| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Proteomics and Protein Markers |
Departments of1
Laboratory Medicine and 2
Internal Medicine, Konventhospital Barmherzige Brueder, Linz, Austria.
3 Institute for Applied System Sciences and Statistics, University of Linz, Linz, Austria.
aAddress correspondence to this author at: Department of Laboratory Medicine, Konventhospital Barmherzige Brueder, Seilerstaette 2, A-4021 Linz, Austria. Fax 43-732-7897-3799; e-mail thomas.mueller{at}bblinz.at.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We studied a consecutive series of 149 asymptomatic patients referred for echocardiographic evaluation of the cardiac effects of systemic arterial hypertension. Diagnosis of cardiac structural disease was based on the presence of systolic or diastolic dysfunction, left atrial dilatation, left ventricular dilatation or hypertrophy, pulmonary hypertension, and wall motion or valvular abnormalities. Blood concentrations of BNP and NT-proBNP were measured by 2 commercially available assays (Abbott AxSYM and Roche Elecsys, respectively). Diagnostic accuracies of BNP and NT-proBNP were assessed by ROC curve analysis. Areas under the curves were compared by analysis of equivalency.
Results: In distinguishing between hypertensive patients with cardiac structural disease (n = 118) and hypertensive patients without (n = 31), areas under the curves were 0.740 (95% confidence interval, 0.662–0.808) for BNP and 0.762 (0.685–0.828) for NT-proBNP and were significantly equivalent (P = 0.015). Cutoff values with a 90% sensitivity for cardiac structural disease were 17 ng/L for BNP and 39 ng/L for NT-proBNP, with 29% and 32% specificity, respectively.
Conclusions: BNP and NT-proBNP have similar capabilities for detecting cardiac structural disease in asymptomatic patients with systemic arterial hypertension. However, in the setting evaluated, a screening strategy relying on measurement of BNP or NT-proBNP may be of limited value because of the low specificity at the selected cutoff values.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The measurement of BNP and NT-proBNP concentrations in plasma may aid decision making in a variety of clinical settings. The diagnostic value of circulating BNP and NT-proBNP is well established in symptomatic patients with suspected heart failure, but their utility as a screening tool for cardiac structural disease in the general population appears to be limited (1)(2)(3). However, in populations at high risk for the development of symptomatic heart failure, the role of the 2 analytes for the detection of asymptomatic cardiac dysfunction remains to be clarified.
In the American College of Cardiology/American Heart Association guidelines for the evaluation and management of chronic heart failure in the adult, it has been suggested that 4 stages be considered in the evolution of heart failure (heart failure stages A through D) (4). Typically, in asymptomatic patients at high risk for developing heart failure (e.g., patients with systemic arterial hypertension, diabetes mellitus, or coronary artery disease), echocardiography is used to detect patients progressing from stage A to stage B along the heart failure pathway, which is important because these patients require more intensive therapy (4). Thus, a screening test (such as plasma BNP or NT-proBNP) to be performed before echocardiography would be clinically useful for ruling out cardiac structural disease in this high-risk population and thereby reducing the frequency of echocardiography. Consequently, if plasma BNP and/or NT-proBNP cutoff concentrations, set at a high sensitivity, were associated with a clinically useful specificity, both markers could be used as a screening test to select patients for further assessment by echocardiography. Given the relatively high incidence of structural heart disease (e.g., diastolic dysfunction and left ventricular dilatation or hypertrophy) in patients with systemic arterial hypertension, the proposed application of BNP and NT-proBNP as a screening test could be clinically useful. The aims of the present study, therefore, were to prospectively evaluate the diagnostic utility of plasma BNP and NT-proBNP measurements for the detection of cardiac structural disease in asymptomatic patients with systemic arterial hypertension and to test the hypothesis that the 2 analytes are equally useful in this clinical setting.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Hypertensive patients without known cardiac structural disease were included if echocardiography was performed satisfactorily, as detailed below, and if blood samples for the measurement of plasma BNP and NT-proBNP concentrations and serum creatinine were collected on the same day. The estimated glomerular filtration rate was calculated as recommended recently (5). Patients with acute coronary syndromes (cardiac troponin positive or negative), patients with actual or previous symptomatic heart failure, and critically ill patients were excluded from the study. Other exclusion criteria included indications other than arterial hypertension for echocardiography, such as chest pain, palpitations, any symptoms related to heart failure (i.e., dyspnea on exertion, orthopnea, paroxysmal nocturnal dyspnea, nocturnal cough, jugular venous distension, pulmonary rales, third heart sound, and peripheral edema), prosthetic valves, pericardial disease, pulmonary embolism, acute stroke or actual transient or temporary stroke, and congenital heart disease.
Before echocardiography, each patient enrolled in the study was thoroughly evaluated by use of a history, physical examination, 12-lead electrocardiogram, and chest x-ray. The participants’ height and weight were routinely measured. The body mass index was obtained from the ratio of weight to height squared. Resting systolic and diastolic blood pressure and the patients’ actual medications were recorded on the same day the echocardiography was performed. Heart rate was determined during echocardiography. Echocardiography was performed without knowledge of the BNP and NT-proBNP plasma concentrations.
echocardiography
M-Mode and 2-dimensional images and spectral and color-flow Doppler recordings were obtained with a single commercially available instrument operating at 2.0–3.5 MHz. Two-dimensional imaging examinations were performed in the standard fashion in parasternal long- and short-axis views and apical 4- and 2-chamber views. All measurements were performed as recommended (6).
Two-dimensional echocardiograms were subjected to careful visual analyses to detect regional contractile abnormalities. Left ventricular systolic and diastolic volumes and ejection fraction were derived from biplane apical (2- and 4-chamber) views with the modified Simpson’s rule algorithm. Left ventricular dimensions were measured from M-mode images by the leading-edge technique, which included interventricular septal thickness at end diastole, posterior wall thickness at end diastole, and left ventricular internal dimension at end diastole. The greatest thickness measured at any site in the left ventricular wall was considered to represent maximal left ventricular wall thickness. Diastolic dysfunction was determined according to a previous report (7). For measurement of right ventricular systolic pressure, we used continuous-wave Doppler recording of tricuspid regurgitation according to the modified Bernoulli equation (four times the square peak velocity of the tricuspid jet plus adding to the right atrial pressure 5, 10, or 15 mmHg in relation to the clinical aspect, jugular venous pulse, and x-ray of the chest).
For the assessment of aortic stenosis, the pressure gradient across the aortic valve was estimated by the simplified Bernoulli equation from flow velocity detection by continuous-wave Doppler integrated throughout systole. For the evaluation of mitral stenosis, the transmitral gradient at rest during diastole was assessed by the pressure half-time using continuous-wave Doppler. Color-flow Doppler of the aortic valve from the parasternal long- and short-axis views was used to detect and assess the severity of aortic regurgitation based on the ratio of the width of regurgitant jet to the width of the regurgitant jet in the outflow tract. Color-flow Doppler was also applied for the detection of mitral regurgitation, and severity was estimated by the regurgitant jet area expressed as percentage of the left atrial area.
definitions
Arterial hypertension was defined as the use of any antihypertensive medication, resting systolic blood pressure 
140 mmHg, or resting diastolic blood pressure 90 mmHg. Diabetes mellitus was defined as the use of any glucose-lowering medication or as fasting blood glucose concentration 7 mmol/L (126 mg/dL). Renal dysfunction was defined as an estimated glomerular filtration rate <60 mL · min–1 · (1.73 m2)–1. Coronary artery disease was defined as remote myocardial infarction by history, occult myocardial infarction by electrocardiography, and previous coronary bypass surgery or percutaneous transluminal coronary angioplasty. The diagnosis of atrial fibrillation was defined as first episode, paroxysmal, persistent, or permanent.
Systolic dysfunction was defined as a left ventricular ejection fraction <50%. Diastolic dysfunction was defined as impaired relaxation or a restrictive or pseudonormal pattern. Left atrial dilatation was defined as left atrial diameter >40 mm. Left ventricular dilatation was defined as left ventricular end diastolic diameter >56 mm. Left ventricular hypertrophy was defined as septal thickness at end diastole or posterior wall thickness at end diastole >11 mm. Pulmonary hypertension was defined as right ventricular systolic pressure 35 mmHg. Valvular abnormality was defined as stenosis or regurgitation grade 2 or higher for the mitral or aortic valve.
A study participant was considered to have cardiac structural disease if at least one of the following echocardiographic findings was present: systolic or diastolic dysfunction, left atrial dilatation, left ventricular dilatation, left ventricular hypertrophy, pulmonary hypertension, wall motion abnormalities, or valvular abnormalities.
BNP and NT-PROBNP measurements
Blood for measurement of natriuretic peptide concentrations was collected by venipuncture in Vacuette polyethylene terephthalate glycol EDTA tubes (Greiner Bio-One) on the day of the echocardiographic evaluation. Blood samples were centrifuged at 3500g for 10 min at 4 °C immediately after collection. Both BNP and NT-proBNP were assayed within 4 h after blood withdrawal by commercially available assays. Plasma BNP was assayed on an AxSYM analyzer using the AxSYM BNP assay (Abbott Laboratories), and plasma NT-proBNP was measured by the Elecsys proBNP assay on an Elecsys 2010 instrument (Roche Diagnostics).
The precision of the 2 methods was evaluated according to the Clinical and Laboratory Standards Institute (CLSI, formerly NCCLS) guideline EP5-A (8). Three pooled patient plasma samples were aliquoted into 40 tubes (1.5 mL) for each concentration and frozen at –70 °C. We analyzed these samples in duplicate in 2 runs every day for 20 days on the 2 analyzers. Total imprecision was calculated by the CLSI double-run precision evaluation test (8). The precision data for the 2 methods were as follows: the AxSYM BNP assay had a total CV of 8.1% at a mean concentration of 108 ng/L (pool 1), a total CV of 7.5% at a mean concentration of 524 ng/L (pool 2), and a total CV of 10% at a mean concentration of 2117 ng/L (pool 3). The Elecsys NT-proBNP assay had a total CV of 3.8% at a mean concentration of 246 ng/L (pool 1), a total CV of 4.7% at a mean concentration of 891 ng/L (pool 2), and a total CV of 2.2% at a mean concentration of 10666 ng/L (pool 3).
statistical analysis
Data were analyzed statistically with SPSS 13.0 software (SPSS Inc.), the MedCalc 8.0.0.0 package (MedCalc Software), and the software N package (IDV). All probabilities were 2-tailed, and P values <0.05 were regarded as significant. Sample size calculation was based on the following assumptions: AUC for BNP 
0.750 with an estimated SE of 0.045 and a correlation coefficient of 0.800 for BNP and NT-proBNP plasma concentrations. Thus, testing for equivalency, our calculation showed that a sample size of 120 patients had to be enrolled into this study if the limit of equivalency was set at ±15% difference of the areas under the curves (AUCs) for the 2 analytes (
= 0.050, 2-sided; ß = 0.100).
Using the nonparametric Mann–Whitney U-test or Fisher exact test, as appropriate, we compared the univariate data for the demographic and clinical features between the patients with cardiac structural disease caused by arterial hypertension and hypertensive patients without cardiac structural disease; the P values were not adjusted for multiple comparisons and are therefore only descriptive. We used the Spearman coefficient of rank correlation to assess the relationship between BNP and NT-proBNP concentrations in the study population. To determine the diagnostic accuracy of plasma BNP and NT-proBNP for cardiac structural disease, we analyzed the ROC curves and calculated AUCs for both analytes. According to our study hypothesis, AUCs were compared by analysis of equivalency. Cutoff concentrations for BNP and NT-proBNP were determined at the 90% sensitivity criterion derived directly from the ROC curves.
To determine whether BNP and NT-proBNP plasma concentrations were independent predictors for cardiac structural disease and to calculate multivariate odds ratios, we performed 2 logistic regression analyses without variable selection (all relevant variables were included simultaneously into the models). Dichotomous risk factors were coded with an indicator variable of 1 for having the condition and 0 for its absence; an incremental approach was used for continuous variables (i.e., the calculated odds ratios are related to the increase of a certain continuous variable by the given amount).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
50%, concomitant atrial fibrillation during echocardiography prohibited exact classification of diastolic dysfunction; therefore, the final number studied was 149. Classification of the study participants showed that 118 of the hypertensive patients had cardiac structural disease and 31 did not (disease prevalence, 79%). The demographic and clinical characteristics of the study participants according to this classification are listed in Table 1
. The median BNP and NT-proBNP concentrations for the patients with cardiac structural disease were 2-fold and 3-fold higher, respectively, than for the patients without cardiac structural disease.
|
In distinguishing between hypertensive patients with cardiac structural disease (n = 118) and hypertensive patients without cardiac structural disease (n = 31), the AUCs were 0.740 (SE = 0.045; 95% confidence interval, 0.662–0.808) for BNP and 0.762 (SE = 0.043; 95% confidence interval, 0.685–0.828) for NT-proBNP (Fig. 1
). The Spearman coefficient of rank correlation for BNP and NT-proBNP plasma concentrations was 0.841 (95% confidence interval, 0.787–0.883; P <0.001). Equivalence testing was based on a ±15% range as detailed in the Materials and Methods. Accordingly, given the standard AUC of 0.740, the confidence limits were 0.629 and 0.851. Because the 97% confidence interval for the AUC of NT-proBNP was 0.671–0.851, both AUCs were significantly equivalent (P = 0.015).
|
The cutoff values with a 90% sensitivity (95% confidence interval for both BNP and NT-proBNP, 83%–95%) for cardiac structural disease were 17 ng/L for BNP [specificity, 29% (14%–48%)] and 39 ng/L for NT-proBNP [specificity, 32% (17%–51%)]. On the basis of a disease prevalence of 79%, the positive and negative predictive values at these cutoff concentrations were 83% and 43% for BNP and 83% and 45% for NT-proBNP, respectively, and the diagnostic accuracy was 77% for BNP and 78% for NT-proBNP. When we used the respective cutoff concentrations, classification by both BNP and NT-proBNP was correct in 106 and incorrect in 24 patients, whereas we obtained discordant false classifications for 10 cases by the BNP assay and 9 cases by the NT-proBNP assay.
As shown in Table 2
, BNP and NT-proBNP plasma concentrations were predictors of cardiac structural disease independent of possible confounding variables. In the first multivariate model, BNP was an independent and significant predictor (P = 0.005) of cardiac structural disease with an odds ratio of 2.91 (95% confidence interval, 1.37–6.14) for an increment of 50 ng/L. The second statistical model revealed that NT-proBNP was also independently related to cardiac structural disease (P = 0.004), displaying an odds ratio of 2.57 (95% confidence interval, 1.35–4.89) for an increment of 100 ng/L. All other variables included failed to reach statistical significance.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-fold and 3-fold, respectively, higher than for the patients without cardiac structural disease and despite the fact that BNP and NT-proBNP were the only significant predictors of cardiac structural disease in multivariate analysis, a screening strategy relying on measurement of plasma BNP or NT-proBNP to select patients with arterial hypertension for further echocardiography is of limited value, in our opinion, in the investigated setting because of the low specificity of this proposed strategy. In general, many studies have shown a correlation of BNP and/or NT-proBNP with several indicators of left ventricular and right ventricular function as assessed by echocardiography (2)(3)(9). This, of course, should also hold true for hypertensive patients. Indeed, published data indicate that increased plasma BNP (10)(11)(12)(13)(14)(15)(16)(17) and NT-proBNP (10)(18) concentrations are associated with left ventricular systolic and diastolic dysfunction, left ventricular dilatation, left ventricular hypertrophy, pulmonary hypertension, and significant valve disease in asymptomatic or mildly symptomatic patients with arterial hypertension. However, most of these studies focused on either 1 or 2 of the above-mentioned indicators of cardiac impairment. Of them, only the study by Bhalla et al. (10), retrospective in design, used a more complex approach measuring both plasma BNP and NT-proBNP concentrations and demonstrating that these markers might be useful to identify the presence of left ventricular dysfunction (determined by the presence of systolic or diastolic dysfunction, valvular or wall motion abnormalities, or left ventricular hypertrophy). In fact, such an approach is convincing because BNP and NT-proBNP are considered indicators of increased intracardiac pressure, irrespective of whether the increased intracardiac pressure is caused by left ventricular hypertrophy, left ventricular systolic dysfunction, valve disease, or even fast atrial fibrillation (19). In our study, we used an approach for classifying the study participants as having cardiac structural disease if at least one of the following echocardiographic findings was present: systolic or diastolic dysfunction, left atrial dilatation, left ventricular dilatation, left ventricular hypertrophy, pulmonary hypertension, wall motion abnormalities, or valvular abnormalities (see Materials and Methods), and, as detailed, we found equal diagnostic accuracies for BNP and NT-proBNP for differentiating between patients with and without asymptomatic cardiac structural disease.
Because most patients in our study were on oral antihypertensive medications, it is important to emphasize that drugs, including diuretics, angiotensin-converting enzyme inhibitors, and adrenergic agonists, can modify circulating concentrations of natriuretic peptides (1)(2)(3). A recent study demonstrated that BNP concentrations for treated hypertensive patients were not statistically significantly different from those for an age-matched control population (20). Our study design, however, with inclusion of asymptomatic patients referred for echocardiographic evaluation of the cardiac effects of systemic arterial hypertension, represents a real-world scenario; thus, it is conceivable that this population is often on therapy before the evaluation of structural heart disease by echocardiography.
Our study has several potential limitations. It is a single-center study performed in a relatively small sample of hypertensive patients and may not accurately represent the general demographics of patients referred for echocardiographic evaluation of the cardiac effects of systemic arterial hypertension. Because the diagnostic accuracies of BNP and NT-proBNP assays are considered to depend on the disease prevalence in the population studied and on the reference standard used for classification of study participants (21), the relatively high prevalence of structural heart disease in our population (79%) compared with other populations might be an important issue. Although we believe that the high disease prevalence is related to both the tertiary care setting and the consideration of diastolic dysfunction for the reference standard, the reproduction of our results in other centers or by multicenter studies would argue for their validity. Furthermore, we must acknowledge that the optimal cutoff BNP and NT-proBNP concentrations assessed in our study, and probably the diagnostic accuracies as well, may have been influenced by the assay choice, because authors of some recent reports have suggested that the diagnostic information provided by BNP and NT-proBNP measurements might be method-dependent (21)(22)(23). This may specifically hold true for the comparison of older generation assays (e.g., IRMAs and enzyme immunoassays) with fully automated newer generation assays. However, in the present evaluation, plasma BNP and NT-proBNP concentrations were determined by newer generation assays that have been proposed for widespread use in clinical practice.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. Emdin, C. Passino, C. Prontera, M. Fontana, R. Poletti, A. Gabutti, C. Mammini, A. Giannoni, L. Zyw, G. Zucchelli, et al. Comparison of Brain Natriuretic Peptide (BNP) and Amino-Terminal ProBNP for Early Diagnosis of Heart Failure Clin. Chem., July 1, 2007; 53(7): 1289 - 1297. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Clerico, M. Fontana, L. Zyw, C. Passino, and M. Emdin Comparison of the Diagnostic Accuracy of Brain Natriuretic Peptide (BNP) and the N-Terminal Part of the Propeptide of BNP Immunoassays in Chronic and Acute Heart Failure: A Systematic Review Clin. Chem., May 1, 2007; 53(5): 813 - 822. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Herrmann, O. Taban-Shoma, U. Hubner, A. Pexa, H. Kilter, N. Umanskaya, R. Hans Straub, M. Bohm, and W. Herrmann Hyperhomocysteinemia and Myocardial Expression of Brain Natriuretic Peptide in Rats Clin. Chem., April 1, 2007; 53(4): 773 - 780. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Herrmann, S. Muller, I. Kindermann, L. Gunther, J. Konig, M. Bohm, and W. Herrmann Plasma B vitamins and their relation to the severity of chronic heart failure Am. J. Clinical Nutrition, January 1, 2007; 85(1): 117 - 123. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. E. Malavazos, L. Morricone, A. Marocchi, F. Ermetici, B. Ambrosi, and M. M. Corsi N-terminal pro-B-type natriuretic Peptide and echocardiographic abnormalities in severely obese patients: correlation with visceral fat. Clin. Chem., June 1, 2006; 52(6): 1211 - 1213. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
