Summary
Background
B-type natriuretic peptide (BNP) and left atrial volume index (LAVi) are used as surrogate measures for global myocardial function and are recommended for the diagnosis of heart failure with normal ejection fraction. Little is known, however, about predictors in patients with preserved systolic function.
Aims
To identify factors that influence the relation of BNP and left atrial size to invasively determined left ventricular end-diastolic pressure in stable patients with preserved left ventricular systolic function.
Methods
Fifty-nine consecutive patients were included prospectively. Clinical, biological, Doppler echocardiographic and invasive variables were collected simultaneously.
Results
BNP was predicted independently by left ventricular ejection fraction, diastolic function and age ( p < 0.05). LAVi was predicted independently by left ventricular mass index and invasive left ventricular end-diastolic pressure ( p < 0.01). BNP predicted increased left ventricular end-diastolic pressure greater than 16 mmHg ( p = 0.004); the optimal cut-off value was 33 pg/mL (area under the receiver-operating characteristic curve [AUC] 0.74 [0.6–0.84], p < 0.001, sensitivity 72%, specificity 70%). LAVi predicted increased left ventricular end-diastolic pressure ( p < 0.001); the optimal cut-off value for LAVi was 26 mL/m 2 (AUC 0.87 [0.75–0.94], p < 0.001; sensitivity 85%, specificity 80%). Unlike BNP ( p = 0.1), LAVi performed well in patients with abnormal relaxation at mitral filling ( p < 0.01).
Conclusion
BNP is influenced by age in stable patients with preserved systolic function and should be interpreted cautiously. LAVi is a powerful surrogate for invasively determined left ventricular end-diastolic pressure regardless of age and mitral filling.
Résumé
Contexte
Le peptide natriurétique de type B (BNP) et le volume de l’oreillette gauche reflètent la fonction myocardique globale et sont recommandés dans le diagnostic de l’insuffisance cardiaque diastolique, mais leurs facteurs prédictifs sont peu connus en cas de fonction systolique préservée.
Objectifs
Identifier les facteurs influençant la corrélation de ces deux paramètres avec la pression télédiastolique ventriculaire gauche chez les patients stables avec fonction systolique préservée.
Méthodes
Cinquante-neuf patients consécutifs ont été inclus de manière prospective. Plusieurs données cliniques, biologiques, échographiques et invasives ont été collectées simultanément.
Résultats
Le BNP était indépendamment prédit par la fraction d’éjection, la fonction diastolique et l’âge ( p < 0,05). Le volume de l’oreillette gauche était indépendamment prédit par la masse ventriculaire et la pression télédiastolique ventriculaire gauche invasive ( p < 0,01). Le BNP prédisait une élévation de la pression télédiastolique ventriculaire gauche supérieure à 16 mmHg ( p = 0,004) ; la valeur seuil optimale était 33 pg/mL (aire sous la courbe ROC de 0,74 [0,6–0,84], p < 0,001, sensibilité 72 %, spécificité 70 %). L’oreillette gauche prédisait une élévation de la pression télédiastolique ventriculaire gauche ( p < 0,001) ; la valeur seuil optimale était 26 mL/m 2 (aire sous la courbe ROC de 0,87 [0,75–0,94], p < 0,001, sensibilité 85 %, spécificité 80 %). À la différence du BNP ( p = 0,1), l’oreillette gauche restait prédictive chez les patients avec anomalie de relaxation au flux Doppler mitral ( p < 0,01).
Conclusion
La concentration en BNP est influencée par l’âge chez les patients stables avec fonction systolique préservée et doit être interprétée avec prudence. Le volume de l’oreillette gauche est un puissant prédicteur de la pression télédiastolique ventriculaire gauche invasive indépendamment de l’âge et du profil mitral.
Background
HFnlEF accounts for half of patients with the clinical syndrome of heart failure. Establishing this medical condition is difficult in daily clinical practice, although specific recommendations have been proposed recently by French and European cardiology societies . In particular, affirmation of HFnlEF is challenging in stable patients because exertional symptoms are non-specific for the diagnosis . Natriuretic peptides and left atrial size have been related successfully to invasively determined left ventricular end-diastolic pressure – a well-recognized marker of global myocardial dysfunction – and have been proposed subsequently for the diagnosis of HFnlEF in patients with normal ejection fraction . However, these two variables are likely to be influenced by several confounders and little is known about factors affecting natriuretic peptides and left atrial size in stable patients with preserved left ventricular systolic function . The aim of the present study was to identify factors that influence the relation of BNP and left atrial size to invasively determined left ventricular end-diastolic pressure in stable patients with preserved left ventricular systolic function.
Methods
Patients
Seventy-eight consecutive patients referred for clinically indicated catheterization were enrolled prospectively after written informed consent was obtained. Patients had to be in sinus rhythm, clinically stable and free of symptoms and signs of decompensated heart failure and symptomatic coronary artery disease (including angina, acute coronary syndromes and acute myocardial infarction) to be enrolled. All data (laboratory tests, Doppler echocardiography and catheterization) were collected on the same morning for each patient. Patients with a left ventricular ejection fraction less than 45%, severe left-sided valve disease and/or mild to moderate mitral stenosis at Doppler echocardiography were excluded. The final study group comprised 59 patients.
Laboratory tests
A blood sample was collected from all the patients during bed rest, early in the morning, to avoid the potential influence of circadian variations and exercise on BNP concentration , which was measured at our institution’s laboratory with an ADVIA Centaur system (Siemens Healthcare Diagnostics Inc., Tarrytown, NY, USA; range of 5–5000 pg/mL). ADVIA Centaur BNP is harmonized to the Biosite Triage assay with similar decision threshold values. Serum creatinine and serum haemoglobin were measured at the same time. Creatinine clearance (mL/min/m 2 ) was calculated with the Cockroft formula.
Doppler echocardiography
All patients underwent bedside Doppler echocardiography with an Acuson Sequoia ultrasound system (Siemens, Mountain View, CA, USA). All echocardiography measurements were made by the same operator, who was blinded to the data; the measurements were performed according to recommendations from the American Society of Echocardiography and validated by a second experienced operator. Optimal Doppler gain and filter settings were adjusted carefully. Three to five beats were averaged for all variables. Fractional shortening was measured in all patients. Left ventricular ejection fraction was measured with Simpson’s biplane method; Teichholz’s method was used in association with a visual estimate in patients with inadequate two-dimensional images in apical views. Peak early and late diastolic mitral velocities (E and A, cm/s) were recorded placing the pulsed-wave Doppler sample volume between the tips of mitral leaflets. Diastolic function was categorized according to mitral filling pattern and invasive data as follows: normal mitral filling (E/A ratio 1–2 and normal left ventricular end-diastolic pressure); abnormal relaxation (E/A ratio < 1); pseudonormal mitral filling (E/A ratio 1–2 and increased left ventricular end-diastolic pressure); and restrictive mitral filling (E/A ratio > 2 and increased left ventricular end-diastolic pressure). The severity of mitral regurgitation was assessed by the flow convergence method. The tissue Doppler-derived peak systolic s’ velocity was measured by spectral tissue Doppler at the lateral side of the mitral annulus and used as a surrogate for longitudinal systolic function. The peak early diastolic e’ velocity was measured by spectral tissue Doppler at the lateral side of the mitral annulus and used as a surrogate for left ventricular relaxation. The E/e’ ratio was calculated and used as a non-invasive surrogate for instantaneous left atrial pressure. Maximal left atrial volume was measured at end-systole in the four-chamber apical view by Simpson’s method and indexed to the body surface area (LAVi, mL/m 2 ) .
Invasive data
Baseline systolic aortic pressure and left ventricular end-diastolic pressure were derived from left-sided heart catheterization with a 4–6F fluid-filled pigtail catheter. Five consecutive beats were averaged at the end of expiration. The presence of coronary artery disease at coronary angiography was defined as a stenosis greater than 70% in one or more epicardial arteries, detected by two experienced operators. Increased left ventricular end-diastolic pressure was defined by a value greater than 16 mmHg in this setting .
Statistical analysis
Descriptive data are given as means ± standard deviations or medians [25th–75th percentiles] when appropriate. Log-transformed data were used for BNP concentrations. Variables of influence were identified by multiple regression analysis. The multivariable model included only variables that reached significance in the univariate model (i.e. p < 0.05), using a stepwise approach. The diagnostic performance of BNP, LAVi and E/e’ in predicting invasive left ventricular end-diastolic pressure greater than 16 mmHg was assessed by logistic regression analysis. The diagnostic performance of variables was also evaluated using the area under the receiver-operating characteristic curve (AUC), which was provided with its 95% confidence interval. The optimal cut-off corresponded to the value with the greatest accuracy. A p -value less than 0.05 was considered to be statistically significant. Medcalc statistical software (version 11.1.0, MedCalc Software, Mariakerke, Belgium) was used for the purpose of statistical analysis.
Results
Baseline clinical characteristics are summarized in Table 1 . Age ranged from 34–89 years. Five patients presented with chronic kidney disease defined by a creatinine clearance less than 45 mL/min/m 2 . Invasive left ventricular end-diastolic pressure was increased in 39 patients (66%). BNP correlated poorly with LAVi ( r 2 = 0.07, p = 0.036).
| Variable | |
|---|---|
| Age (years) | 64 ± 12 |
| Men | 37 (63) |
| Systolic blood pressure (mmHg) | 148 ± 31 |
| Heart rate (beats/min) | 65 [55–74] |
| Body mass index (kg/m 2 ) | 27 ± 5 |
| New York Heart Association class | 2 ± 0.7 |
| Left ventricular ejection fraction (%) | 59 ± 7 |
| Fractional shortening (%) | 34 ± 6 |
| Tissue Doppler s’ velocity (cm/s) | 9.4 [8–10] |
| Tissue Doppler e’ velocity (cm/s) | 10.5 [9.4–12.5] |
| Tissue Doppler E/e’ ratio | 6.6 [5.2–8.8] |
| Diastolic function | |
| Normal | 7 (12) |
| Impaired relaxation | 28 (47) |
| Pseudonormal | 21 (36) |
| Restrictive | 3 (5) |
| Left ventricular mass index (g/m 2 ) | 92 [80–109] |
| Indexed left ventricular end-diastolic diameter (mm/m 2 ) | 27 ± 3.5 |
| Left atrial volume index (mL/m 2 ) | 31 [22–39] |
| B-type natriuretic peptide concentration (pg/mL) | 53 [20–137] |
| Medical history | |
| Coronary artery disease | 29 (49) |
| Hypertension | 34 (58) |
| Diabetes mellitus | 21 (36) |
| Dyslipidaemia | 32 (54) |
| Pulmonary oedema | 5 (9) |
| Chronic pulmonary disease | 9 (15) |
| Chronic medication | |
| Diuretic | 20 (34) |
| Beta-blocker | 35 (59) |
| ACE inhibitor/ARB | 40 (68) |
| Invasive left ventricular end-diastolic pressure (mm Hg) | 20 ± 7 |
| Coronary artery disease | 30 (51) |
| One vessel | 11 (19) |
| Two vessels | 8 (14) |
| Three vessels | 11 (19) |
| Serum haemoglobin (g/dL) | 12.8 ± 1.7 |
| Creatinine clearance (mL/min) | 85 [68–108] |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
