were blind to the BNP level. Blood sampling. As soon as possible following admission, 5 ml of blood from each patient was collected into ethylenediaminetetraacetic acid tubes and the BNP concentration was measured using the Triage BNP test (Biosite Diagnostics Inc., San Diego, California). Precision, analytical sensitivity, and stability characteristics of the system have been described previously (18,19). The BNP tests were run in triplicate, and final results were reported as the mean of the three samples. Reference standard definition for the diagnosis of HF. Approximately 30 days after the ED visit, two cardiologists at each study center reviewed the case report form from the ED; ECG; chest radiograph; echocardiogram (if performed); and all other clinical tests, consultations, and chart information for each patient, except for the BNP value. On the basis of this information, the two cardiologists categorized all patients in one of the following categories: 1) dyspnea due to acute HF; 2) known chronic HF, but current dyspnea due to other reason; and 3) dyspnea due to a reason other than acute HF. For further statistical analyses, the latter two groups were pooled. Statistical analysis. SPSS for Windows version 11.0 (SPSS Inc., Chicago, Illinois) was used for all statistical analyses. Baseline characteristics are reported in percent or median and interquartile range. Univariate comparisons were made with chi-square tests for categorical variables and with Mann-Whitney U tests for continuous variables. A two-sided p value 0.05 was considered significant. For the evaluation of diagnostic precision of BNP in patients with and without AF, patients were divided into two groups: patients with permanent/paroxysmal AF and patients without AF. Sensitivity, specificity, accuracy, positive and negative predictive values, and positive and negative likelihood ratios for the detection of acute HF were calculated for BNP in both groups by use of different cutoff values (50, 100, 200, 300, 400, 500, 600, 700, and 800 pg/ml). Accuracy is the degree to which the diagnostic decision agrees the true value of the diagnosis in this case, the proportion of correct diagnoses (true positives true negatives)/(true positives false positives true negatives false negatives). As with all diagnostic statistics, this assumes the adjudicated diagnosis is correct. Accuracy changes with prevalence unless sensitivity and specificity are equal. Median BNP value and interquartile range were calculated in patients with permanent/paroxysmal AF with and without acute HF and in patients without AF with and without acute HF. The independent diagnostic value of permanent/paroxysmal AF in patients without a final diagnosis of HF with regard to an elevated BNP level ( 100 pg/ml) was assessed in a multivariate logistic model for all variables that proved to be significant by univariate analysis in patients with and without a final diagnosis of HF. The following dichotomous variables were evaluated in the

3 840 Knudsen et al. JACC Vol. 46, No. 5, 2005 BNP and AF September 6, 2005: Table 1. Patient Characteristics multivariate model: medical history of CHF, myocardial infarction and hypertension, clinical signs of jugular vein distension, orthopnea, lower extremity edema and rales, use of digoxin and diuretics, QRS duration on the ECG, and estimated creatinine clearance. To obtain an index of overall diagnostic performance of BNP in patients with and without AF, receiver-operating characteristic (ROC) curves were generated and the area under the curve calculated. On the basis of the ROC curves, optimal diagnostic cutoff values for the diagnosis of acute HF were obtained in patients both with and without AF. RESULTS Permanent/ Paroxysmal Atrial Fibrillation No Atrial Fibrillation p Value Total number Patients 292 1,139 Demographics Age (yrs) 77 (67 82) 62 (49 74) Gender (female) 113 (38.7) 502 (44.1) Medical history Congestive heart failure 136 (46.6) 344 (30.2) Myocardial infarction 94 (32.2) 259 (22.7) Aterial hypertension 165 (56.5) 634 (55.7) Diabetes mellitus 74 (25.3) 251 (22.0) Clinical findings Jugular vein distension 122 (41.8) 215 (18.9) Orthopnea 172 (58.9) 633 (55.6) Lower extremity edema 178 (61.0) 397 (34.9) Rales 186 (63.7) 446 (39.2) Body mass index (kg/m 2 ) 25.7 (22 31) 27.4 (23 33) Medication at home ACE inhibitors 117 (40.1) 322 (28.3) Beta-blockers 74 (25.3) 230 (20.2) Calcium channel blockers 87 (29.8) 208 (18.3) Digoxin 128 (43.8) 131 (11.5) Other antiarrhythmics 47 (16.1) 44 (3.9) Diuretics 184 (63.0) 475 (41.7) ECG Atrial fibrillation 158 (54.1) Left bundle branch block 27 (9.2) 58 (5.1) Right bundle branch 29 (9.9) 54 (4.7) block QRS duration (ms) 100 (86 128) 90 (80 104) Blood values Creatinine clearance 51 (34 72) 65 (43 90) BNP 100 pg/ml 251 (86.0) 541 (47.5) BNP (pg/ml) 421 ( ) 82 (18 506) Final diagnosis Acute heart failure 219 (75.0) 445 (39.1) ACE angiotensin-converting enzyme; BNP B-type natriuretic peptide; ECG electrocardiogram. The current analysis includes 1,431 patients who all presented with a main complaint of dyspnea. Table 1 summarizes the characteristics of the total number of patients, subdivided according to the presence or absence of AF. Of the 292 patients with permanent/paroxysmal AF, 256 patients had a medical history of AF. Of this population, 122 patients presented with ECG-documented AF in the ED. Thirty-six patients who had no previous history of AF presented with AF on the ECG at the current admission. The patients with permanent/paroxysmal AF were significantly older than patients without AF (p ) and they were more likely to have a medical history of CHF, myocardial infarction (p for both), and diabetes (p 0.039). Jugular vein distension, lower extremity edema, and rales (p for all) were significantly more common among patients with permanent/paroxysmal AF, as well as use of angiotensin-converting enzyme inhibitors (p ), beta-blockers (p 0.042), calcium-channel blockers, digoxin, other antiarrhythmics, and diuretics (p for all). A final diagnosis of acute HF was significantly more common in the AF group of patients than in the group of patients without AF (p ). Circulating BNP levels were significantly higher in patients with permanent/paroxysmal AF compared to those without AF: 421 (interquartile range 168 to 911) pg/ml versus 82 (interquartile range 18 to 506) pg/ml (p ). Circulating BNP levels in patients with/without AF and with/without a final diagnosis of acute HF are depicted in Figure 1. In patients without HF, the BNP values in patients with permanent/paroxysmal AF were significantly higher than in those without AF (p ). Conversely, in patients with HF, the BNP values did not differ significantly between patients with or without AF (p 0.811). In patients without a final diagnosis of HF, AF (odds ratio 3.2, 95% confidence interval 1.3 to 8.1) remained an independent predictor of elevated BNP levels ( 100 pg/ml) after adjustment for medical history of CHF, myocardial infarction and hypertension, clinical signs of jugular vein distension, orthopnea, lower extremity edema and rales, use of digoxin and diuretics, QRS duration on the ECG, and estimated creatinine clearance. Characteristics of Figure 1. B-type natriuretic peptide (BNP) levels in patients with and without atrial fibrillation and with and without a final diagnosis of acute heart failure. In patients without heart failure, BNP levels were significantly higher in patients with atrial fibrillation than in those without (p 0.001). In patients with heart failure, BNP levels were not significantly different between patients with and without atrial fibrillation. B-type natriuretic peptide levels are plotted on a log scale. The boxes represent the 25th, 50th, and 75th percentiles, and the whiskers indicate the 10th and 90th percentiles.

4 JACC Vol. 46, No. 5, 2005 September 6, 2005: Knudsen et al. BNP and AF 841 Table 2. Patient Characteristics in Patients With Permanent/Paroxysmal Atrial Fibrillation Subdivided According to a Final Diagnosis of Acute Congestive Heart Failure or Not Permanent/Paroxysmal Atrial Fibrillation and CHF Permanent/Paroxysmal Atrial Fibrillation and No CHF p Value Total number of patients Demographics Age (yrs) 77 (67 83) 74 (69 79) Gender (female) 87 (39.7) 26 (35.6) Medical history Congestive heart failure 123 (56.2) 13 (17.8) Myocardial infarction 81 (37.0) 13 (17.8) Arterial hypertension 130 (59.4) 35 (47.9) Diabetes mellitus 55 (25.1) 19 (26.0) Clinical findings Jugular vein distension 109 (49.8) 13 (17.8) Orthopnea 144 (65.8) 28 (38.4) Lower extremity edema 152 (69.4) 26 (35.6) Rales 158 (72.1) 28 (38.4) Body mass index (kg/m 2 ) 26 (22 32) 25 (22 30) Medication at home ACE inhibitors 94 (42.9) 23 (31.5) Beta-blockers 59 (26.9) 15 (20.5) Calcium channel blockers 63 (28.8) 24 (32.9) Digoxin 105 (47.9) 23 (31.5) Other antiarrhythmics 39 (17.8) 8 (11.0) Diuretics 151 (68.9) 33 (45.2) ECG Left bundle branch block 20 (9.1) 7 (9.6) Right bundle branch block 23 (10.5) 6 (8.2) QRS duration (ms) 102 (92 132) 90 (81 114) Blood values Creatinine clearance 48 (33 68) 59 (44 80) BNP 100 pg/ml 207 (94.5) 44 (60.3) BNP (pg/ml) 569 (290 1,031) 119 (69 244) Values given as n (%) or median (interquartile range). CHF congestive heart failure; other abbreviations as in Table 2. patients with permanent/paroxysmal AF subdivided according to a final diagnosis of acute HF are summarized in Table 2. Patients with a final diagnosis of acute HF were more likely to have a medical history of CHF, myocardial infarction and hypertension, clinical signs compatible with HF, medication with digoxin and diuretics, and decreased creatinine clearance than patients without a final diagnosis of CHF. Median BNP value was 569 (interquartile range 290 to 1,031) and 119 (interquartile range 69 to 244) in patients with and without HF, respectively. Diagnostic performance for different cutoff values for BNP in the diagnosis of acute HF subdivided according to the presence or absence of AF are presented in Table 3. In patients with permanent/paroxysmal AF, a BNP cutoff value of 100 pg/ml had a specificity of only 40%, whereas the sensitivity was 95%. Not surprisingly, specificity increased and sensitivity decreased with increasing BNP cutoff values in both patient groups. The highest accuracy in both groups of patients was achieved at a BNP cutoff value of 200 pg/ml: 82.2 for patients with permanent/paroxysmal AF and 84.5 for patients without AF. To assess the total diagnostic performance of BNP for diagnosing acute HF, ROC analyses (Fig. 2) revealed an area under the curve of 0.84 (interquartile range 0.78 to 0.89) and 0.91 (interquartile range 0.89 to 0.93) for patients with permanent/ paroxysmal AF and patients without AF, respectively. There was no significant difference in ROC area under the curve for patients with current (n 158: 0.80 [95% confidence interval 0.69 to 0.90]) versus those with a history of but no current AF (n 134: 0.86 [95% confidence interval 0.80 to 0.93]). Neither was there any significant difference in the BNP value between patients with current versus those with a history of but no current AF (377.8 [interquartile range to 912.3] vs [interquartile range to 910.1], p 0.189) (Fig. 3). DISCUSSION Our study demonstrates that in dyspneic patients without HF, permanent/paroxysmal AF is associated with increased circulating BNP levels. This pattern was not evident in patients with a final diagnosis of HF. Although the overall diagnostic performance of BNP for HF tended to be lower in patients with AF than in those without, BNP performed well in both groups. Our data also demonstrate that the conventional cutoff value of 100 pg/ml was associated with

5 842 Knudsen et al. JACC Vol. 46, No. 5, 2005 BNP and AF September 6, 2005: Table 3. Diagnostic Performance of Different BNP Cutoff Values in the Diagnosis of Acute Heart Failure Patients BNP Sensitivity Specificity Positive Predictive Value Negative Predictive Value Accuracy Positive Likelihood Ratio Negative Likelihood Ratio Atrial fibrillation 50 pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml No atrial fibrillation 50 pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml pg/ml BNP B-type natriuretic peptide. a markedly lower specificity and positive likelihood ratio in patients with AF than in those without, suggesting that a higher diagnostic threshold should be used to diagnose HF in patients with AF. The most widely accepted indication for BNP measurement in clinical practice is for the emergency diagnosis of HF in patients presenting with acute dyspnea. The first Figure 2. Receiver-operating characteristic curve for various cutoff levels of B-type natriuretic peptide in differentiating between dyspnea due to acute heart failure and dyspnea due to other causes in patients with and without atrial fibrillation. The area under the curve was 0.84 (95% confidence interval 0.78 to 0.89) and 0.91 (95% confidence interval 0.89 to 0.93) for those with and without atrial fibrillation, respectively. study suggesting that BNP measurement could prove useful in this setting was published 10 years ago (20). However, clinical use of this test was limited until the publication of the Breathing Not Properly Multinational Study, a multicenter diagnostic test evaluation study that included 1,586 acutely dyspneic patients, which documented that measurement of BNP on admission provides valuable diagnostic information in this patient group (16,17) complementary and superior to clinical evaluation. However, even in the absence of left-sided HF, a number of other factors, including advanced patient age and female gender (21 23); decreased body mass index (24 26); decreased renal function (9,27); left ventricular hypertrophy (28); history of myocardial infarction and ongoing cardiac ischemia (29 33), pulmonary embolism (34); chronic right ventricular failure (35); and cardiac arrhythmias, including AF (12,36); have previously been associated with elevated BNP levels. There have been concerns that the presence of one or more of these factors could obscure the association between BNP and HF, thereby reducing diagnostic performance. Our findings show that AF is associated with increased BNP levels in patients without HF, but this pattern was not observed in dyspneic patients with HF, probably because ventricular dysfunction is associated with greater BNP release than is AF. Whereas the ventricular myocardium is believed to be the major source of circulating BNP in HF, the cardiac atria, containing much less muscle mass than the left ventricle, may be the predominant source of BNP in AF. Accordingly, increased pro-bnp messenger ribonucleic acid expression in right atrial appendages has been documented in patients with persistent AF (37). Moreover, selective catheterization of the coronary sinus and the

6 JACC Vol. 46, No. 5, 2005 September 6, 2005: Knudsen et al. BNP and AF 843 diagnostic threshold should be used in patients with AF. Moreover, the total diagnostic performance of BNP in the diagnosis of acute HF was somewhat lower for patients with AF compared to those without, but BNP still performed well in both groups. Reprint requests and correspondence: Dr. Torbjørn Omland, Department of Medicine, Akershus University Hospital, University of Oslo, N-1474 Nordbyhagen, Norway. REFERENCES Figure 3. B-type natriuretic peptide levels in patients with atrial fibrillation on the electrocardiogram (ECG) on admission (n 158), and in patients with a medical history of atrial fibrillation but sinus rhythm on the ECG on admission (n 134) (p 0.189). B-type natriuretic peptide (BNP) levels are plotted on a log scale. The boxes represent the 25th, 50th, and 75th percentiles, and the whiskers indicate the 10th and 90th percentiles. anterior interventricular vein in patients with lone fibrillation has suggested that BNP is produced in the atrium in AF (38). Even in patients without AF, BNP levels appear to be higher in atrial tissue than ventricular tissue on a gram per wet weight basis (3). The diagnostic performance of BNP tended to be lower in patients with AF than in those without. This can probably be ascribed to the higher levels of BNP in patients with AF and no HF than in those without AF and no HF. The difference was modest and the overall diagnostic performance, as assessed by ROC analyses, was not markedly different between patients with and without AF. Looking at the effect of different specific thresholds, however, it seemed that a cutoff of 200 pg/ml resulted in a marked improvement in specificity and positive likelihood ratio compared with the conventional cutoff of 100 pg/ml, with little loss of sensitivity in patients with AF. Accordingly, ROC analysis suggested that 262 pg/ml represented the optimal tradeoff between sensitivity and specificity in patients with AF (optimal defined here as closest to the upper left corner), which was higher than for patients without AF (162 pg/ml). Clinicians are therefore probably best advised to interpret mildly elevated BNP tests with caution, and additional, more specific investigations may be required. Study limitations. The interpretation of permanent/ paroxysmal AF was based on previous medical records or information from the patient, and misclassifications possibly exist. These might have led to an underestimation of the presence of permanent/paroxysmal AF. The fact that there was no objective gold standard for the diagnosis of acute HF is another important limitation. Conclusions. We have found that in patients without, but not in those with HF, the presence of AF was associated with higher circulating BNP levels, suggesting that a higher 1. Yasue H, Yoshimura M, Sumida H, et al. Localization and mechanism of secretion of B-type natriuretic peptide in comparison with those of A-type natriuretic peptide in normal subjects and patients with heart failure. Circulation 1994;90: Ogawa Y, Nakao K, Mukoyama M, et al. Rat brain natriuretic peptide: tissue distribution and molecular form. Endocrinology 1990;126: Mukoyama M, Nakao K, Hosoda K, et al. Brain natriuretic peptide as a novel cardiac hormone in humans. Evidence for an exquisite dual natriuretic peptide system, atrial natriuretic peptide and brain natriuretic peptide. J Clin Invest 1991;87: Hosoda K, Nakao K, Mukoyama M, et al. Expression of brain natriuretic peptide gene in human heart. Production in the ventricle. Hypertension 1991;17: Ogawa Y, Nakao K, Mukoyama M, et al. Natriuretic peptides as cardiac hormones in normotensive and spontaneously hypertensive rats. The ventricle is a major site of synthesis and secretion of brain natriuretic peptide. Circ Res 1991;69: Hama N, Itoh H, Shirakami G, et al. Rapid ventricular induction of brain natriuretic peptide gene expression in experimental acute myocardial infarction. Circulation 1995;92: Mukoyama M, Nakao K, Saito Y, et al. Increased human brain natriuretic peptide in congestive heart failure. N Engl J Med 1990; 323: Richards AM. The natriuretic peptides in heart failure. Basic Res Cardiol 2004;99: McCullough PA, Duc P, Omland T, et al. B-type natriuretic peptide and renal function in the diagnosis of heart failure: an analysis from the Breathing Not Properly Multinational Study. Am J Kidney Dis 2003;41: McCullough PA, Joseph K, Mathur VS. Diagnostic and therapeutic utility of B-type natriuretic peptide in patients with renal insufficiency and decompensated heart failure. Rev Cardiovasc Med 2004;5: Wallen T, Landahl S, Hedner T, Saito Y, Masuda I, Nakao K. Brain natriuretic peptide in an elderly population. J Intern Med 1997;242: Silvet H, Young-Xu Y, Walleigh D, Ravid S. Brain natriuretic peptide is elevated in outpatients with atrial fibrillation. Am J Cardiol 2003;92: Beck-da-Silva L, de Bold A, Fraser M, Williams K, Haddad H. Brain natriuretic peptide predicts successful cardioversion in patients with atrial fibrillation and maintenance of sinus rhythm. Can J Cardiol 2004;20: Vinch CS, Rashkin J, Logsetty G, et al. Brain natriuretic peptide levels fall rapidly after cardioversion of atrial fibrillation to sinus rhythm. Cardiology 2004;102: Wozkowska-Kapton B. Effect of sinus rhythm restoration on plasma brain natriuretic peptide in patients with atrial fibrillation. Am J Cardiol 2004;93: Maisel A, Krishnaswamy P, Nowak R, et al. Rapid measurement of B-type natriuretic peptide in the emergency diagnosis of heart failure. N Engl J Med 2002;347: McCollough PA, Nowak RM, McCord J, et al. B-type natriuretic peptide and clinical judgement in the emergency diagnosis of heart failure. Circulation 2002;106: Harrison A, Morrison LK, Krishnaswamy P, Kazanegra R, Clopton P, Dao Q. B-type natriuretic peptide predicts future cardiac events in patients presenting to the emergency department with dyspnea. Ann Emerg Med 2002;39:131 8.

7 844 Knudsen et al. JACC Vol. 46, No. 5, 2005 BNP and AF September 6, 2005: Morrison LK, Harrison A, Krishnaswamy P, Kazanegra R, Clopton P, Maisel A. Utility of a rapid B-natriuretic peptide assay in differentiating congestive heart failure from lung disease in patients presenting with dyspnea. J Am Coll Cardiol 2002;39: Davis M, Espiner E, Richards G, et al. Plasma brain natriuretic peptide in assessment of acute dyspnea. Lancet 1994;343: Raymond I, Groenning BA, Hildebrandt PR, et al. The influence of age, sex and other variables on the plasma level of N-terminal pro brain natriuretic peptide in a large sample of the general population. Heart 2003;89: Redfield MM, Rodeheffer RJ, Jacobsen SJ, Mahoney DW, Bailey KR, Burnett JC, Jr. Plasma brain natriuretic peptide concentration: impact of age and gender. J Am Coll Cardiol 2002;40: Wang TJ, Larson MG, Levy D, et al. Impact of age and sex on plasma natriuretic peptide levels in healthy adults. Am J Cardiol 2002;90: Mehra MR, Uber PA, Park MH, et al. Obesity and suppressed B-type natriuretic peptide levels in heart failure. J Am Coll Cardiol 2004;43: Wang TJ. Impact of obesity on plasma natriuretic peptide levels. Circulation 2004;109: McCord J, Mundy BJ, Hudson MP, et al. The relationship between obesity and B-type natriuretic peptide levels. Arch Intern Med 2004;164: McCullough PA, Joseph K, Mathur VS. Diagnostic and therapeutic utility of B-type natriuretic peptide in patients with renal insufficiency and decompensated heart failure. Rev Cardiovasc Med 2004;5: Yamamoto K, Burnett JC, Jr., Jougasaki M, et al. Superiority of brain natriuretic peptide as a hormonal marker of ventricular systolic and diastolic dysfunction and ventricular hypertrophy. Hypertension 1996; 28: Foy SG, Crozier IG, Richards AM, et al. Neurohormonal changes after acute myocardial infarction. Relationships with haemodynamic indices and effects of ACE inhibition. Eur Heart J 1995;16: Morita E, Yasue H, Yoshimura M, et al. Increased plasma levels of brain natriuretic peptide in patients with acute myocardial infarction. Circulation 1993;88: Omland T, Aakvaag A, Bonarjee VVS, et al. Plasma brain natriuretic peptide as an indicator of left ventricular systolic function and long-term prognosis after acute myocardial infarction. Comparison with plasma atrial natriuretic peptide and N-terminal proatrial natriuretic peptide. Circulation 1996;93: Mukoyama M, Nakao K, Obata K, et al. Augmented secretion of brain natriuretic peptide in acute myocardial infarction. Biochem Biophys Res Commun 1991;180: Talwar S, Squire IB, Downie PF, Davies JE, Ng LL. Plasma N terminal pro-brain natriuretic peptide and cardiotrophin 1 are raised in unstable angina. Heart 2000;84: Kruger S, Graf J, Merx MW, et al. Brain natriuretic peptide predicts right heart failure in patients with acute pulmonary embolism. Am Heart J 2004;147: Ishii J, Nomura M, Ito M, et al. Plasma concentration of brain natriuretic peptide as a biochemical marker for the evaluation of right ventricular overload and mortality in chronic respiratory disease. Clin Chim Acta 2000;301: Rossi A, Enriquez-Sarano M, Burnett JC Jr., Lerman A, Abel MD, Seward JB. Natriuretic peptide levels in atrial fibrillation: a prospective hormonal and Doppler-echocardiographic study. J Am Coll Cardiol 2000;35: Tuinenburg AE, Brundel BJ, Van Gelder IC, et al. Gene expression of the natriuretic peptide system in atrial tissue of patients with paroxysmal and persistent atrial fibrillation. J Cardiovasc Electrophysiol 1999;10: Inoue S, Murakami Y, Sano K, Katoh H, Shimada T. Atrium as a source of brain natriuretic polypeptide in patients with atrial fibrillation. J Card Fail 2000;6:92 6.