- Email: [email protected]
N-Terminal Pro-B-Type Natriuretic Peptide and B-Type Natriuretic Peptide for Identifying Coronary Artery Disease and Left Ventricular Hypertrophy in Ambulatory Chronic Kidney Disease Patients
N-Terminal Pro-B-Type Natriuretic Peptide and B-Type Natriuretic Peptide for Identifying Coronary Artery Disease and Left Ventricular Hypertrophy in Ambulatory Chronic Kidney Disease Patients Ijaz A. Khan, MDa, Jeffrey Fink, MD, MSb, Caitlin Nass, NPa, Hegang Chen, PhDc, Robert Christenson, PhDd, and Christopher R. deFilippi, MDa,* Elevated natriuretic peptide levels are common in patients with chronic kidney disease (CKD), as is the presence of coronary artery disease (CAD) and left ventricular hypertrophy (LVH). It was hypothesized that N-terminal pro–B-type natriuretic peptide (NT–pro-BNP) and B-type natriuretic peptide (BNP) levels could identify CAD and LVH in asymptomatic patients with CKD. Clinical, laboratory, and echocardiographic data were collected prospectively in 54 ambulatory patients with CKD not requiring dialysis. CAD was defined by previous myocardial infarction or coronary revascularization. The median age was 70 years (interquartile range [IQR] 57 to 76). Fourteen patients (26%) had CAD, and 30 (56%) had LVH. Median NT–proBNP was 724 pg/ml (IQR 168 to 2,950), median BNP was 137 pg/ml (IQR 31 to 391), and the median glomerular filtration rate (GFR) was 31 ml/min/1.73 m2 (IQR 21 to 42). A strong correlation was found between NT–pro-BNP and BNP levels (R ⴝ 0.74, p <0.0001), but only moderate correlations were found between NT–pro-BNP and GFR (R ⴝ ⴚ0.45, p ⴝ 0.0006) and between BNP and GFR (R ⴝ ⴚ0.38, p ⴝ 0.005). There was no trend of an increase in the prevalence of LVH or CAD with decreasing GFR. However, across progressive NT–pro-BNP and BNP quartiles, the prevalences of LVH and CAD increased significantly. Receiver-operating characteristic curves showed that these 2 markers are similar and significant predictors for indicating LVH (area under the curve [AUC] 0.72, p ⴝ 0.005 for NT–pro-BNP; AUC 0.72, p ⴝ 0.007 for BNP) and CAD (AUC 0.80, p ⴝ 0.001 for NT–pro-BNP; AUC 0.82, p ⴝ 0.0004 for BNP; p ⴝ 0.45 for NT–pro-BNP vs BNP). In conclusion, NT–pro-BNP and BNP levels are significant and equivalent indicators of CAD and LVH in asymptomatic patients with CKD. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol 2006; 97:1530 –1534)
The prevalences of coronary artery disease (CAD) and left ventricular hypertrophy (LVH) are greater in patients with chronic kidney disease (CKD) than in the general population.1–3 Most of these patients remain asymptomatic despite marked increases in the prevalences of CAD and LVH. The detection of CAD in these patients has remained elusive, and the detection of LVH often requires echocardiography, which is expensive when the magnitude of the population is considered.3 N-terminal pro–B-type natriuretic peptide (NT–pro-BNP) and B-type natriuretic peptide (BNP) levels can predict the presence of CAD and LVH in the general population.4 – 8 However, uncertainty remains as to whether
elevated levels of 1 or both of these markers are predominantly a function of a reduced glomerular filtration rate (GFR) or maintains a predictive value for heart disease.9 –12 Because these 2 markers correlate well for predicting patients with congestive heart failure,13,14 we hypothesized that NT–pro-BNP and BNP levels would have similar abilities to identify CAD and LVH in asymptomatic patients with CKD. To test this hypothesis, we prospectively collected simultaneous clinical, laboratory, and echocardiographic data from ambulatory patients with CKD being examined at routine medical visits.
Divisions of aCardiology and bNephrology and Departments of Epidemiology and dPathology, University of Maryland School of Medicine, Baltimore, Maryland. Manuscript received August 26, 2005; revised manuscript received and accepted November 28, 2005. This study was funded by an investigator-initiated grant from Roche Diagnostics Corporation, Indianapolis, Indiana. * Corresponding author: Tel: 410-328-7204; fax: 410-328-3530. E-mail address: [email protected] (C.R. deFilippi).
Methods Study population and design: Enrollment into the study required that patients have serum creatinine of ⬎1.2 mg/dl, not be on dialysis, and be ⱖ18 years of age. The evaluation involved reviewing the patients’ cardiac histories and cardiac risk factors, taking blood samples, and performing transthoracic echocardiography. The protocol was ap-
c
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2005.11.090
www.AJConline.org
Miscellaneous/B-Type Natriuretic Peptides in Renal Disease
proved by the institutional review board. Informed consent was obtained from each participant. Fifty-four of 70 patients who met enrollment criteria and signed the informed consent form completed all testing. Data collection: Patients’ demographic and clinical information was obtained from electronic medical record reviews. History of diabetes, hypertension, peripheral vascular disease, smoking, current alcohol use, and CAD were included if they were identified in a patient’s history and/or physical examination results. CAD was defined by previous myocardial infarction or coronary revascularization. Smoking history was stratified as never having smoked or ever having smoked, currently or in the past. Height and weight data obtained during clinic visits were used to calculate body mass index. Values for total cholesterol and lowdensity lipoprotein cholesterol were collected from the clinical records. Hypercholesterolemia was defined as total cholesterol ⬎200 mg/dl or low-density lipoprotein cholesterol ⬎100 mg/dl. Creatinine values were obtained clinically for all patients and were used to estimate the GFR using the abbreviated Modification of Diet in Renal Disease study formula.15 For biomarker assays, plasma samples were collected and centrifuged at 1,000g for 12 minutes. The resulting serum and plasma were aliquoted, frozen, and maintained at ⫺70°C. NT–pro-BNP was measured from serum using an immunoassay (Elecsys proBNP, Roche Diagnostics Corporation, Indianapolis, Indiana). BNP was measured from plasma using an immunoassay (Triage BNP, Biosite Diagnostics, San Diego, California). A second assessment of renal function was made by measuring cystatin C from serum using a nephelometric immunoassay (BN Prospect N Latex Cystatin C, Dade Behring, Glasgow, Delaware). Transthoracic echocardiograms were obtained on the day of enrollment for the purposes of this study and not for clinical indications. All studies were performed by the same technician using a Phillips Sonos 5500 (Philips Medical Systems, Andover, Massachusetts) or SonoSite SonoHeart Elite (SonoSite, Inc., Bothell, Washington). Echocardiograms were analyzed to quantify the left ventricular ejection fraction and left ventricular mass. Echocardiographic interpretations were performed without knowledge of biomarker results. Left ventricular mass was determined using the formula of Devereux et al16 and was adjusted for patient height.17 LVH was determined after adjustment for patient height.17 The left ventricular ejection fraction was calculated using end-diastolic volume and end-systolic volume measurements from apical 4- and 2-chamber views calculated by Simpson’s method.18 Statistical analysis: Data from continuous variables are expressed as medians with interquartile ranges (IQRs). Categorical variables are expressed as absolute numbers and percentages. The Cochran-Armitage test for trend was used to analyze trends among BNP quartiles, NT–pro-BNP quartiles, and CKD stages with categorical clinical and echocar-
1531
Table 1 Patient characteristics (n ⫽ 54) Characteristic Age (yrs) Men African–American Diabetes mellitus Hypertension Smoking Hypercholesterolemia Coronary artery disease Left ventricular ejection fraction ⱕ40% LVH Diastolic dysfunction Body mass index (kg/m2) Glomerular filtration rate (ml/min/1.73 m2) NT–pro-BNP (pg/ml) BNP (pg/ml)
Value 70 (57.3–76), 66.5 ⫾ 13.8 53 (98%) 24 (44%) 28 (52%) 51 (94%) 45 (83%) 29 (53%) 14 (26%) 4 (7%) 30 (56%) 31 (57%) 28.3 (25.3–32.3), 29.7 ⫾ 6.7 31 (21–42), 34 ⫾ 21 728 (168–2,950), 2,552 ⫾ 4,143 138 (31–391) 271 ⫾ 326
Data are presented or number (percentages), mean ⫾ SD, and median (IQR).
diographic variables. The Kruskal-Wallis test is a nonparametric test that was used to evaluate the relations among NT–pro-BNP quartiles, BNP quartiles, and CKD stages with continuous clinical, laboratory, and echocardiographic variables. For each subject, GFR versus NT–pro-BNP level, GFR versus BNP level, 1/cystatin C versus NT–pro-BNP level, 1/cystatin C versus BNP level, and NT–pro-BNP level versus BNP level were plotted, and Pearson’s correlation coefficient was calculated for these relationss. Receiver-operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was calculated to determine and compare the ability of NT–pro-BNP and BNP levels to indicate LVH and CAD. The AUC of NT– pro-BNP was compared with that of BNP using DeLong, DeLong, and Clarke-Pearson’s method of the comparison of 2 ROC curves derived from correlated samples. Two-sided p values ⬍0.05 were considered statistically significant. Analyses were performed using SAS version 8.2 (SAS Institute Inc., Cary, North Carolina) and AccuROC version 2.5 (Accumetric Corporation, Montreal, Quebec, Canada).
Results Of 54 study patients, 53 were men (98%), and 24 were African-Americans (44%). The median age was 70 years (IQR 57 to 76). Fourteen patients (26%) had CAD, and 30 (56%) had LVH. Only 4 patients (7%) had left ventricular ejection fractions ⱕ40%. The median NT–pro-BNP level was 724 pg/ml (IQR 168 to 2,950), and the median BNP level was 137 pg/ml (IQR 31 to 391). The median GFR was 31 ml/min/1.73 m2 (IQR 21 to 42) (Table 1). A significant positive correlation was found between NT–pro-BNP and BNP levels (R ⫽ 0.74, p ⬍0.0001).
1532
The American Journal of Cardiology (www.AJConline.org)
Table 2 Trends of CAD and LVH stratified by glomerular filtration rate, BNP, and NT–pro-BNP Variable
GFR Stages 1 and 2 (⬎90–60 ml/min/1.73 m2) (n ⫽ 6)
Stage 3 (30–59 ml/min/1.73 m2) (n ⫽ 26)
Stage 4 (15–29 ml/min/1.73 m2) (n ⫽ 11)
Stage 5 (⬍15 ml/min/1.73 m2) (n ⫽ 11)
p Value
1 (17%) 4 (67%) 239 (115–271) 82 (41–107)
6 (23%) 12 (46%) 315 (90–782) 45 (18–228)
4 (36%) 7 (64%) 1414 (226–5,788) 200 (60–388)
3 (27%) 7 (64%) 4376 (1,070–13,798) 423 (310–634)
0.50 0.56 0.0003 0.001
CAD LVH NT–pro-BNP, median (IQR) BNPm median (IQR)
NT–pro-BNP
CAD LVH
Quartile 1 (16–144 pg/ml) (n ⫽ 13)
Quartile 2 (160–685 pg/ml) (n ⫽ 14)
Quartile 3 (762–2,273 pg/ml) (n ⫽ 13)
Quartile 4 (3,175–1,7493 pg/ml) (n ⫽ 14)
p Value
0 (0%) 4 (31%)
1 (7%) 7 (50%)
7 (54%) 7 (54%)
6 (43%) 12 (86%)
0.001 0.005
BNP
CAD LVH
Quartile 1 (4–31 pg/ml) (n ⫽ 14)
Quartile 2 (32–114 pg/ml) (n ⫽ 13)
Quartile 3 (161–388 pg/ml) (n ⫽ 13)
Quartile 4 (392–1,520 pg/ml) (n ⫽ 14)
p Value
0 (0%) 4 (29%)
1 (8%) 8 (62%)
6 (46%) 6 (46%)
7 (50%) 12 (86%)
0.0004 0.008
Figure 1. ROC curves of NT–pro-BNP (AUC 0.80, p ⫽ 0.001) and BNP (AUC 0.82, p ⫽ 0.0004) as indicators of CAD (p ⫽ 0.45, for comparison between ROC curves).
Moderate correlations were found between NT–pro-BNP and GFR (R ⫽ ⫺0.45, p ⫽ 0.0006) and between BNP and GFR (R ⫽ ⫺0.38, p ⫽ 0.005). Using the reciprocal of cystatin C as an alternative method to estimate renal function, the correlations between NT–pro-BNP and 1/cystatin C (R ⫽ ⫺0.45, p ⫽ 0.0006) and BNP and 1/cystatin C (R
⫽ ⫺0.35, p ⫽ 0.01) remained moderate and were similar to the correlations determined by the estimated GFR. When patients were stratified by CKD stages, the NT– pro-BNP level increased progressively from stage 1 to 2 to stage 5 CKD (p ⫽ 0.0003). Similarly, BNP levels also increased from stage 1 to 2 to stage 5 of the GFR, with a positive trend (p ⫽ 0.001). However, there was no trend for an increase in the prevalence of LVH or CAD by increasing CKD stage. In contrast, when patients were stratified across progressive NT–pro-BNP quartiles, the prevalence of LVH and CAD increased (p ⫽ 0.005 and p ⫽ 0.001, respectively). Similarly, when patients were stratified across progressive BNP quartiles, the prevalence of LVH and CAD increased (p ⫽ 0.008 and p ⫽ 0.0004, respectively) (Table 2). The ROC curves for NT–pro-BNP (AUC 0.80, p ⫽ 0.001) and BNP (AUC 0.82, p ⫽ 0.0004) as indicators of CAD are shown in Figure 1. The ROC curves of NT–proBNP (AUC ⫽ 0.72, p ⫽ 0.005) and BNP (AUC ⫽ 0.72, p ⫽ 0.007) as predictors of LVH are shown in Figure 2. A comparison between the ROC curves for NT–pro-BNP and BNP for indicating CAD demonstrated that the AUCs for the 2 tests were not different (p ⫽ 0.45, for the AUC comparison). Similarly, a comparison between the ROC curves of NT–pro-BNP and BNP for identifying LVH demonstrated that the AUCs were not different between the 2 tests (p ⫽ 0.78, for the AUC comparison). On the basis of the ROC analysis, the optimal value of NT–pro-BNP as an indicator of CAD was 979 pg/ml, with a sensitivity of 79%, a specificity of 70%, a positive predictive value (PPV) of 48%, and a negative predictive value (NPV) of 90%. The optimal value of BNP to indicate CAD
Miscellaneous/B-Type Natriuretic Peptides in Renal Disease
Figure 2. ROC curve of NT–pro-BNP (AUC 0.72, p ⫽ 0.005) and BNP (AUC 0.72, p ⫽ 0.007) as indicators of LVH (p ⫽ 0.78, for comparison between ROC curves).
was 228 pg/ml, with a sensitivity of 86%, a specificity of 73%, a PPV of 52%, and a NPV of 94%. The optimal value of NT–pro-BNP as a predictor of LVH was 762 pg/ml, with a sensitivity of 63%, a specificity of 67%, a PPV of 70%, and a NPV of 57%. The optimal value of BNP as a predictor of LVH was 200 pg/ml, with a sensitivity of 60%, a specificity of 71%, a PPV of 72%, and a NPV of 59%.
Discussion The present study demonstrates that in ambulatory patients with CKD not requiring dialysis, NT–pro-BNP and BNP can identify, with similar accuracy, patients with LVH and CAD. Serum levels of NT–pro-BNP and BNP generally increase with worsening renal impairment.8,11,12 Furthermore, it has been hypothesized that NT–pro-BNP may be more dependent on renal clearance than BNP,12 limiting the diagnostic accuracy of the test in this setting. Our results do indicate a greater increase in NT–pro-BNP levels compared with BNP levels from the established reference ranges.19,20 However, such a finding is also common in a variety of disease states, such as congestive heart failure13,14,21 and acute coronary syndromes22,23 in patients with normal renal function. To further support the equivalency of the response of NT–pro-BNP and BNP in the CKD setting, independent of their predictive value, the correlation between the 2 tests was good, and NT–pro-BNP and BNP had only a moderate correlation with the extent of renal impairment. The only moderate correlation of the 2 markers with renal function remained consistent by either estimating GFR by the abbre-
1533
viated Modification of Diet in Renal Disease equation or measuring cystatin C. The 2 measures appear to have comparable accuracy for estimating renal function,24 but cystatin C may have particular efficacy in patients with mild to moderate renal impairment.25 This only moderate correlation with either measure of renal function makes it difficult to account for elevated NT–pro-BNP and BNP levels primarily on the basis of impaired renal clearance. Consistent with the finding that natriuretic peptide levels are in large part independent of renal function is the absence of a progressive trend in the prevalence of LVH or CAD by progressive stages of CKD. Although long-term outcome studies are still needed in the CKD population to determine if natriuretic peptides can specifically predict adverse cardiac events, the ability of NT–pro-BNP and BNP to identify CAD and LVH suggests that elevated levels may be a method for identifying CKD populations at the highest cardiovascular risk. The elevated levels of natriuretic peptides in renal failure could be due to a combination of real insufficiency and the presence of underlying cardiac disease. In such cases, renal insufficiency will increase the optimal level of a marker for the prediction of cardiac disease.26 In our study, the optimal value of BNP was 228 pg/ml as a predictor of CAD and 220 pg/ml as a predictor of LVH. Similarly, the optimal value of NT–pro-BNP was 979 pg/ml as a predictor of CAD and 762 pg/ml as a predictor of LVH. Our study was not designed to show a cause-and-effect relation between NT–pro-BNP and BNP and CAD or LVH in patients with CKD but was designed to examine the trends of CAD and LVH in concordance with these markers. Our patients with CKD demonstrated trends for a greater prevalence of CAD and LVH with increasing quartiles of NT–pro-BNP and BNP, whereas the correlation between either of these markers and GFR was only modest, implying that the elevated levels of NT–pro-BNP and BNP are in part secondary to CAD or LVH. The possible mechanisms of elevated NT–pro-BNP and BNP in CAD and LVH include residual ischemia from CAD, subendocardial ischemia from hypertension, and greater intracardiac pressures, which could be from CAD or LVH. However, on the basis of the small sample size, the present study does not recommend replacing noninvasive testing for cardiac ischemia or echocardiography for LVH assessment but suggests that the NT–pro-BNP and BNP could be useful for indicating trends of CAD and LVH in patients with CKD and that these 2 markers are comparable for such a purpose. 1. Culleton BF, Larson MG, Wilson PW, Evans JC, Parfrey PS, Levy D. Cardiovascular disease and mortality in a community-based cohort with mild renal insufficiency. Kidney Int 1999;56:2214 –2219. 2. Levin A, Singer J, Thompson CR, Ross H, Lewis M. Prevalent left ventricular hypertrophy in the predialysis population: identifying opportunities for intervention. Am J Kidney Dis 1996;27:347–354. 3. Shlipak MG, Fried LF, Cushman M, Manolio TA, Peterson D, Stehman-Breen C, Bleyer A, Newman A, Siscovick D, Psaty B. Cardio-
1534
4.
5.
6.
7.
8.
9.
10. 11.
12. 13.
14.
15.
16.
The American Journal of Cardiology (www.AJConline.org)
vascular mortality risk in chronic kidney disease: comparison of traditional and novel risk factors. JAMA 2005;293:1737–1745. Bibbins-Domingo K, Ansari M, Schiller NB, Massie B, Whooley MA. B-Type natriuretic peptide and ischemia in patients with stable coronary disease. Circulation 2003;108:2987–2992. Weber M, Dill T, Arnold R, Rau M, Ekinci O, Muller KD, Berkovitsch A, Mitrovic V, Hamm C. N-terminal B-type natriuretic peptide predicts the extent of coronary artery disease and ischemia in patients with stable angina pectoris. Am Heart J 2004;148:612– 620. Foote R, Pearlman J, Siegal A, Yeo KJ. Detection of exercise-induced ischemia by changes in B-type natriuretic peptides. J Am Coll Cardiol 2004;44:1980 –1987. Uusimaa P, Tokola H, Ylitalo A, Vuolteenaho O, Ruskoaho H, Risteli J, Linnaluoto M, Peuhkurinen K. Plasma B-type natriuretic peptide reflects left ventricular hypertrophy and diastolic function in hypertension. Int J Cardiol 2004;97:251–256. Luchner A, Hengstenberg C, Lowel H, Trawinski J, Baumann M, Riegger GA, Schunkert H, Holmer S. N-terminal pro-brain natriuretic peptide after myocardial infarction: a marker of cardio-renal function. Hypertension 2002;39:99 –104. Cataliotti A, Malatino LS, Jougasaki M, Zoccali C, Castellino P, Giacone G, Bellanuova I, Tripepi R, Seminara G, Parlongo S, et al. Circulating natriuretic peptide concentrations in patients with endstage renal disease: role of brain natriuretic peptide as a biomarker for ventricular remodeling. Mayo Clin Proc 2001;76:1111–1119. Munagala VK, Burnett JC Jr, Redfield MM. The natriuretic peptides in cardiovascular medicine. Curr Probl Cardiol 2004;29:707–769. Takami Y, Horio T, Iwashima Y, Takiuchi S, Kamide K, Yoshihara F, Nakamura S, Nakahama H, Inenaga T, Kangawa K, Kawano Y. Diagnostic and prognostic value of plasma brain natriuretic peptide in non-dialysis-dependent CRF. Am J Kidney Dis 2004;44:420 – 428. McCullough PA, Sandberg KR. Sorting out the evidence on natriuretic peptides. Rev Cardiovasc Med 2003;4(suppl):S13–S19. Lainchbury JG, Campbell E, Frampton CM, Yandle TG, Nicholls MG, Richards AM. Brain natriuretic peptide and n-terminal brain natriuretic peptide in the diagnosis of heart failure in patients with acute shortness of breath. J Am Coll Cardiol 2003;42:728 –735. Mueller T, Gegenhuber A, Poelz W, Haltmayer M. Diagnostic accuracy of B type natriuretic peptide and amino terminal proBNP in the emergency diagnosis of heart failure. Heart 2005;91:606 – 612. Levey AS, Greene T, Kusek JW. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 2000;11:155A. Devereux RB, Casale PN, Kligfield P, Eisenberg RR, Miller D, Campo E, Alonso DR. Performance of primary and derived M-mode echocardiographic measurements for detection of left ventricular hypertrophy in necropsied subjects and in patients with systemic hypertension,
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
mitral regurgitation and dilated cardiomyopathy. Am J Cardiol 1986; 57:1388 –1393. de Simone G, Daniels SR, Devereux RB, Meyer RA, Roman MJ, de Divitiis O, Alderman MH. Left ventricular mass and body size in normotensive children and adults: assessment of allometric relations and impact of overweight. J Am Coll Cardiol 1992;20:1251–1260. Vuille C, Weyman AE. Left ventricle I: general considerations. Assessment of chamber size and function. In: Weyman AE, ed. Principles and Practice of Echocardiography. Philadelphia, PA: Lea & Febiger, 1994:575– 620. Maisel AS, Krishnaswamy P, Nowak RM, McCord J, Hollander JE, Duc P, Omland T, Storrow AB, Abraham WT, Wu AH, et al. Rapid measurement of B-type natriuretic peptide in the emergency diagnosis of heart failure. N Engl J Med 2002;347:161–167. Dasgupta A, Chow L, Tso G, Nazareno L. Stability of NT-pro BNP in serum specimens collected in Becton Dickinson Vacutainer (SST) tubes. Clin Chem 2003;49:958 –960. Januzzi JL, Camargo CA, Anwaruddin S, Baggish AL, Chen AA, Krauser DG, Tung R, Cameron R, Nagurney JT, Chae CU, et al. The N-terminal Pro-BNP Investigation of Dyspnea in the Emergency Department (PRIDE) study. Am J Cardiol 2005;95:948 –954. Morrow DA, de Lemos JA, Sabatine MS, Murphy SA, Demopoulos LA, DiBattiste PM, McCabe CH, Gibson CM, Cannon CP, Braunwald E. Evaluation of B-type natriuretic peptide for risk assessment in unstable angina/non-ST-elevation myocardial infarction: B-type natriuretic peptide and prognosis in TACTICS-TIMI 18. J Am Coll Cardiol 2003;41:1264 –1272. James SK, Lindahl B, Siegbahn A, Stridsberg M, Venge P, Armstrong P, Barnathan ES, Califf R, Topol EJ, Simoons ML, Wallentin L. N-terminal pro-brain natriuretic peptide and other risk markers for the separate prediction of mortality and subsequent myocardial infarction in patients with unstable coronary artery disease: a Global Utilization of Strategies to Open Occluded Arteries (GUSTO)-IV substudy. Circulation 2003;108:275–281. Lamb EJ, Stowe HJ, Simpson DE, Coakley AJ, Newman DJ, Leahy M. Diagnostic accuracy of cystatin C as a marker of kidney disease in patients with multiple myeloma: calculated glomerular filtration rate formulas are equally useful. Clin Chem 2004;50:1848 –1851. Shlipak MG, Sarnak MJ, Katz R, Fried LF, Seliger SL, Newman AB, Siscovick DS, Stehman-Breen C. Cystatin C and the risk of death and cardiovascular events among elderly persons. N Engl J Med 2005;352: 2049 –2060. McCullough PA, Duc P, Omland T, McCord J, Nowak RM, Hollander JE, Herrmann HC, Steg PG, Westheim A, Knudsen CW, 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;41:571–579.
The prevalences of coronary artery disease (CAD) and left ventricular hypertrophy (LVH) are greater in patients with chronic kidney disease (CKD) than in the general population.1–3 Most of these patients remain asymptomatic despite marked increases in the prevalences of CAD and LVH. The detection of CAD in these patients has remained elusive, and the detection of LVH often requires echocardiography, which is expensive when the magnitude of the population is considered.3 N-terminal pro–B-type natriuretic peptide (NT–pro-BNP) and B-type natriuretic peptide (BNP) levels can predict the presence of CAD and LVH in the general population.4 – 8 However, uncertainty remains as to whether
elevated levels of 1 or both of these markers are predominantly a function of a reduced glomerular filtration rate (GFR) or maintains a predictive value for heart disease.9 –12 Because these 2 markers correlate well for predicting patients with congestive heart failure,13,14 we hypothesized that NT–pro-BNP and BNP levels would have similar abilities to identify CAD and LVH in asymptomatic patients with CKD. To test this hypothesis, we prospectively collected simultaneous clinical, laboratory, and echocardiographic data from ambulatory patients with CKD being examined at routine medical visits.
Divisions of aCardiology and bNephrology and Departments of Epidemiology and dPathology, University of Maryland School of Medicine, Baltimore, Maryland. Manuscript received August 26, 2005; revised manuscript received and accepted November 28, 2005. This study was funded by an investigator-initiated grant from Roche Diagnostics Corporation, Indianapolis, Indiana. * Corresponding author: Tel: 410-328-7204; fax: 410-328-3530. E-mail address: [email protected] (C.R. deFilippi).
Methods Study population and design: Enrollment into the study required that patients have serum creatinine of ⬎1.2 mg/dl, not be on dialysis, and be ⱖ18 years of age. The evaluation involved reviewing the patients’ cardiac histories and cardiac risk factors, taking blood samples, and performing transthoracic echocardiography. The protocol was ap-
c
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2005.11.090
www.AJConline.org
Miscellaneous/B-Type Natriuretic Peptides in Renal Disease
proved by the institutional review board. Informed consent was obtained from each participant. Fifty-four of 70 patients who met enrollment criteria and signed the informed consent form completed all testing. Data collection: Patients’ demographic and clinical information was obtained from electronic medical record reviews. History of diabetes, hypertension, peripheral vascular disease, smoking, current alcohol use, and CAD were included if they were identified in a patient’s history and/or physical examination results. CAD was defined by previous myocardial infarction or coronary revascularization. Smoking history was stratified as never having smoked or ever having smoked, currently or in the past. Height and weight data obtained during clinic visits were used to calculate body mass index. Values for total cholesterol and lowdensity lipoprotein cholesterol were collected from the clinical records. Hypercholesterolemia was defined as total cholesterol ⬎200 mg/dl or low-density lipoprotein cholesterol ⬎100 mg/dl. Creatinine values were obtained clinically for all patients and were used to estimate the GFR using the abbreviated Modification of Diet in Renal Disease study formula.15 For biomarker assays, plasma samples were collected and centrifuged at 1,000g for 12 minutes. The resulting serum and plasma were aliquoted, frozen, and maintained at ⫺70°C. NT–pro-BNP was measured from serum using an immunoassay (Elecsys proBNP, Roche Diagnostics Corporation, Indianapolis, Indiana). BNP was measured from plasma using an immunoassay (Triage BNP, Biosite Diagnostics, San Diego, California). A second assessment of renal function was made by measuring cystatin C from serum using a nephelometric immunoassay (BN Prospect N Latex Cystatin C, Dade Behring, Glasgow, Delaware). Transthoracic echocardiograms were obtained on the day of enrollment for the purposes of this study and not for clinical indications. All studies were performed by the same technician using a Phillips Sonos 5500 (Philips Medical Systems, Andover, Massachusetts) or SonoSite SonoHeart Elite (SonoSite, Inc., Bothell, Washington). Echocardiograms were analyzed to quantify the left ventricular ejection fraction and left ventricular mass. Echocardiographic interpretations were performed without knowledge of biomarker results. Left ventricular mass was determined using the formula of Devereux et al16 and was adjusted for patient height.17 LVH was determined after adjustment for patient height.17 The left ventricular ejection fraction was calculated using end-diastolic volume and end-systolic volume measurements from apical 4- and 2-chamber views calculated by Simpson’s method.18 Statistical analysis: Data from continuous variables are expressed as medians with interquartile ranges (IQRs). Categorical variables are expressed as absolute numbers and percentages. The Cochran-Armitage test for trend was used to analyze trends among BNP quartiles, NT–pro-BNP quartiles, and CKD stages with categorical clinical and echocar-
1531
Table 1 Patient characteristics (n ⫽ 54) Characteristic Age (yrs) Men African–American Diabetes mellitus Hypertension Smoking Hypercholesterolemia Coronary artery disease Left ventricular ejection fraction ⱕ40% LVH Diastolic dysfunction Body mass index (kg/m2) Glomerular filtration rate (ml/min/1.73 m2) NT–pro-BNP (pg/ml) BNP (pg/ml)
Value 70 (57.3–76), 66.5 ⫾ 13.8 53 (98%) 24 (44%) 28 (52%) 51 (94%) 45 (83%) 29 (53%) 14 (26%) 4 (7%) 30 (56%) 31 (57%) 28.3 (25.3–32.3), 29.7 ⫾ 6.7 31 (21–42), 34 ⫾ 21 728 (168–2,950), 2,552 ⫾ 4,143 138 (31–391) 271 ⫾ 326
Data are presented or number (percentages), mean ⫾ SD, and median (IQR).
diographic variables. The Kruskal-Wallis test is a nonparametric test that was used to evaluate the relations among NT–pro-BNP quartiles, BNP quartiles, and CKD stages with continuous clinical, laboratory, and echocardiographic variables. For each subject, GFR versus NT–pro-BNP level, GFR versus BNP level, 1/cystatin C versus NT–pro-BNP level, 1/cystatin C versus BNP level, and NT–pro-BNP level versus BNP level were plotted, and Pearson’s correlation coefficient was calculated for these relationss. Receiver-operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was calculated to determine and compare the ability of NT–pro-BNP and BNP levels to indicate LVH and CAD. The AUC of NT– pro-BNP was compared with that of BNP using DeLong, DeLong, and Clarke-Pearson’s method of the comparison of 2 ROC curves derived from correlated samples. Two-sided p values ⬍0.05 were considered statistically significant. Analyses were performed using SAS version 8.2 (SAS Institute Inc., Cary, North Carolina) and AccuROC version 2.5 (Accumetric Corporation, Montreal, Quebec, Canada).
Results Of 54 study patients, 53 were men (98%), and 24 were African-Americans (44%). The median age was 70 years (IQR 57 to 76). Fourteen patients (26%) had CAD, and 30 (56%) had LVH. Only 4 patients (7%) had left ventricular ejection fractions ⱕ40%. The median NT–pro-BNP level was 724 pg/ml (IQR 168 to 2,950), and the median BNP level was 137 pg/ml (IQR 31 to 391). The median GFR was 31 ml/min/1.73 m2 (IQR 21 to 42) (Table 1). A significant positive correlation was found between NT–pro-BNP and BNP levels (R ⫽ 0.74, p ⬍0.0001).
1532
The American Journal of Cardiology (www.AJConline.org)
Table 2 Trends of CAD and LVH stratified by glomerular filtration rate, BNP, and NT–pro-BNP Variable
GFR Stages 1 and 2 (⬎90–60 ml/min/1.73 m2) (n ⫽ 6)
Stage 3 (30–59 ml/min/1.73 m2) (n ⫽ 26)
Stage 4 (15–29 ml/min/1.73 m2) (n ⫽ 11)
Stage 5 (⬍15 ml/min/1.73 m2) (n ⫽ 11)
p Value
1 (17%) 4 (67%) 239 (115–271) 82 (41–107)
6 (23%) 12 (46%) 315 (90–782) 45 (18–228)
4 (36%) 7 (64%) 1414 (226–5,788) 200 (60–388)
3 (27%) 7 (64%) 4376 (1,070–13,798) 423 (310–634)
0.50 0.56 0.0003 0.001
CAD LVH NT–pro-BNP, median (IQR) BNPm median (IQR)
NT–pro-BNP
CAD LVH
Quartile 1 (16–144 pg/ml) (n ⫽ 13)
Quartile 2 (160–685 pg/ml) (n ⫽ 14)
Quartile 3 (762–2,273 pg/ml) (n ⫽ 13)
Quartile 4 (3,175–1,7493 pg/ml) (n ⫽ 14)
p Value
0 (0%) 4 (31%)
1 (7%) 7 (50%)
7 (54%) 7 (54%)
6 (43%) 12 (86%)
0.001 0.005
BNP
CAD LVH
Quartile 1 (4–31 pg/ml) (n ⫽ 14)
Quartile 2 (32–114 pg/ml) (n ⫽ 13)
Quartile 3 (161–388 pg/ml) (n ⫽ 13)
Quartile 4 (392–1,520 pg/ml) (n ⫽ 14)
p Value
0 (0%) 4 (29%)
1 (8%) 8 (62%)
6 (46%) 6 (46%)
7 (50%) 12 (86%)
0.0004 0.008
Figure 1. ROC curves of NT–pro-BNP (AUC 0.80, p ⫽ 0.001) and BNP (AUC 0.82, p ⫽ 0.0004) as indicators of CAD (p ⫽ 0.45, for comparison between ROC curves).
Moderate correlations were found between NT–pro-BNP and GFR (R ⫽ ⫺0.45, p ⫽ 0.0006) and between BNP and GFR (R ⫽ ⫺0.38, p ⫽ 0.005). Using the reciprocal of cystatin C as an alternative method to estimate renal function, the correlations between NT–pro-BNP and 1/cystatin C (R ⫽ ⫺0.45, p ⫽ 0.0006) and BNP and 1/cystatin C (R
⫽ ⫺0.35, p ⫽ 0.01) remained moderate and were similar to the correlations determined by the estimated GFR. When patients were stratified by CKD stages, the NT– pro-BNP level increased progressively from stage 1 to 2 to stage 5 CKD (p ⫽ 0.0003). Similarly, BNP levels also increased from stage 1 to 2 to stage 5 of the GFR, with a positive trend (p ⫽ 0.001). However, there was no trend for an increase in the prevalence of LVH or CAD by increasing CKD stage. In contrast, when patients were stratified across progressive NT–pro-BNP quartiles, the prevalence of LVH and CAD increased (p ⫽ 0.005 and p ⫽ 0.001, respectively). Similarly, when patients were stratified across progressive BNP quartiles, the prevalence of LVH and CAD increased (p ⫽ 0.008 and p ⫽ 0.0004, respectively) (Table 2). The ROC curves for NT–pro-BNP (AUC 0.80, p ⫽ 0.001) and BNP (AUC 0.82, p ⫽ 0.0004) as indicators of CAD are shown in Figure 1. The ROC curves of NT–proBNP (AUC ⫽ 0.72, p ⫽ 0.005) and BNP (AUC ⫽ 0.72, p ⫽ 0.007) as predictors of LVH are shown in Figure 2. A comparison between the ROC curves for NT–pro-BNP and BNP for indicating CAD demonstrated that the AUCs for the 2 tests were not different (p ⫽ 0.45, for the AUC comparison). Similarly, a comparison between the ROC curves of NT–pro-BNP and BNP for identifying LVH demonstrated that the AUCs were not different between the 2 tests (p ⫽ 0.78, for the AUC comparison). On the basis of the ROC analysis, the optimal value of NT–pro-BNP as an indicator of CAD was 979 pg/ml, with a sensitivity of 79%, a specificity of 70%, a positive predictive value (PPV) of 48%, and a negative predictive value (NPV) of 90%. The optimal value of BNP to indicate CAD
Miscellaneous/B-Type Natriuretic Peptides in Renal Disease
Figure 2. ROC curve of NT–pro-BNP (AUC 0.72, p ⫽ 0.005) and BNP (AUC 0.72, p ⫽ 0.007) as indicators of LVH (p ⫽ 0.78, for comparison between ROC curves).
was 228 pg/ml, with a sensitivity of 86%, a specificity of 73%, a PPV of 52%, and a NPV of 94%. The optimal value of NT–pro-BNP as a predictor of LVH was 762 pg/ml, with a sensitivity of 63%, a specificity of 67%, a PPV of 70%, and a NPV of 57%. The optimal value of BNP as a predictor of LVH was 200 pg/ml, with a sensitivity of 60%, a specificity of 71%, a PPV of 72%, and a NPV of 59%.
Discussion The present study demonstrates that in ambulatory patients with CKD not requiring dialysis, NT–pro-BNP and BNP can identify, with similar accuracy, patients with LVH and CAD. Serum levels of NT–pro-BNP and BNP generally increase with worsening renal impairment.8,11,12 Furthermore, it has been hypothesized that NT–pro-BNP may be more dependent on renal clearance than BNP,12 limiting the diagnostic accuracy of the test in this setting. Our results do indicate a greater increase in NT–pro-BNP levels compared with BNP levels from the established reference ranges.19,20 However, such a finding is also common in a variety of disease states, such as congestive heart failure13,14,21 and acute coronary syndromes22,23 in patients with normal renal function. To further support the equivalency of the response of NT–pro-BNP and BNP in the CKD setting, independent of their predictive value, the correlation between the 2 tests was good, and NT–pro-BNP and BNP had only a moderate correlation with the extent of renal impairment. The only moderate correlation of the 2 markers with renal function remained consistent by either estimating GFR by the abbre-
1533
viated Modification of Diet in Renal Disease equation or measuring cystatin C. The 2 measures appear to have comparable accuracy for estimating renal function,24 but cystatin C may have particular efficacy in patients with mild to moderate renal impairment.25 This only moderate correlation with either measure of renal function makes it difficult to account for elevated NT–pro-BNP and BNP levels primarily on the basis of impaired renal clearance. Consistent with the finding that natriuretic peptide levels are in large part independent of renal function is the absence of a progressive trend in the prevalence of LVH or CAD by progressive stages of CKD. Although long-term outcome studies are still needed in the CKD population to determine if natriuretic peptides can specifically predict adverse cardiac events, the ability of NT–pro-BNP and BNP to identify CAD and LVH suggests that elevated levels may be a method for identifying CKD populations at the highest cardiovascular risk. The elevated levels of natriuretic peptides in renal failure could be due to a combination of real insufficiency and the presence of underlying cardiac disease. In such cases, renal insufficiency will increase the optimal level of a marker for the prediction of cardiac disease.26 In our study, the optimal value of BNP was 228 pg/ml as a predictor of CAD and 220 pg/ml as a predictor of LVH. Similarly, the optimal value of NT–pro-BNP was 979 pg/ml as a predictor of CAD and 762 pg/ml as a predictor of LVH. Our study was not designed to show a cause-and-effect relation between NT–pro-BNP and BNP and CAD or LVH in patients with CKD but was designed to examine the trends of CAD and LVH in concordance with these markers. Our patients with CKD demonstrated trends for a greater prevalence of CAD and LVH with increasing quartiles of NT–pro-BNP and BNP, whereas the correlation between either of these markers and GFR was only modest, implying that the elevated levels of NT–pro-BNP and BNP are in part secondary to CAD or LVH. The possible mechanisms of elevated NT–pro-BNP and BNP in CAD and LVH include residual ischemia from CAD, subendocardial ischemia from hypertension, and greater intracardiac pressures, which could be from CAD or LVH. However, on the basis of the small sample size, the present study does not recommend replacing noninvasive testing for cardiac ischemia or echocardiography for LVH assessment but suggests that the NT–pro-BNP and BNP could be useful for indicating trends of CAD and LVH in patients with CKD and that these 2 markers are comparable for such a purpose. 1. Culleton BF, Larson MG, Wilson PW, Evans JC, Parfrey PS, Levy D. Cardiovascular disease and mortality in a community-based cohort with mild renal insufficiency. Kidney Int 1999;56:2214 –2219. 2. Levin A, Singer J, Thompson CR, Ross H, Lewis M. Prevalent left ventricular hypertrophy in the predialysis population: identifying opportunities for intervention. Am J Kidney Dis 1996;27:347–354. 3. Shlipak MG, Fried LF, Cushman M, Manolio TA, Peterson D, Stehman-Breen C, Bleyer A, Newman A, Siscovick D, Psaty B. Cardio-
1534
4.
5.
6.
7.
8.
9.
10. 11.
12. 13.
14.
15.
16.
The American Journal of Cardiology (www.AJConline.org)
vascular mortality risk in chronic kidney disease: comparison of traditional and novel risk factors. JAMA 2005;293:1737–1745. Bibbins-Domingo K, Ansari M, Schiller NB, Massie B, Whooley MA. B-Type natriuretic peptide and ischemia in patients with stable coronary disease. Circulation 2003;108:2987–2992. Weber M, Dill T, Arnold R, Rau M, Ekinci O, Muller KD, Berkovitsch A, Mitrovic V, Hamm C. N-terminal B-type natriuretic peptide predicts the extent of coronary artery disease and ischemia in patients with stable angina pectoris. Am Heart J 2004;148:612– 620. Foote R, Pearlman J, Siegal A, Yeo KJ. Detection of exercise-induced ischemia by changes in B-type natriuretic peptides. J Am Coll Cardiol 2004;44:1980 –1987. Uusimaa P, Tokola H, Ylitalo A, Vuolteenaho O, Ruskoaho H, Risteli J, Linnaluoto M, Peuhkurinen K. Plasma B-type natriuretic peptide reflects left ventricular hypertrophy and diastolic function in hypertension. Int J Cardiol 2004;97:251–256. Luchner A, Hengstenberg C, Lowel H, Trawinski J, Baumann M, Riegger GA, Schunkert H, Holmer S. N-terminal pro-brain natriuretic peptide after myocardial infarction: a marker of cardio-renal function. Hypertension 2002;39:99 –104. Cataliotti A, Malatino LS, Jougasaki M, Zoccali C, Castellino P, Giacone G, Bellanuova I, Tripepi R, Seminara G, Parlongo S, et al. Circulating natriuretic peptide concentrations in patients with endstage renal disease: role of brain natriuretic peptide as a biomarker for ventricular remodeling. Mayo Clin Proc 2001;76:1111–1119. Munagala VK, Burnett JC Jr, Redfield MM. The natriuretic peptides in cardiovascular medicine. Curr Probl Cardiol 2004;29:707–769. Takami Y, Horio T, Iwashima Y, Takiuchi S, Kamide K, Yoshihara F, Nakamura S, Nakahama H, Inenaga T, Kangawa K, Kawano Y. Diagnostic and prognostic value of plasma brain natriuretic peptide in non-dialysis-dependent CRF. Am J Kidney Dis 2004;44:420 – 428. McCullough PA, Sandberg KR. Sorting out the evidence on natriuretic peptides. Rev Cardiovasc Med 2003;4(suppl):S13–S19. Lainchbury JG, Campbell E, Frampton CM, Yandle TG, Nicholls MG, Richards AM. Brain natriuretic peptide and n-terminal brain natriuretic peptide in the diagnosis of heart failure in patients with acute shortness of breath. J Am Coll Cardiol 2003;42:728 –735. Mueller T, Gegenhuber A, Poelz W, Haltmayer M. Diagnostic accuracy of B type natriuretic peptide and amino terminal proBNP in the emergency diagnosis of heart failure. Heart 2005;91:606 – 612. Levey AS, Greene T, Kusek JW. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 2000;11:155A. Devereux RB, Casale PN, Kligfield P, Eisenberg RR, Miller D, Campo E, Alonso DR. Performance of primary and derived M-mode echocardiographic measurements for detection of left ventricular hypertrophy in necropsied subjects and in patients with systemic hypertension,
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
mitral regurgitation and dilated cardiomyopathy. Am J Cardiol 1986; 57:1388 –1393. de Simone G, Daniels SR, Devereux RB, Meyer RA, Roman MJ, de Divitiis O, Alderman MH. Left ventricular mass and body size in normotensive children and adults: assessment of allometric relations and impact of overweight. J Am Coll Cardiol 1992;20:1251–1260. Vuille C, Weyman AE. Left ventricle I: general considerations. Assessment of chamber size and function. In: Weyman AE, ed. Principles and Practice of Echocardiography. Philadelphia, PA: Lea & Febiger, 1994:575– 620. Maisel AS, Krishnaswamy P, Nowak RM, McCord J, Hollander JE, Duc P, Omland T, Storrow AB, Abraham WT, Wu AH, et al. Rapid measurement of B-type natriuretic peptide in the emergency diagnosis of heart failure. N Engl J Med 2002;347:161–167. Dasgupta A, Chow L, Tso G, Nazareno L. Stability of NT-pro BNP in serum specimens collected in Becton Dickinson Vacutainer (SST) tubes. Clin Chem 2003;49:958 –960. Januzzi JL, Camargo CA, Anwaruddin S, Baggish AL, Chen AA, Krauser DG, Tung R, Cameron R, Nagurney JT, Chae CU, et al. The N-terminal Pro-BNP Investigation of Dyspnea in the Emergency Department (PRIDE) study. Am J Cardiol 2005;95:948 –954. Morrow DA, de Lemos JA, Sabatine MS, Murphy SA, Demopoulos LA, DiBattiste PM, McCabe CH, Gibson CM, Cannon CP, Braunwald E. Evaluation of B-type natriuretic peptide for risk assessment in unstable angina/non-ST-elevation myocardial infarction: B-type natriuretic peptide and prognosis in TACTICS-TIMI 18. J Am Coll Cardiol 2003;41:1264 –1272. James SK, Lindahl B, Siegbahn A, Stridsberg M, Venge P, Armstrong P, Barnathan ES, Califf R, Topol EJ, Simoons ML, Wallentin L. N-terminal pro-brain natriuretic peptide and other risk markers for the separate prediction of mortality and subsequent myocardial infarction in patients with unstable coronary artery disease: a Global Utilization of Strategies to Open Occluded Arteries (GUSTO)-IV substudy. Circulation 2003;108:275–281. Lamb EJ, Stowe HJ, Simpson DE, Coakley AJ, Newman DJ, Leahy M. Diagnostic accuracy of cystatin C as a marker of kidney disease in patients with multiple myeloma: calculated glomerular filtration rate formulas are equally useful. Clin Chem 2004;50:1848 –1851. Shlipak MG, Sarnak MJ, Katz R, Fried LF, Seliger SL, Newman AB, Siscovick DS, Stehman-Breen C. Cystatin C and the risk of death and cardiovascular events among elderly persons. N Engl J Med 2005;352: 2049 –2060. McCullough PA, Duc P, Omland T, McCord J, Nowak RM, Hollander JE, Herrmann HC, Steg PG, Westheim A, Knudsen CW, 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;41:571–579.