Assays Specific for BNP1-32 and NT-proBNP Exhibit a Similar Performance to Two Widely Used Assays in the Diagnosis of Heart Failure.

BACKGROUND: Secretion of cardioprotective B-type natriuretic peptide 1-32 (BNP1-32) is increased proportionately with cardiac dysfunction, but its measurement in plasma is difficult. Therefore, less specific BNP and amino-terminal proBNP (NT-proBNP) assays that detect the precursor molecule proBNP alongside BNP or NT-proBNP metabolites were developed to reflect BNP1-32 secretion and are now mandated in the diagnosis of heart failure (HF). We compared the diagnostic performance of 2 widely used clinical assays: the Roche proBNPII assay, and Abbott BNP assay, against our recently developed in-house assays that measure either intact BNP1-32 or NT-proBNP. METHODS: EDTA plasma samples obtained from patients presenting with breathlessness (n = 195, 60 [31%] with clinically adjudicated HF) were assayed using the Roche NT-proBNP and our specific in-house BNP1-32 and NTBNP assays. A subset (n = 75) were also assessed with the Abbott BNP assay. RESULTS: Roche NT-proBNP was highly correlated with BNP1-32 and NTBNP (Spearman rho = 0.92 and 0.90, respectively, both Ps < 0.001), and all 3 assays similarly discriminated acute HF from other causes of breathlessness (ROC analysis areas under the curve 0.85-0.89). The Abbott BNP assay performed similarly to the other assays. Roche NT-proBNP and BNP1-32 assays had similar sensitivity (83% and 80%), specificity (83% and 84%), positive (70% and 71%) and negative (91% and 90%) predictive values, and accuracy (both 83%) at their optimal cutoffs of 1536 and 12 ng/L, respectively. CONCLUSIONS: Since all assays exhibited similar performance in the diagnosis of HF, currently mandated assays provide a reliable proxy for circulating concentrations of active BNP1-32 in HF diagnosis.

ProBNP That Is Not Glycosylated at Threonine 71 Is Decreased with Obesity in Patients with Heart Failure.

BACKGROUND: Heart failure (HF) is a leading cause of morbidity and mortality worldwide. Plasma concentrations of B-type natriuretic peptide (BNP) or its amino terminal congener (NT-proBNP) are used for HF diagnosis and risk stratification. Because BNP concentrations are inexplicably lowered in obese patients, we investigated the relationship between proBNP glycosylation, plasma NT-proBNP, and body mass index (BMI) in HF patients. METHODS: Three assays were developed to distinguish between total proBNP (glycosylated plus nonglycosylated proBNP), proBNP not glycosylated at threonine 71 (NG-T71), and proBNP not glycosylated in the central region (NG-C). Intraassay and interassay CVs were <15%; limits of detection were <21 ng/L; and samples diluted in parallel. RESULT: Applying these assays and an NT-proBNP assay to plasma samples from 106 healthy volunteers and 238 HF patients determined that concentrations [median (interquartile range)] of proBNP, NG-T71, and NT-proBNP were greater in HF patients compared with controls [300 (44-664), 114 (18-254), and 179 (880-3459) ng/L vs 36 (18-229), 36 (18-175), and 40 (17-68) ng/L, respectively; all P < 0.012]. NG-C was undetectable in most samples. ProBNP concentrations in HF patients with BMI more or less than 30 kg/m2 were not different (P = 0.85), whereas HF patients with BMI >30 kg/m2 had lower NT-proBNP and NG-T71 concentrations (P < 0.003) and higher proBNP/NT-proBNP and proBNP/NG-T71 ratios (P = 0.001 and P = 0.02, respectively) than those with BMI <30 kg/m2. CONCLUSIONS: Increased BMI is associated with decreased concentrations of proBNP not glycosylated at T71. Decreased proBNP substrate amenable to processing could partially explain the lower NT-proBNP and BNP concentrations observed in obese individuals, including those presenting with HF.

Development of a BNP1-32 Immunoassay That Does Not Cross-React with proBNP.

BACKGROUND: Plasma B-type natriuretic peptide (BNP) concentration reflects cardiac dysfunction and assists in determining the diagnosis and prognosis of heart failure (HF). Current BNP assays overestimate circulating bioactive BNP1-32 concentrations as they also detect less active BNP metabolites and proBNP. A specific BNP1-32 assay with negligible cross-reactivity to proBNP and/or BNP metabolites may be advantageous. METHODS: We developed a Luminex-based specific BNP1-32 immunoassay and compared results obtained from 3 other BNP assays (a Luminex-based total-BNP assay, our BNP RIA, and the commercially available Abbott Architect BNP assay) in plasma from 42 patients with HF and 22 healthy controls. RESULTS: The BNP1-32 assay showed 57% cross-reactivity with BNP2-32, but ≤0.1% cross-reactivity to BNP3-32, other BNP metabolites, and proBNP; its detection limit was 0.35 ng/L; and intra- and interassay CVs were <15%. BNP immunoreactivity increased with HF severity (median concentrations being 0.3, 0.8, 26.2, and 17.3 ng/L in healthy controls and 40.7, 139, 465, and 1778 ng/L in HF patients for the BNP1-32, total-BNP, BNP RIA, and Abbott BNP assays respectively). The fold increase between HF cases with the New York Heart Association (NYHA) class IV was significantly greater with the BNP1-32 assay than the Abbott BNP (P = 0.026) and the BNP RIA (P < 0.0001) but not the total-BNP assay. CONCLUSIONS: We have developed the first assay that measures BNP1-32 in plasma without interference by proBNP. Analysis of larger patient cohorts is now required to compare the performance of this assay with current less specific assays for the diagnosis or prognosis of HF.

Comparison of immunoassays for NTproBNP conducted on three analysis systems: Milliplex, Elecsys and RIA.

OBJECTIVES: The aim of this study was to evaluate a Milliplex NTproBNP immunoassay and to compare its performance with the Roche Elecsys assay and an in house NTproBNP radioimmunoassay (RIA). DESIGN AND METHODS: Samples were analyzed for NTproBNP in all 3 assays following the manufacturer's instructions and/or our previously published methodologies. RESULTS: Similar plasma concentrations and assay reproducibility were observed for the Roche assay and in house RIA. However, the Milliplex assay produced much lower concentrations than the other two assays (versus Roche Elecsys bias of -367.9 pg/mL, (LOA -907.6 and 171.8)), and only quantified results for 63 of the 111 samples assessed. Of these 46 samples had acceptable errors of measurement (within sample coefficients of variation <20%) equating to 41% of the total samples measured. CONCLUSIONS: The Milliplex assay, in its current form, does not produce results comparable with the Roche Elecsys 2010 assay or our in house RIA, and does not perform well as a quantitative assay. It may provide qualitative results (ie high, medium or low concentration), and serve as a rudimentary screening assay.

Generation and characterization of a mouse model of the metabolic syndrome: apolipoprotein E and aromatase double knockout mice.

The aim of this study was to create a comprehensive mouse model of the metabolic syndrome by crossing aromatase-deficient (ArKO) mice with apolipoprotein E-deficient (ApoE(-/-)) mice. Successive crossbreeding of ArKO with ApoE(-/-)-deficient mice generated double knockout, MetS-Tg mice. The phenotypic characteristics of the MetS-Tg mice were assessed at 3, 6, and 12 mo of age and compared with age- and sex-matched wild-type (WT) controls. Blood pressure and heart rate were recorded by a noninvasive, computerized tail-cuff system. Oral glucose and intraperitoneal insulin tolerance tests were performed. Serum cholesterol levels were measured by a combined quantitative colorimetric assay. Plasma adiponectin, C-reactive protein (CRP), insulin, interleukin-6 (IL-6), leptin, resistin, and tumor necrosis factor-α (TNF-α) were measured by multiplexed ELISA. MetS-Tg mice displayed significantly increased body weight, central obesity, and elevated blood pressure at all three ages compared with WT mice. Elevated serum cholesterol was associated with higher triglycerides and LDL/VLDL cholesterol particles and was accompanied by a decrease in HDL and histological evidence of fatty liver. MetS-Tg mice of all ages showed impaired glucose tolerance. At 12 mo, MetS-Tg mice had elevated plasma levels of CRP, IL-6, leptin, and TNF-α, but resistin levels were largely unchanged. We now report that this combination of gene knockouts produces a novel strain of mice that display the diverse clinical features of the metabolic syndrome, including central obesity, progressive hypertension, an adverse serum lipid profile, fatty liver, glucose intolerance, insulin resistance, and evidence of an inflammatory state.

Prandial regulation of ghrelin secretion in humans: does glucagon contribute to the preprandial increase in circulating ghrelin?

OBJECTIVE: Glucagon secretion is stimulated by fasting and inhibited postprandially, a pattern that mimics the secretory profiles of both ghrelin and GH. We thus hypothesized that glucagon may be a determinant of the changes in circulating ghrelin and GH that occur in relation to meals. The objective of the study was to explore this hypothesis by determining the ghrelin and GH response to a bolus of glucagon or saline in healthy subjects. SUBJECTS AND MEASUREMENTS: Nine healthy volunteers, mean age 47 years (range 33-58) and body mass index (BMI) 24 kg/m2 (range 20.9-27.6) were recruited and received either 1 mg glucagon (n = 9) or 1 ml saline (n = 6) subcutaneously on separate days between 0800 and 0830 h after an overnight fast. Venous blood was then sampled at 15-min intervals during the first hour, followed by 30-min intervals up to 4 h for glucose, insulin, GH, cortisol, somatostatin and ghrelin. RESULTS: Mean +/- SE basal ghrelin was 213.1 +/- 34.3 pmol/l and decreased significantly by 15 min after glucagon administration to 179.3 +/- 28 pmol/l (P = 0.01), then remaining suppressed relative to the basal value until 240 min after glucagon. Plasma insulin increased from a basal value of 46.7 +/- 7.7 pmol/l to a peak of 327.1 +/- 54.9 pmol/l (P < 0.0001). There was an inverse statistical relationship between the increase in insulin over the first 120 min and the decrease in ghrelin (P = 0.005), while somatostatin, GH and glucose were not significant contributors to the decrease in ghrelin (P > 0.05). Mean +/- SE basal GH was 7.3 +/- 2.9 microg/l and increased by 150 min after glucagon to a peak of 20.5 +/- 6.8 microg/l (P = 0.006). Changes in neither ghrelin nor glucose were related to the increase in GH (P = 0.7). Saline administration did not produce any significant change in ghrelin, insulin or somatostatin although the expected diurnal reduction in cortisol (P < 0.05) was observed. CONCLUSIONS: Our study found no evidence that glucagon stimulates ghrelin secretion in humans and supports the hypothesis that insulin is a negative regulator of ghrelin secretion in the postprandial state. We did not find a negative relationship between endogenous somatostatin and ghrelin despite earlier reports that exogenously administered somatostatin analogues suppress plasma ghrelin. Finally, glucagon-induced GH secretion is not mediated by an increase in plasma ghrelin.

Plasma cardiotrophin-1 is elevated in human hypertension and stimulated by ventricular stretch.

OBJECTIVE: Cardiotrophin-1 (CT-1) is an interleukin-6-related cytokine with known hypertrophic and protective actions upon cardiac myocytes. We provide here the first report of cardiac tissue and plasma levels of CT-1 in human and experimental hypertension, demonstrate cardiac CT-1 secretion stimulated by ventricular stretch, and characterise molecular forms of CT-1 in tissue and plasma. METHODS: CT-1 levels in human and rat plasma and in rat cardiac tissue extracts were determined by specific radioimmunoassay (RIA). Cardiac CT-1 secretion during ventricular stretch was studied in isolated, perfused hearts. Molecular forms of CT-1 were identified using RIA coupled with high performance liquid chromatography (HPLC). Results are given as mean+/-SEM. RESULTS: Plasma levels of CT-1 in patients with untreated hypertension (UTH, 606+/-18 pmol/L, n=24) were significantly higher than those in age-and BMI-matched normotensive volunteers (NT, 546+/-12 pmol/L, n=31, P<0.01 vs. UTH). CT-1 levels in matched patients with treated hypertension (THT, 618+/-10 pmol/L, n=35) were similar to those in UTH patients, but higher than in NT controls (P<0.01). Plasma CT-1 demonstrated a weak but significant correlation with systolic blood pressure in all patients (r=0.241, P<0.05, n=90). In contrast, CT-1 levels in male, 40-week-old, NT-WKY rats (1295+/-98 pmol/L) were significantly higher than those in matched UTH-SHR (937+/-31 pmol/L, P<0.01). In both WKY and SHR rats, atrial tissue concentrations of CT-1 were 8-fold higher than ventricular levels. Left ventricular tissue CT-1 protein concentrations were significantly higher in 40-week-old SHR compared with age-matched WKY (SHR 12.6+/-0.5 fmol/g vs. WKY 9.5+/-0.8 fmol/g, P<0.01). Ventricular stretch of Langendorff perfused, isolated WKY/SHR hearts resulted in significant, acute release of CT-1 and BNP. HPLC coupled with specific RIA revealed CT-1 in human/rat plasma, isolated rat heart perfusate, and rat heart tissue extracts to consist of complex, high molecular weight forms. CONCLUSIONS: This is the first report to show increased levels of plasma CT-1 in hypertensive disease. CT-1 is a unique cardiac cytokine whose release is stimulated by ventricular stretch. The atrium contains the highest levels of the protein. The stored and circulating molecular form of CT-1 is complex, which may modulate its in vivo role in cardiovascular disease.