Midregional pro-A-type natriuretic peptide measurements for diagnosis of acute destabilized heart failure in short-of-breath patients: comparison with B-type natriuretic peptide (BNP) and amino-terminal proBNP.
The role of BNP and NT-proBNP measurements as an aid in the emergency diagnosis of acute destabilized HF is well established in patients with dyspnea (5-8). Because there are currently no published data on whether ANP or NT-proANP may also be useful for diagnostic purposes in this setting, we aimed at assessing the utility of determining circulating NT-proANP concentrations, using a novel sandwich immunoassay covering midregional epitopes (MR-proANP) in comparison with BNP and NT-proBNP measurements for the diagnosis of acute destabilized HF in short-of-breath patients presenting to an emergency department.
Materials and Methods
The present evaluation is a post hoc analysis of a previously described study cohort (8). In brief, 251 consecutive patients consulting the emergency department of the St. John of God Hospital (Linz, Austria) with dyspnea as a chief complaint were prospectively studied to compare head to head the diagnostic accuracies of BNP and NT-proBNP for acute destabilized HF in an emergency setting. The diagnosis of acute destabilized HF was based on the Framingham score for HF (9) plus echocardiographic evidence of systolic or diastolic dysfunction. During the initial patient examination in our emergency department, blood samples were collected for the measurement of plasma BNP and NT-proBNP concentrations and were analyzed within the next 4 h by two commercially available assays (Abbott AxSYM and Roche Elecsys, respectively). Aliquots of these EDTA-plasma samples were stored at -80[degrees]C for further analyses. These samples were used for the determination of plasma MR-proANP concentrations by an immunoluminometric assay for midregional NT-proANP (B.R.A.H.M.S SERISTRA; B.R.A.H.M.S AG) (10). All 251 plasma samples were measured in one batch approximately 1 year after collection in a blinded fashion to the clinical features and the biochemical data. The precision of the 3 methods has been evaluated and described previously (8, 10). It has been recommended that both BNP and NT-proBNP measurements should be reported in nanograms per liter (11). However, because circulating NT-proANP concentrations are usually reported in picomoles per liter in the literature, we decided to do so for MR-proANP in the present report as well.
Median MR-proANP, BNP, and NT-proBNP plasma concentrations in patients with dyspnea caused by acute destabilized HF and patients with dyspnea attributable to other reasons were compared by the nonparametric Mann-Whitney U-test. To determine the diagnostic accuracy of MR-proANP in comparison with BNP and NTproBNP for acute destabilized HF, ROC plots were analyzed, and areas under the curve (AUCs) were calculated for all 3 analytes. AUCs were compared according to the method of Hanley and McNeil (respective P values were not adjusted for multiple comparisons and are therefore only descriptive) (12). Cutoff concentrations for MR-proANP, BNP, and NT-proBNP were determined according to the 90% and 95% sensitivity criteria derived directly from the ROC curves. Furthermore, cutoff concentrations with the highest diagnostic accuracy (i.e., optimal cutoff concentrations, defined as the points on the ROC curves at which the sum of the false-negative and false-positive results was lowest) were evaluated for all 3 analytes. Positive and negative predictive values at these cutoff concentrations were calculated by use of the ratio of cases in the positive and negative groups (reflecting the prevalence of the disease in our study population). Discordant false biochemical classifications for MR-proANP vs BNP and NT-proBNP, respectively, at optimal cutoff concentrations were compared with the McNemar test. The Spearman coefficient of rank correlation was used to assess the relationship of MR-proANP with BNP and NT-proBNP concentrations in the study population. Logistic regression analyses were performed to determine whether MR-proANP, BNP, and NT-proBNP plasma concentrations above the optimal cutoff concentrations were predictors for acute destabilized HF (i.e., to calculate odds ratios; unadjusted and controlling for confounding covariates). Dichotomous variables were coded with an indicator variable of 1 for having the condition and 0 for its absence. Statistical analyses were performed with SPSS 13.0 software (SPSS Inc.), the MedCalc 18.104.22.168 package (MedCalc Software), and software N (IDV). All probabilities were two-tailed, and P values <0.05 were regarded as significant.
Of the 251 patients with dyspnea as a chief complaint, 137 patients were classified as having dyspnea attributable to acute destabilized HF, and 114 patients were classified as having dyspnea attributable to other reasons. Comprehensive information on the demographic and clinical characteristics of the study participants is given in the previous publication on this study cohort (8). Median plasma MR-proANP, BNP, and NT-proBNP concentrations were significantly higher in the patients with dyspnea attributable to acute destabilized HF than in the patients with dyspnea attributable to other reasons: MRproANP, 338 pmol/L (interquartile range, 234-521 pmol/L) vs 98 pmol/L (interquartile range, 54-166 pmol/L; P <0.001); BNP, 792 ng/L (interquartile range, 363-1714 ng/L) vs 65 ng/L (interquartile range, 10-196 ng/L; P <0.001); NT-proBNP, 3275 ng/L (interquartile range, 1398-6979 ng/L) vs 248 ng/L (interquartile range, 77-729 ng/L; P <0.001).
[FIGURE 1 OMITTED]
In distinguishing between patients with dyspnea caused by acute destabilized HF (n = 137) and patients with dyspnea attributable to other causes (n = 114), the AUCs were 0.876 [SE = 0.022; 95% confidence interval (CI), 0.829-0.914] for MR-proANP, 0.916 (SE = 0.018; 95% CI, 0.874-0.947) for BNP, and 0.903 (SE = 0.019; 95% CI, 0.859-0.939) for NT-proBNP (Fig. 1). Comparison of the ROC curves revealed no significant difference between the AUCs for MR-proANP and NT-proBNP (difference between AUCs, 0.027; SE = 0.017; 95% CI, -0.007 to 0.061; P = 0.123), and between the AUCs for BNP and NTproBNP (difference between AUCs, 0.013; SE = 0.012; 95% CI, -0.011 to 0.037; P = 0.277). Power calculations showed that the power of these 2 analyses was 88% and 94%, respectively. In contrast, there was a statistically significant difference between the AUCs for MR-proANP and BNP (difference between AUCs, 0.040; SE = 0.017; 95% CI, 0.006-0.073; P = 0.020). The complete information, including the appropriate decision statistics, for the biochemical diagnosis of acute destabilized HF in short-ofbreath patients are listed in Table 1.
When we used the cutoff concentrations with the highest diagnostic accuracy according to Table 1, classification by both BNP and MR-proANP was correct in 185 and incorrect in 19 patients. When we compared the 24 misclassifications by the BNP assay with the 23 by the MR-proANP assay with the McNemar test, the difference was not significant (P >0.999). Accordingly, classification by both NT-proBNP and MR-proANP was correct in 192 and incorrect in 23 patients at the cutoff concentrations with the highest diagnostic accuracy. When we compared the 17 misclassifications by the NT-proBNP assay with the 19 by the MR-proANP assay, the difference was also not significant (P = 0.868).
When we assessed the relationship of MR-proANP with BNP and NT-proBNP values in the entire study population (n = 251), nonparametric correlation analysis of MR-proANP vs BNP revealed a correlation coefficient ([r.sub.s]) of 0.835 (95% CI, 0.794-0.869; P <0.001); accordingly, the nonparametric correlation coefficient ([r.sub.s]) for MR-proANP vs NT-proBNP was 0.832 (95% CI, 0.789-0.866; P <0.001). Because the cusum test for linearity showed a significant deviation from linearity (P <0.010) for MR-proANP vs BNP and for MR-proANP vs NT-proBNP, regression analyses were not performed.
We calculated the univariate odds ratios for the detection of acute destabilized HF with MR-proANP, BNP, or NT-proBNP as independent variables dichotomized according to optimal cutoff concentrations. In addition, we determined MR-proANP, BNP, and NT-proBNP odds ratios for acute destabilized HF adjusted for age, sex, and estimated glomerular filtration rate (eGFR; adjusted regression model 1) to estimate the influence of these potential confounders. We also included a history of acute destabilized HF as well as clinical variables (signs and symptoms of HF) in another analysis (adjusted regression model 2) to determine the impact of these covariates on the predictive value of the 3 analytes for acute destabilized HF. The results of these analyses are shown in Table 2. When we performed a logistic regression with acute destabilized HF as the dependent variable and age, sex, eGFR, history of acute destabilized HF, and the presence of orthopnea, paroxysmal nocturnal dyspnea, nocturnal cough, jugular venous distension, pulmonary rales, third heart sound, and peripheral edema as independent variables (excluding MR-proANP, BNP, and NT-proBNP), we obtained a diagnostic accuracy of 79% for this statistical model.
The main finding of the present post hoc analysis is that MR-proANP measurements in plasma may be useful for ruling out acute destabilized HF in short-of-breath patients presenting to the emergency department. Furthermore, in this setting, the diagnostic value of MR-proANP is comparable to those of BNP and NT-proBNP, both of which are established markers for the emergency diagnosis of acute destabilized HF in patients with dyspnea (5-8). In this context it should be mentioned that BNP and NT-proBNP have been investigated for prognostic purposes and also as candidate guides in the monitoring of management of HF (2), whereas there is no such body of investigation at present for MR-proANP.
Although the present evaluation revealed that the AUC for MR-proANP was somewhat smaller than the AUCs for BNP (statistically significant) and NT-proBNP (not statistically significant), this issue is probably not clinically relevant. Interpretation of the respective ROC curves in Fig. 1 indicated that the smaller AUC for MR-proANP is the result of a lower specificity for MRproANP compared with BNP and NT-proBNP for sensitivities <80%. However, because the measurement of these analytes is considered useful for ruling out acute destabilized HF in short-of-breath patients, the clinically relevant range covers a sensitivity of 80%-100%. For sensitivities >80%, the ROC curves for MR-proANP, BNP, and NT-proBNP are very similar, and accordingly, as detailed in Table 1, our evaluation showed comparable sensitivities, specificities, and diagnostic accuracies at selected cutoff concentrations for all 3 markers. In addition, comparison of discordant false classifications on the basis of optimal cutoff values for MR-proANP vs BNP and NT-proBNP, respectively, underlined the similar diagnostic utility of the analytes. Thus, our findings indicate that MR-proANP, BNP, and NT-proBNP may be equally useful as an aid for the diagnosis of acute destabilized HF in patients consulting an emergency department with shortness of breath as a chief complaint. Of course, different cutoff concentrations must be considered for the 3 analytes, as Table 1 shows.
Of note, the proportional difference in mean concentrations of the peptides in persons with acute destabilized HF compared with those without was ~3-fold for the MR-proANP assay in contrast to BNP and NT-proBNP, for which the differences were 12- and 13-fold, respectively. Nevertheless, this did not markedly alter the diagnostic test performance for MR-proANP. As shown by logistic regression analysis, the odds ratios of MR-proANP for the prediction of acute destabilized HF did not change when we calculated the model unadjusted for potential confounders and the model controlling for age, sex, and renal function (eGFR). In addition, the overall accuracy of the 2 logistic regression models for MR-proANP was equal (i.e., 83%), indicating that age, sex, and eGFR did not add any relevant diagnostic information in the setting evaluated. If we included all clinical information available in the emergency department to the above logistic regression model (i.e., history of acute destabilized HF and the presence of orthopnea, paroxysmal nocturnal dyspnea, nocturnal cough, jugular venous distension, pulmonary rales, third heart sound, and peripheral edema), the diagnostic accuracy of the whole model was increased to 87%. Conversely, the odds ratios for MR-proANP decreased as expected, indicating the predictive value of the clinical signs and symptoms of HF. MR-proANP alone (similar to BNP and NT-proBNP) appeared to have a greater predictive value for acute destabilized HF in an emergency department (accuracy, 83%) than did taking together all of the clinical information (accuracy, 79%).
The major limitation of the present study is that this was a post hoc evaluation of the diagnostic capability of MR-proANP in an emergency setting. Therefore, future prospectively planned and adequately powered investigations should focus on the question of whether MRproANP can be used in place of BNP and/or NT-proBNP for ruling out acute destabilized HF in patients with dyspnea or whether there might be additional value for a strategy based on a combination of these markers. Furthermore, as there are currently no commercially available assays for the rapid measurement of MR-proANP (e.g., total duration of assay <20 min), the proposed application of MR-proANP for the emergency diagnosis of acute destabilized HF is hindered by the design of the assay used in the present study (i.e., sandwich immunoassay with a total assay time of ~3 h). This issue might reduce the practicability of MR-proANP measurements in an emergency setting, but the findings in the present evaluation could initiate the development of a fully automated assay for rapid MR-proANP measurements, making them suitable as an aid for routine decision-making in an emergency department.
Received December 17, 2005; accepted February 27, 2006.
Previously published online at DOI: 10.1373/clinchem.2005.065441
(1.) Clerico A, Del Ry S, Giannessi D. Measurement of cardiac natriuretic hormones (atrial natriuretic peptide, brain natriuretic peptide, and related peptide) in clinical practice: the need for a new generation of immunoassay methods [Review]. Clin Chem 2000;46:1529-34.
(2.) Cowie MR, Jourdain P, Maisel A, Dahlstrom U, Follath F, Isnard R, et al. Clinical applications of B-type natriuretic peptide (BNP) testing [Review]. Eur Heart J 2003;24:1710-8.
(3.) Mair J. Role of cardiac natriuretic peptide testing in heart failure [Editorial]. Clin Chem 2002;48:977-8.
(4.) Ala-Kopsala M, Magga J, Peuhkurinen K, Leipala J, Ruskoaho H, Leppaluoto J, et al. Molecular heterogeneity has a major impact on the measurement of circulating N-terminal fragments of A- and B-type natriuretic peptides. Clin Chem 2004;50:1576-88.
(5.) Maisel AS, Krishnaswamy P, Nowak RM, McCord J, Hollander JE, Duc P, et al. Rapid measurement of B-type natriuretic peptide in the emergency diagnosis of heart failure. N Engl J Med 2002;347: 161-7.
(6.) Januzzi JL Jr, Camargo CA, Anwaruddin S, Baggish AL, Chen AA, Krauser DG, et al. The N-terminal pro-BNP investigation of dyspnea in the emergency department (PRIDE) study. Am J Cardiol 2005; 95:948-54.
(7.) 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;42:728-35.
(8.) 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;91:606-12.
(9.) Senni M, Tribouilloy CM, Rodeheffer RJ, Jacobsen SJ, Evans JM, Bailey KR, et al. Congestive heart failure in the community: a study of all incident cases in Olmsted County, Minnesota, in 1991. Circulation 1998;98:2282-9.
(10.) Morgenthaler NG, Struck J, Thomas B, Bergmann A. Immunoluminometric assay for the midregion of pro-atrial natriuretic peptide in human plasma [Technical Brief]. Clin Chem 2004;50:234-6.
(11.) Apple FS, Panteghini M, Ravkilde J, Mair J, Wu AH, Tate J, et al. Quality specifications for B-type natriuretic peptide assays. Clin Chem 2005;51:486-93.
(12.) Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 1983;148:839-43.
ALFONS GEGENHUBER,  JOACHIM STRUCK,  WERNER POELZ,  RICHARD PACHER,  NILS G. MORGENTHALER,  ANDREAS BERGMANN,  MEINHARD HALTMAYER, [5,6] and THOMAS MUELLER  *
Departments of  Internal Medicine and  Laboratory Medicine, Konventhospital Barmherzige Brueder, Linz, Austria.
 Research Department, B.R.A.H.M.S AG, Hennigsdorf/Berlin, Germany.
 Institute for Applied System Sciences and Statistics, University of Linz, Linz, Austria.
 Department of Cardiology, Medical University of Vienna, Vienna, Austria.
 Paracelsus Private Medical University, Salzburg, Austria.
 Nonstandard abbreviations: HF, heart failure; ANP, A-type (atrial) natriuretic peptide; BNP, B-type natriuretic peptide; NT-proANP and NT-proBNP, amino-terminal fragment of the ANP and BNP prohormone, respectively; MR-proANP, midregion of the ANP prohormone; AUC, area under the curve; CI, confidence interval; and eGFR, estimated glomerular filtration rate.
* Address correspondence to this author at: Department of Laboratory Medicine, Konventhospital Barmherzige Brueder, Seilerstaette 2-4, A-4020 Linz, Austria. Fax 43-732-7677-3799; e-mail firstname.lastname@example.org.
Table 1. Diagnostic information for the biochemical diagnosis of acute destabilized heart failure by MR-proANP, BNP, and NT-proBNP in short-of-breath patients presenting to the emergency department. Cutoff Sensitivity Specificity PPV Diagnostic concentration (95% CI), % (95% CI), % (a) NPV accuracy, % MR-proANP 109 pmol/L 95 (90-98) 56 (47-65) 0.72 0.90 77 147 pmol/L 90 (84-95) 68 (59-77) 0.77 0.86 80 169 pmol/L 89 (83-94) 76 (67-84) 0.82 0.85 83 (b) BNP 118 ng/L 95 (90-98) 64 (55-73) 0.76 0.91 81 160 ng/L 90 (84-95) 73 (64-81) 0.80 0.87 83 295 ng/L 80 (73-87) 86 (78-92) 0.87 0.78 83 (b) NT-proBNP 292 ng/L 95 (90-98) 53 (43-62) 0.71 0.90 76 476 ng/L 90 (84-95) 65 (55-74) 0.76 0.85 79 825 ng/L 87 (80-92) 81 (72-88) 0.84 0.84 84 (b) (a) PPV, positive predictive value; NPV, negative predictive value. (b) Optimal cutoff concentration, defined as the concentration corresponding to the highest diagnostic accuracy (i.e., the point on the ROC curve at which the sum of the false-negative and false-positive results was lowest). Table 2. Results of logistic regression analyses for prediction of acute destabilized HF. (a) Cutoff Odds Independent concen- ratio DACC, Model variable tration (95% CI) P (b) % Unadjusted MR-proANP 169 pmol/L 26 (13-52) <0.001 83 model BNP 295 ng/L 25 (13-49) <0.001 83 NT-proBNP 825 ng/L 28 (14-55) <0.001 84 Adjusted MR-proANP 169 pmol/L 34 (15-76) <0.001 83 model 1 (c) BNP 295 ng/L 27 (13-56) <0.001 83 NT-proBNP 825 ng/L 29 (14-61) <0.001 84 Adjusted MR-proANP 169 pmol/L 23 (8-63) <0.001 88 model 2 (d) BNP 295 ng/L 10 (4-24) <0.001 87 NT-proBNP 825 ng/L 13 (6-32) <0.001 87 Incorrect classification False- False- Model positive, n negative, n Unadjusted 27 15 model 16 27 22 18 Adjusted 27 15 model 1 (c) 16 27 22 18 Adjusted 17 13 model 2 (d) 16 18 19 15 (a) Odds ratios are related to the 3 analytes. Diagnostic accuracy and false-positive and false-negative results of classification are derived from the respective whole statistical model. Observed frequencies: patients with dyspnea caused by acute destabilized HF, n = 137; patients with dyspnea attributable to other causes, n = 114. (b) DACC, diagnostic accuracy of the statistical model including all given covariates simultaneously. (c) Controlling for age, sex, and eGFR. (d) Controlling for age, sex, eGFR, history of acute destabilized HF, and the presence of orthopnea, paroxysmal nocturnal dyspnea, nocturnal cough, jugular venous distension, pulmonary rales, third heart sound, and peripheral edema.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Proteomics and Protein Markers|
|Author:||Gegenhuber, Alfons; Struck, Joachim; Poelz, Werner; Pacher, Richard; Morgenthaler, Nils G.; Bergmann|
|Article Type:||Clinical report|
|Date:||May 1, 2006|
|Previous Article:||High maternal blood S100B concentrations in pregnancies complicated by intrauterine growth restriction and intraventricular hemorrhage.|
|Next Article:||The antibody configurations of cardiac troponin I assays may determine their clinical performance.|