Author + information
- Received May 19, 1997
- Revision received September 4, 1997
- Accepted September 25, 1997
- Published online January 1, 1998.
- Noritoshi Nagaya, MDA,
- Toshio Nishikimi, MD, PhDA,* (, )
- Yoshiaki Okano, MDA,
- Masaaki Uematsu, MD, PhDC,
- Toru Satoh, MD, PhDA,
- Shingo Kyotani, MD, PhDA,
- Sachio Kuribayashi, MD, PhDB,
- Seiki Hamada, MD, PhDB,
- Mikio Kakishita, MDA,
- Norifumi Nakanishi, MD, PhDA,
- Makoto Takamiya, MD, PhDB,
- Takeyoshi Kunieda, MD, PhDA,
- Hisayuki Matsuo, PhDC and
- Kenji Kangawa, PhDC
- ↵*Dr. Toshio Nishikimi, Department of Hypertension, National Cardiovascular Center, 5-7-1 Fujishirodai, Suita, Osaka, 565 Japan.
Objectives. This study sought to investigate the influence of right ventricular (RV) hemodynamic variables and function on the secretion of brain natriuretic peptide (BNP) in patients with isolated RV overload.
Background. Plasma BNP is known to increase in proportion to the degree of left ventricular (LV) overload. However, whether BNP secretion is also regulated in the presence of RV overload remains unknown.
Methods. Plasma BNP and atrial natriuretic peptide (ANP) levels in the pulmonary artery were measured in 44 patients with RV overload: 18 with RV volume overload (RVVO) due to atrial septal defect and 26 with RV pressure overload (RVPO) due to primary or thromboembolic pulmonary hypertension. Right heart catheterization was performed in all patients. RV and LV ejection fraction, myocardial mass and volume of the four chambers were determined by using electron beam computed tomography.
Results. Although both plasma BNP and ANP levels were significantly elevated in patients with RV overload compared with values in control subjects, plasma BNP and the BNP/ANP ratio were significantly higher in patients with RVPO than with RVVO (BNP 294 ± 72 vs. 48 ± 14 pg/ml; BNP/ANP 1.6 ± 0.2 vs. 0.8 ± 0.2, both p < 0.05). Plasma BNP correlated positively with mean pulmonary artery pressure (r = 0.73), total pulmonary resistance (r = 0.79), mean right atrial pressure (r = 0.79), RV end-diastolic pressure (r = 0.76) and RV myocardial mass (r = 0.71); it correlated negatively with cardiac output (r = −0.33) and RV ejection fraction (r = −0.71). Plasma BNP significantly decreased from 315 ± 120 to 144 ± 54 pg/ml with long-term vasodilator therapy (total pulmonary resistance decreased from 23 ± 4 to 15 ± 3 Wood U).
Conclusions. Plasma BNP increases in proportion to the extent of RV dysfunction in pulmonary hypertension.
Brain natriuretic peptide (BNP) was first isolated from the porcine brain and was later found in the human heart . BNP shares with atrial natriuretic peptide (ANP) a high degree of structural homology and a profile of diuretic, natriuretic and vasodilator activities and inhibition of the renin-angiotensin-aldosterone system . Unlike ANP, however, plasma BNP has as its main source the cardiac ventricles, suggesting that BNP may be more sensitive and specific than ANP as an indicator of ventricular disorders. Earlier studies have shown that plasma BNP levels are increased in states of left ventricular (LV) overload, such as LV failure or LV hypertrophy and this can be a useful biochemical marker for LV overload.
Regarding right ventricular (RV) overload, an experimental study has shown that plasma BNP levels are elevated in association with increased pulmonary hypertension in rats. In humans, plasma ANP levels have been shown to be elevated in patients with pulmonary hypertension. However, there has been no report regarding the modes of BNP secretion in isolated RV overload, in part because the complicated configuration of the RV has made it difficult to measure RV volume and mass by conventional methods. Electron beam computed tomography (CT) has made it possible to evaluate RV volume and mass with a high degree of accuracy [8, 9].
Thus, the purpose of this study was to use this method to investigate the influence of RV hemodynamics and function on the secretion of BNP in patients with RV overload.
1.1 Study Patients
We studied 44 patients with isolated RV overload (19 men and 25 women, mean age 47 years [range 20 to 68]). Eighteen patients with atrial septal defect served as a model of RV volume overload (RVVO) and 26 with pulmonary artery hypertension (10 with primary pulmonary hypertension and 16 with chronic thromboembolic pulmonary hypertension) as a model of RV pressure overload (RVPO). The diagnosis of atrial septal defect was made by echocardiography and was confirmed by cardiac catheterization . All patients had uncomplicated secundum atrial septal defect. No patient had evidence of the Eisenmenger syndrome. Patients with pulmonary valve stenosis were not included. Pulmonary artery hypertension was defined by a mean pulmonary artery pressure >25 mm Hg at rest . Primary pulmonary hypertension was defined as pulmonary hypertension unexplained by any secondary cause, based on the criteria of National Institutes of Health registry on primary pulmonary hypertension . Chronic thromboembolic pulmonary hypertension was identified by radionuclide perfusion lung scans and pulmonary angiography [12, 13]. All patients were in a steady state during this study protocol. Patients with one or more of the following conditions were excluded: 1) significant LV diseases, such as primary myocardial, valvular or coronary artery disease; 2) acute heart failure; 3) chronic renal impairment (serum creatinine ≥133 μmol/liter); and 4) sustained arrhythmia, such as atrial fibrillation. The study included 11 age-matched control subjects (5 men and 6 women, mean age 50 years [range 29 to 65]). All subjects provided written informed consent.
1.2 Hemodynamic Studies
Right heart catheterization was performed in all subjects after an overnight fast. The following hemodynamic variables were obtained at end-expiration: mean pulmonary artery pressure, mean right atrial pressure, RV end-diastolic pressure, pulmonary capillary wedge pressure and mean systemic arterial pressure. Cardiac output was determined by the Fick method. Total pulmonary resistance was calculated by dividing mean pulmonary artery pressure by cardiac output. Control subjects demonstrated normal hemodynamics by right heart catheterization. Pulmonary to systemic flow ratio in patients with atrial septal defect was obtained by the oximetric principle of Fick .
1.3 Electron Beam CT
Ventricular and atrial volume and myocardial mass were measured by using electron beam CT in 35 patients (14 of the 18 patients with RVVO and 21 of the 26 with RVPO) within 1 month of cardiac catheterization and blood sampling. No changes in the clinical status or medication regimens occurred between the CT and cardiac catheterization studies. Three patients with primary pulmonary hypertension and one with thromboembolic pulmonary hypertension were excluded from the CT study because of severe RV dysfunction. Four patients with atrial septal defect and one with thromboembolic pulmonary hypertension rejected the protocol. Electron beam CT was performed with a C-150 scanner (Imatron), as previously reported . In brief, cine mode scanning (scanning time 50 ms for 256 matrix images) was performed after the administration of 40 to 50 ml of nonionic contrast medium (Iopamidol 370, Nippon Schering). The scanner table was rotated 25° in a clockwise horizontal direction to obtain near short-axis views of the heart. Eight-level (10 contiguous images/level) or 10-level (8 contiguous images/level) cine mode scans of the heart were obtained with electrocardiographic (ECG) gating. The end-diastolic (R wave on the ECG) and end-systolic (smallest ventricular chamber volume during the cardiac cycle) frames were identified at each tomographic level. Endocardial and epicardial borders of the ventricles were determined by using previously described methods of edge detection for electron beam CT (Fig. 1) .
We used a modified version of Simpson’s method to obtain multisection cine mode scans and then calculated the following variables: RV end-diastolic volume index, RV ejection fraction, RV myocardial mass index, right atrial end-systolic volume index, LV end-diastolic volume index, LV ejection fraction, LV myocardial mass index and left atrial end-systolic volume index.
Reproducibility of CT measurements (mean ± SD) was assessed in 10 subjects randomly allocated from the patients with RVVO and RVPO. The mean differences between the measurements in each subject were 2 ± 1% for RV ejection fraction and 3 ± 3 ml/m2for RV myocardial mass index. Interobserver variability was also assessed in the same 10 subjects by two independent observers (N.N. and S.H.). The mean differences between the measurements of the two observers were 3 ± 2% for RV ejection fraction and 3 ± 4 ml/m2for RV myocardial mass index.
1.4 Blood Sampling and Assay for ANP and BNP
Blood samples were drawn simultaneously from the main pulmonary artery and femoral vein during cardiac catheterization. Blood was immediately transferred into chilled glass tubes containing disodium EDTA (1 mg/ml) and aprotinin (500 U/ml); it was centrifuged immediately at 4°C, and the plasma was frozen and stored at −80°C until assayed. Plasma ANP and BNP levels were measured directly with specific immunoradiometric assay kits (Shiono RIA ANP assay kit and Shiono RIA BNP assay kit, Shionogi Co., Ltd., Osaka, Japan) . Data analyses were done on the samples from the main pulmonary artery unless otherwise stated.
1.5 Effects of Acute Hemodynamic Changes on Plasma ANP and BNP Levels
Nitric oxide (NO) was inhaled by 18 patients with pulmonary hypertension (8 with primary and 10 with thromboembolic pulmonary hypertension). NO gas (2,000 ppm in nitrogen gas) was mixed with air with use of an originally assembled apparatus that controlled each gas flow by computerized mass flow meters . The mixture was then introduced into a nonrebreathing circuit consisting of large bore tubing and a continuous positive airway pressure mask. After a complete baseline hemodynamic evaluation and blood sampling for the measurement of plasma ANP and BNP, NO at 80 ppm was inhaled for 10 min. Subsequently, the hemodynamic evaluation and blood sampling were repeated. Patients were classified as responders and nonresponders on the basis of hemodynamic responses to the NO trial. Responders were defined as those whose total pulmonary resistance decreased by >20% of the basal value . Eight patients were responders and 10 were nonresponders. Of the eight responders, two had primary pulmonary hypertension and six had thromboembolic pulmonary hypertension.
1.6 Effects of Long-Term Hemodynamic Changes on Plasma ANP and BNP Levels
We examined nine patients with primary pulmonary hypertension who underwent long-term vasodilator therapy: 4 to 9 ng/kg per min of prostacyclin in four patients and 5 to 12 ng/kg per min of prostaglandin E1in five patients. The mean follow-up period was 35 days (range 11 to 63). Right heart catheterization and blood sampling for ANP and BNP measurements were simultaneously performed before and after these treatments.
1.7 Statistical Analysis
All data were expressed as mean value ± SEM unless otherwise indicated. Log transformation was used to normalize the distribution of plasma ANP and BNP levels unless otherwise indicated. Comparisons of hemodynamic data and cardiac function between RVVO and RVPO groups were performed with the Student ttest for unpaired values. Comparisons of the mean values among the three groups were made by using one-way analysis of variance followed by the Scheffé multiple comparison test. To identify relevant relations, linear regression analyses of hemodynamics and cardiac function with plasma ANP or BNP were performed. Those variables found to be significant in the linear regression analysis were further examined by using multivariate regression, and independent relations of hemodynamics and cardiac function with plasma ANP or BNP were determined. The effects of vasodilator therapy on plasma ANP and BNP were analyzed with the Student ttest for paired values. A p value <0.05 was considered statistically significant.
2.1 Hemodynamic Variables and Cardiac Function
The mean pulmonary artery pressure, total pulmonary resistance, mean right atrial pressure, RV end-diastolic pressure and RV myocardial mass index were significantly higher in patients with RVPO than with RVVO (Table 1). Cardiac output and RV ejection fraction were significantly lower in patients with RVPO. In contrast, RV end-diastolic volume index and right atrial end-systolic volume index were significantly higher in patients with RVVO than with RVPO. There were no significant differences between the two groups in pulmonary capillary wedge pressure, mean systemic arterial pressure and in all left-sided functional variables.
2.2 Plasma ANP and BNP Levels
Plasma ANP levels were significantly higher in patients with RVVO and RVPO than in control subjects (Fig. 2A), and they tended to be higher in patients with RVPO than in patients with RVVO (p = NS). Plasma BNP levels were also significantly higher in patients with RVVO or RVPO than in control subjects (Fig. 2B). Elevations in plasma BNP levels versus control values were more marked in patients with RVPO (∼27-fold) than in those with RVVO (∼4-fold). The BNP/ANP ratio was also significantly higher in patients with RVPO than in those with RVVO or control subjects (Fig. 2C). Plasma ANP or BNP levels did not show significant correlations with age or heart rate in the patients studied (data not shown).
2.3 Relation Between Hemodynamic Variables and Plasma ANP and BNP Levels
Plasma ANP levels positively correlated with mean pulmonary artery pressure and inversely with cardiac output; therefore, they showed a strong positive correlation with total pulmonary resistance (Table 2). Plasma ANP levels also correlated positively with mean right atrial pressure and RV end-diastolic pressure but not with pulmonary capillary wedge pressure or mean systemic arterial pressure. Similarly, plasma BNP levels correlated positively with mean pulmonary artery pressure and inversely with cardiac output, thereby showing a strong positive correlation with total pulmonary resistance (Table 2Fig. 3). Plasma BNP levels correlated positively with mean right atrial pressure and RV end-diastolic pressure but not with pulmonary capillary wedge pressure or mean systemic arterial pressure.
2.4 Relation Between Cardiac Function and Plasma ANP and BNP Levels
Plasma ANP levels correlated inversely with RV ejection fraction and positively with RV myocardial mass index (Table 2). However, there were no significant correlations between plasma ANP levels and RV end-diastolic volume index, right atrial end-systolic volume index, LV ejection fraction, LV end-diastolic volume index, LV myocardial mass index or left atrial end-systolic volume index. Plasma BNP levels correlated inversely with RV ejection fraction and positively with RV myocardial mass index, but there were no significant correlations between plasma BNP levels and RV end-diastolic volume index, right atrial end-systolic volume index, LV ejection fraction, LV end-diastolic volume index, LV myocardial mass index or left atrial end-systolic volume index (Table 2Fig. 4). The four patients who were excluded from the CT study because of severe RV dysfunction had markedly poor hemodynamics and high plasma levels of ANP and BNP (mean pulmonary artery pressure 66 ± 9 mm Hg, cardiac output 2.1 ± 0.1 liters/min, total pulmonary resistance 31 ± 5 Wood U, RV end-diastolic pressure 20 ± 2 mm Hg, mean right atrial pressure 18 ± 2 mm Hg, ANP 520 ± 165 pg/ml and BNP 878 ± 162 pg/ml).
2.5 Multivariate Analysis
As previously described, univariate analyses demonstrated significant correlations of mean pulmonary artery pressure, cardiac output, total pulmonary resistance, mean right atrial pressure, RV end-diastolic pressure, RV ejection fraction or RV myocardial mass index with plasma ANP and BNP levels. Mean pulmonary artery pressure, cardiac output, mean right atrial pressure, RV end-diastolic pressure, RV ejection fraction and RV myocardial mass index, excluding total pulmonary resistance, were examined as independent factors because total pulmonary resistance is a variable derived from mean pulmonary artery pressure and cardiac output. Among these variables, mean right atrial pressure and RV end-diastolic pressure correlated independently with plasma ANP levels (p < 0.05). RV end-diastolic pressure and RV ejection fraction correlated independently with plasma BNP levels (p < 0.05).
2.6 Effects of Acute Hemodynamic Changes on Plasma ANP and BNP Levels
In responders, plasma ANP levels decreased significantly after inhalation of NO (121 ± 41 to 55 ± 23 pg/ml, Fig. 5A). In contrast, plasma BNP levels were not significantly altered even after NO inhalation (208 ± 126 to 194 ± 122 pg/ml, Fig. 5B). Nonresponders showed no significant changes in plasma ANP (206 ± 58 to 194 ± 50 pg/ml) or BNP (321 ± 98 to 329 ± 97 pg/ml).
2.7 Effects of Long-Term Hemodynamic Changes on Plasma ANP and BNP Levels
After long-term vasodilator treatment, total pulmonary resistance decreased in all patients with pulmonary hypertension (23 ± 4 to 15 ± 3 Wood U, p < 0.05). Plasma ANP levels were reduced from 148 ± 34 to 72 ± 21 pg/ml (p < 0.05), and plasma BNP levels from 315 ± 120 to 144 ± 54 pg/ml (p < 0.05). The decrease in plasma ANP and in plasma BNP showed significant correlations with the decrease in total pulmonary resistance (Fig. 6).
2.8 Comparison of Plasma ANP and BNP Levels Between the Pulmonary Artery and Femoral Vein
Both plasma ANP and BNP levels revealed close correlations between the samples from the main pulmonary artery and those from the femoral vein (ANP, r = 0.96, p < 0.001; BNP, r = 0.99, p < 0.001), although the plasma levels were higher in the main pulmonary artery than in the femoral vein (ANP, 130 ± 29 vs. 80 ± 15 pg/ml; BNP, 193 ± 46 vs. 155 ± 36 pg/ml, p < 0.05).
In this study, we demonstrated for the first time 1) that both plasma BNP and ANP levels were elevated in RV overload, particularly the plasma BNP levels elevated in RV pressure overload, and 2) that plasma BNP levels correlated positively with mean pulmonary artery pressure, total pulmonary resistance, mean right atrial pressure, RV end-diastolic pressure and RV myocardial mass index; and correlated negatively with cardiac output and RV ejection fraction, whereas plasma BNP levels did not correlate significantly with any LV variables. Finally, we demonstrated 3) that plasma BNP levels changed in association with long-term changes in hemodynamics, thereby serving as a potential indicator of the efficacy of long-term vasodilator treatment in patients with pulmonary hypertension.
3.1 Plasma BNP in RV Overload
ANP is released mainly from atrial tissue in healthy subjects, although it is also secreted from ventricular tissue in patients with ventricular disorders. In contrast, BNP is secreted predominantly from ventricular tissue [1, 2]. BNP messenger ribonucleic acid (RNA) can be detected not only in the LV but also in the RV of normal human cardiac tissues obtained at autopsy . These findings raise the possibility that different types of overload of the left or right atria or ventricles, or both, may result in differing secretion patterns of ANP and BNP. In fact, plasma BNP levels have been shown to be elevated in patients with LV overload in proportion to the degree of LV dysfunction or LV hypertrophy .
In contrast, little information has previously been available regarding plasma BNP levels in patients with isolated RV overload. An experimental study showed that BNP was secreted mainly from the RV in hypoxic pulmonary hypertension. In humans, Lang et al. showed that plasma BNP levels were elevated in proportion to the degree of hypoxemia in chronic obstructive pulmonary disease. They speculated that BNP release was triggered by the increased pressure and volume stress on the right side of the heart. However, RV pressure and volume were not assessed in their study. Thus, the mechanism of BNP secretion in RV overload remained to be elucidated.
To examine the factors regulating BNP secretion in RV overload, we assessed hemodynamic variables by cardiac catheterization and ventricular volume, mass, and ejection fraction by electron beam CT. Volume, mass and function of the RV have been technically difficult to determine by conventional methods because of its complex anatomic geometry. However, electron beam CT made it possible to measure those RV variables, and the reproducibility and accuracy of these CT measurements have been validated in both animal and human studies elsewhere [8, 9].
In the present study, plasma BNP levels were correlated with RV variables such as RV ejection fraction, RV end-diastolic pressure and total pulmonary resistance. These results suggest that BNP secretion is influenced by the severity of RV dysfunction in pulmonary hypertension.
3.2 Effects of Acute and Long-Term Hemodynamic Changes on BNP Secretion
NO inhalation has been shown to produce selective vasodilation of pulmonary vascular beds without systemic effects because of the rapid inactivation of NO by hemoglobin . Plasma ANP levels in responders decreased significantly after NO inhalation, whereas plasma BNP levels were unaltered. This difference may be attributed to the different patterns of synthesis, secretion and clearance of ANP versus BNP. ANP is released mainly from stored granules in atrial tissue by way of a regulated pathway . In contrast, BNP is secreted predominantly from cardiac ventricular tissue by way of a constitutive pathway . In humans, ANP and BNP have different plasma half-lives (ANP 3 min; BNP 20 min) [21, 22]. Thus, plasma ANP is more likely than plasma BNP to be affected by short-term hemodynamic changes. Plasma ANP levels may be a marker for the efficacy of short-term vasodilator trials, which have usually been performed in patients with pulmonary hypertension to evaluate potential vasodilator response [23, 24].
We investigated the effects of long-term hemodynamic unloading on plasma ANP and BNP levels in patients with long-term vasodilator therapy. Both plasma ANP and BNP levels, but particularly the latter, reflected the improvement in total pulmonary resistance during long-term follow-up periods. Thus, plasma BNP may be a potential marker for the efficacy of long-term treatment for RV dysfunction.
Plasma ANP levels are easily influenced by blood pressure, age, sodium intake, renal function, postural change and exercise [25–29]. In addition, increments in plasma ANP in patients with pulmonary hypertension are small compared with those in plasma BNP. Thus, plasma ANP levels may be significantly affected when blood samples are obtained in nonstandardized settings. In contrast, plasma BNP levels are not easily affected by postural changes or exercise. Therefore, plasma BNP may be more suitable than ANP for the evaluation of the pathophysiology of pulmonary hypertension.
3.3 Pathophysiologic Role of BNP in RV Overload
The present study showed that plasma BNP levels and the BNP/ANP ratio were markedly elevated in patients with RVPO. BNP has been reported [3, 6]to have various biochemical effects such as diuretic and natriuretic activities, potent vasodilator effects both for pulmonary and systemic circulation and suppression of aldosterone secretion. These findings suggest that BNP may play an important role in maintaining cardiopulmonary homeostasis in RV overload through a reduction in preload and afterload. Recently, Pidgeon et al. demonstrated that the natriuretic and blood pressure–lowering effects of BNP are two to three times as powerful as those of equal amounts of ANP. In view of the strong biologic effects of BNP, the disproportionate elevation of BNP over ANP in patients with RVPO may represent a compensatory mechanism under conditions such as severe right heart failure due to pulmonary hypertension.
3.4 Study Limitations
1) Four patients were excluded from the CT study because of severe RV dysfunction. However, inclusion of these patients should not have affected the conclusions drawn from our data, because all four had markedly poor hemodynamic status and high plasma levels of ANP and BNP. 2) It would be interesting to investigate changes in RV ejection fraction associated with long-term vasodilator therapy in patients with primary pulmonary hypertension. However, we could not repeat the CT study in the follow-up period because of the potential side effects of repeated examinations in patients with severe pulmonary hypertension. 3) The usefulness of plasma BNP as a marker for RV dysfunction still needs to be validated in a longitudinal large-scale study.
Plasma BNP levels increase in proportion to the extent of RV dysfunction in pulmonary hypertension.
We thank Yoko Saito, BS for technical assistance and Nobuo Shirahashi, BS for helpful advice regarding statistical analysis.
☆ This work was supported in part by Special Coordination Funds for Promoting Science and Technology (Encouragement System of COE) from the Science and Technology Agency of Japan.
- atrial natriuretic peptide
- brain natriuretic peptide
- computed tomography (tomographic)
- electrocardiogram, electrocardiographic
- left ventricle (ventricular)
- nitric oxide
- right ventricle (ventricular)
- right ventricular pressure overload
- right ventricular volume overload
- Received May 19, 1997.
- Revision received September 4, 1997.
- Accepted September 25, 1997.
- The American College of Cardiology
- Yoshimura M,
- Yasue H,
- Morita E,
- et al.
- Yasue H,
- Yoshimura M,
- Sumida H,
- et al.
- Hill NS,
- Klinger JR,
- Warburton RR,
- Pietras L,
- Wrenn DS
- Nootens M,
- Kaufmann E,
- Rector T,
- et al.
- Reiter SJ,
- Rumberger JA,
- Feiring AJ,
- Stanford W,
- Marcus ML
- Hajduczok ZD,
- Weiss RM,
- Stanford W,
- Marcus ML
- Tada H,
- Shimizu W,
- Ohe T,
- et al.
- Kunieda T,
- Kyotani S,
- Satoh T,
- Okano Y,
- Nakanishi N,
- Hukushima T
- Weir EK,
- Rubin LJ,
- Ayres SM,
- et al.
- Hosoda K,
- Nakao K,
- Mukoyama M,
- et al.
- Raffy O,
- Azarian R,
- Brenot F,
- et al.
- Pidgeon GB,
- Richards AM,
- Nicholls MG,
- Espiner EA,
- Yandle TG,
- Frampton C