Objective
The diagnostic and prognostic value of plasma B-type natriuretic peptide (BNP) level in isolated aortic stenosis (AS) has not been fully understood.
Methods
BNP level was determined in 109 consecutive patients with isolated severe AS (68.1 ± 10.6 years; 53 men; transvalvular peak gradient, 87.2 ± 37.0 mm Hg; valve area index, 0.43 ± 0.14 cm 2 /m 2 ) and 12 healthy volunteers in their stable state. They were followed up for 36 months.
Results
BNP level increased with New York Heart Association (NYHA) class (75.2 ± 95.9 pg/mL, 135.0 ± 112.0 pg/mL, 450.6 ± 366.3 pg/mL, and 1478.9 ± 941.5 pg/mL for NYHA I, II, III, and IV, respectively). Left ventricular (LV) mass index had the best relationship with BNP ( r = 0.73, P < .0001). Aortic valve replacement (AVR) was eventually performed in 95 patients (male = 44, age = 67.8 ± 9.3 years). Echocardiography was repeated early ( n = 88, 13.2 ± 6.2 day) and late ( n = 62, 32 ± 10 months) after AVR. Preoperative BNP level correlated with LV mass index early ( r = 0.74, P < .0001) and late ( r = 0.78, P < .0001) after AVR. Patients with higher BNP level had a tendency to show cardiac symptoms (NYHA > I) late after AVR (NYHA I vs. > I = 160.8 ± 197.9 pg/mL vs. 504.3 ± 567.3 pg/mL, P < .0001). Preoperative BNP level predicted the occurrence of perioperative complications ( P < .0001). During follow-up of the 94 patients (44 ± 10 months after AVR), 10 were readmitted for major cardiac and cerebrovascular events, including 9 patients with congestive heart failure and 1 patient with ischemic stroke. An event-free survival rate was significantly higher in patients with BNP ≤ 312 pg/mL than in patients with BNP > 312 pg/mL (log rank, χ 2 = 10.21, P = .001). Multiple logistic regression analysis revealed that BNP > 312 pg/mL was an independent predictor of AVR complication (odds ratio 5.58; confidence interval, 1.82–20.16; P = .002). Furthermore, BNP was the strongest predictor of major adverse cardiac and cerebrovascular events within 36 months after AVR (odds ratio 8.80; confidence interval, 1.83–42.35; P = .006).
Conclusion
Plasma BNP level reflects the degree of heart failure, is associated with LV structure and function in severe AS, and is an independent predictor of complication and outcome after AVR. BNP level may be useful in risk stratification of patients with AS in conjunction with other clinical and echocardiographic parameters.
Plasma B-type natriuretic peptide (BNP), a cardiac hormone secreted mainly by the cardiac ventricle, has been used as a noninvasive marker of left ventricular (LV) dysfunction and a prognostic indicator in a variety of patients with heart failure. The usefulness of BNP measurement has been shown in the diagnosis of congestive heart failure (CHF). Hutfless et al. reported that elevated preoperative BNP level predicts postoperative complications after heart surgery. Several studies have demonstrated that the plasma levels of natriuretic peptides are related to the severity of aortic stenosis (AS). Specifically, these levels were elevated in symptomatic patients with AS, and they provided important prognostic information for severe AS. Two recent studies showed that N-terminal pro-BNP and BNP were independent predictors of perioperative mortality, and the latter study also demonstrated the superiority of BNP in the prediction of perioperative and long-term mortality compared with the logistic European System for Cardiac Operative Risk Evaluation. However, whether the optimal timing of AVR in patients with isolated severe AS can be predicted by measuring plasma BNP remains unclear. A decline in the incidence of rheumatic fever and an increasing proportion of the population aged more than 65 years have led to an increase in the number of patients with AS, which is now appearing as a common valvular lesion. Thus, the importance of diagnosis and risk stratification of AS has recently increased.
The present study assessed the diagnostic significance of plasma BNP level in patients with isolated severe AS by comparing clinical, echocardiographic, and hemodynamic variables. The value of BNP level was also assessed as a prognosticator after AVR in comparison with the other parameters previously reported.
Materials and Methods
Study Subjects
There were 130 consecutive patients with isolated AS (peak transaortic flow velocity ≥ 2.5 m/s) without significant aortic regurgitation and mitral regurgitation who were referred to the National Cerebral and Cardiovascular Center, Suita, Japan. Among them, 21 patients were excluded for the following reasons: (1) plasma creatinine level ≥ 0.133 mmol/L (=1.5 mg/dL), (2) prior myocardial infarction, (3) more than mild mitral valve disease or aortic regurgitation, (4) severe pulmonary disease, (5) chronic atrial fibrillation or flutter, and (6) acute coronary syndrome. The remaining 109 patients (53 men and 56 women, mean age 68.1 ± 10.6 years, range 22–89 years; transvalvular peak gradient, 86.6 ± 36.6 mm Hg; valve area index, 0.43 ± 0.14 cm 2 /m 2 ) were studied prospectively. Informed written consent was obtained from each patient before the beginning of the study. We studied 12 age-matched healthy volunteers as controls (7 men and 5 women; mean age 62.9 ± 10.3; age range, 40–78 years). New York Heart Association (NYHA) functional class was assessed by experienced cardiologists blinded to the results of echocardiographic and neurohormonal measurements. The presence of coronary artery disease was defined as significant stenosis (≥75% diameter stenosis) with coronary angiography.
Measurement of Plasma BNP
A blood sample was drawn via the antecubital vein while patients were in the supine position and a stable hemodynamic state. The plasma BNP level was measured directly with a specific immunoradiometric assay kit (Shiono RIA BNP Assay Kit; Shionogi Co., Ltd., Osaka, Japan).
Echocardiographic Assessment
Commercially available ultrasound systems (Hewlett Packard 5500 Sonos, Philips Inc., Eindhoven, The Netherlands; Acuson Sequoia, Siemens Inc., Mountain View, CA; Vivid7, General Electric Inc., Agnotes, Norway; Prosound SSD-5500SV, Aloka Inc., Tokyo, Japan; Aplio SSA770, Toshiba Inc., Tokyo, Japan) were used for standard echocardiographic examinations. LV end-diastolic and end-systolic diameters, and septal and posterior wall thicknesses in diastole were obtained from M-mode or two-dimensional echocardiographic recordings. LV fractional shortening and mass were calculated. LV end-diastolic volume, LV end-systolic volume, and LV ejection fraction (LVEF) were calculated using biplane Simpson’s method. The volumes and LV mass were indexed (LVMI) to the body surface area. The pressure gradient across the aortic valve (peak PG) was estimated by the simplified Bernoulli equation from the transaortic flow velocity detected by continuous-wave Doppler echocardiography. The aortic valve area was calculated by the continuity equation, and the aortic valve area index (AVAI) was also calculated.
Transmitral flow (TMF) was assessed in the apical four-chamber view using pulsed-wave Doppler echocardiography, with the Doppler beam aligned parallel to the direction of flow and the sample volume at the leaflet tips. E- and A-wave peak velocities and deceleration time were measured from the TMF profile.
Cardiac Catheterization
Coronary angiography was performed on all 95 patients who were scheduled for AVR. Right-sided heart catheterization was added in 81 patients. Right atrial pressure, pulmonary artery pressure, and pulmonary capillary wedge pressure were then measured. LV catheterization was performed in 46 patients with unclear echocardiographic findings or when additional information was necessary for clinical management. LV end-diastolic and end-systolic volumes were measured by left ventriculogram, and LVEF was obtained. Aortic valve area was calculated by Gorlin and Gorlin’s equation.
Study Protocol
Baseline clinical, hemodynamic variables, echocardiographic data, and BNP levels were assessed in their stable condition. The physicians referred patients for surgery when they presented symptoms related to severe AS or reduced LV function according to the American College of Cardiology/American Heart Association guidelines, irrespective of the BNP level. AVR was performed in 95 patients (41 men, 54 women, mean age 67.8 ± 9.3 years). Among the 94 patients who survived after AVR, 88 were reevaluated clinically and by echocardiography during routine follow-up (44 ± 10 months, 30–70 months). They had echocardiography early after ( n = 88, after 13.2 ± 6.2 days, 7–27 days) and late after ( n = 62, after 32 ± 14 months, 12–65 months) surgery. The primary end point of this study was major adverse cardiac and cerebrovascular events (MACCE), which included cardiovascular death and readmission due to heart failure and stroke.
Statistical Analysis
All data were expressed as mean ± standard deviation. Comparisons of measurements between two patient groups were analyzed with unpaired Student t test. Comparisons among more than three groups were made by one-way analysis of variance, followed by Scheffe’s multiple comparison tests. Logistic regression analyses were performed to identify predictors of postoperative risk and event-free survival. Univariate analyses were performed using nonparametric statistical tests. The model used was built stepwise; the value for entering and staying in the model was set at P < .05. Survival curves were derived by means of the Kaplan–Meier method and compared by means of the log-rank test. P < .05 was considered statistically significant. Receiver operating characteristic curves were generated, and the area under the curve (AUC) and optimal threshold were determined. Event-free survival was analyzed by the Kaplan–Meier method for patient groups with BNP ≤ 312 or > 312 pg/mL, as previously reported. These analyses were performed using JMP 8.0 software (SAS Institute, Inc., Cary, NC).
Results
Characteristics of Patients
Patient characteristics are shown in Table 1 . Most of the patients were in NYHA class II and III. Echocardiographic and catheterization parameters are shown in Table 2 . Although the peak PG varied, ranging from 25.0 to 197.0 mm Hg (mean 86.3 ± 37.0 mm Hg), the AVAI was uniformly small (0.43 ± 0.14 cm 2 /m 2 ). BNP also widely ranged from 8.9 to 2525.0 pg/mL (mean 287.3 ± 413.1 pg/mL). Because we performed LV catheterization only in those patients with unclear echocardiographic findings, or when additional information was necessary for clinical management, most of them had low peak PG and LV dysfunction.
| Value | Range | |
|---|---|---|
| Clinical characteristics | ||
| Age (y) | 68.1 ± 10.6 | 22–89 |
| Male gender (%) | 53 (49) | |
| Bicuspid AS (%) | 36 (33) | |
| NYHA class | ||
| I | 15 (13) | |
| Ⅱ | 54 (50) | |
| Ⅲ | 35 (32) | |
| Ⅳ | 5 (5) | |
| BNP (pg/mL) | 287.3 ± 413.8 | 8.9–2528.0 |
| Systolic BP (mm Hg) | 125.6 ± 19.6 | 94.0–180.0 |
| Diastolic BP (mm Hg) | 67.4 ± 11.3 | 46.0–100.0 |
| Heart rate (bpm) | 68.9 ± 9.9 | 53–96 |
| Hypertension, n (%) | 70 (64) | |
| Diabetes, n (%) | 23 (21) | |
| Dyslipidemia, n (%) | 44 (40) | |
| Coronary artery disease, n (%) | 19 (17) | |
| Medications at entry | ||
| β-blocker | 13 (12) | |
| ACE inhibitors/ARB | 55 (50) | |
| Diuretic | 52 (48) | |
| Digoxin | 28 (26) |
| Value | Range | |
|---|---|---|
| Echocardiography ( n = 109) | ||
| LVEDD (mm) | 48.1 ± 7.5 | 36.0–73.0 |
| LVESD (mm) | 31.6 ± 9.1 | 20.0–66.0 |
| LVFS (%) | 34.7 ± 10.7 | 10.0–54.2 |
| LVEF (%) | 67.3 ± 12.9 | 28.8–86.6 |
| LVMI (g/m 2 ) | 182.9 ± 71.2 | 73.8–560.9 |
| Peak aortic PG (mm Hg) | 87.2 ± 37.0 | 25.0–197.0 |
| AVAI (cm 2 /m 2 ) | 0.43 ± 0.14 | 0.19–0.78 |
| Transmitral flow | ||
| E velocity (m/sec) | 78.3 ± 22.2 | 33.0–135.0 |
| A velocity (m/sec) | 80.7 ± 22.7 | 33.0–127.0 |
| E-Dct (msec) | 215.6 ± 55.7 | 110–433.4 |
| E/A | 1.1 ± 0.6 | 0.3–3.2 |
| Catheterization ( n = 81) | ||
| PCWP (mm Hg) | 10.1 ± 5.3 | 1.0–28.0 |
| Systolic PAP (mm Hg) | 30.3 ± 9.7 | 13.0–80.0 |
| Diastolic PAP (mm Hg) | 11.2 ± 5.4 | 2.0–38.0 |
| Mean PAP (mm Hg) | 17.8 ± 6.7 | 7.0–48.0 |
| LV catheterization ( n = 46) | ||
| LVSP (mm Hg) | 174.1 ± 31.4 | 118.0–244.0 |
| LVEDP (mm Hg) | 14.9 ± 5.9 | 6.0–31.0 |
| LVEDVI (mL/m 2 ) | 96.7 ± 42.1 | 53.0–250.0 |
| LVESVI (mL/m 2 ) | 50.4 ± 33.8 | 22.0–174.0 |
| LVEF (%) | 52.8 ± 33.8 | 25.0–70.0 |
NYHA Class, Echocardiography, and Catheterization Parameters
Preoperative BNP levels increased according to NYHA class ( Figure 1 ). Although the difference between class I and II did not reach statistical significance, the differences between class II and III, and between class III and IV were significant.
The correlations between BNP level and echocardiographic and catheterization parameters are shown in Table 3 . BNP moderately reflected LV function, showing positive correlations with LV end-diastolic and end-systolic diameters and negative correlations with LV fractional shortening and LVEF. Among echocardiographic parameters, the strongest correlation was found between BNP and LVMI. Peak PG had a significant positive correlation with BNP, whereas AVAI was negatively correlated with BNP. The catheterization data revealed that plasma BNP levels showed a significant negative correlation with LV end-diastolic volume, LV end-systolic volume, and LVEF, whereas the correlation was positive with pulmonary artery pressure and LV end-diastolic pressure. TMF parameters also showed significant correlations with BNP ( Table 3 ).
| r value | P value | |
|---|---|---|
| Echocardiography ( n = 109) | ||
| LVEDD (mm) | 0.41 | <.0001 |
| LVESD (mm) | 0.52 | <.0001 |
| LVFS (%) | −0.45 | <.0001 |
| LVEF (%) | −0.4 | <.0001 |
| LVMI (g/m 2 ) | 0.73 | <.0001 |
| Peak aortic PG (mm Hg) | 0.24 | <.01 |
| AVAI (cm 2 /m 2 ) | −0.26 | <.01 |
| Transmitral flow | ||
| E velocity (m/sec) | 0.37 | .001 |
| A velocity (m/sec) | −0.5 | <.0001 |
| E-Dct (msec) | −0.36 | .001 |
| E/A | 0.68 | <.0001 |
| Right heart catheterization ( n = 81) | ||
| Mean PCWP (mm Hg) | 0.1 | NS |
| Systolic PAP (mm Hg) | 0.25 | <.05 |
| Diastolic PAP (mm Hg) | 0.23 | <.05 |
| Mean PAP (mm Hg) | 0.23 | <.05 |
| LV catheterization ( n = 46) | ||
| LVSP (mm Hg) | 0.23 | NS |
| LVEDP (mm Hg) | 0.4 | <.0001 |
| LVEDVI (mL/m 2 ) | 0.46 | <.0001 |
| LVESVI (mL/m 2 ) | 0.61 | <.0001 |
| LVEF (%) | −0.5 | <.0001 |
BNP Levels and Patients’ Outcomes
Of the 95 patients who underwent AVR, 1 died in the hospital. Another 88 of these patients underwent echocardiography 2 weeks after AVR (13.2 ± 6.2 days after AVR), and 62 patients subsequently underwent repeat echocardiography (31 ± 12 months after AVR). Preoperative BNP level showed a strong correlation with LVMI, early and late after AVR ( Table 4 ). Patients with higher preoperative BNP remained symptomatic, despite receiving successful AVR ( Figure 2 ). Postoperative NYHA class was significantly correlated with preoperative BNP (130.6 ± 132.1 pg/mL, 420.2 ± 3653.8 pg/mL, 1502.5 ± 1010.9 pg/mL for NYHA I, II, and III, respectively, P < .0001). Patients with higher preoperative BNP had cardiac symptoms (NYHA > I) more frequently, even late after AVR (postoperative NYHA I vs. NYHA > I; 160.8 ± 197.9 pg/mL vs. 504.3 ± 567.3 pg/mL, P < .0001).
| r value | P value | |
|---|---|---|
| Early after AVR ( n = 82) | ||
| LVEDD (mm) | 0.4 | <.001 |
| LVESD (mm) | 0.38 | <.001 |
| LVFS (%) | −0.43 | <.0001 |
| LVEF (%) | −0.38 | <.001 |
| LVMI (g/m 2 ) | 0.72 | <.0001 |
| Late after AVR ( n = 62) | ||
| LVEDD (mm) | 0.28 | <.05 |
| LVESD (mm) | 0.29 | <.05 |
| LVFS (%) | −0.35 | <.01 |
| LVEF (%) | −0.31 | <.05 |
| LVMI (g/m 2 ) | 0.79 | <.0001 |
Atrial tachyarrhythmia was the most common complication, found in 34 patients, at a median of 2 days after surgery (92% atrial fibrillation, 8% atrial flutter). Seven patients had heart failure requiring intravenous diuretics and oxygenation in the early postoperative period (within 2 weeks). The patients with complications in the early postoperative period had higher levels of preoperative BNP than those with later complications (143.5 ± 157.5 pg/mL vs. 492.5 ± 563.5 pg/mL, P < .0001). We determined the appropriate cutoff value for the prediction of long-term outcome was > 312 pg/mL, as previously reported by Pedrazzini et al . Then, it gained a sensitivity of 83% and specificity of 66% for perioperative complication (AUC = 0.690) and a sensitivity of 78% and specificity of 62% for perioperative atrial tachyarrhythmias (AUC = 0.640). Table 5 shows both univariate and multivariate analyses for the prediction of complications: BNP > 312 pg/mL, severe LV hypertrophy (LVMI > 180 g/m 2 ), ejection fraction < 50%, and age more than 75 years were predictors of complications in the early postoperative periods. By multiple logistic regression analysis, BNP > 312 pg/mL was the only independent predictor of complications in AVR.
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| Variables | Odds ratio | 95% CI | P value | Odds ratio | 95% CI | P value |
| Age > 75 y (yes) | 6.26 | 1.62–24.09 | .08 | 2.74 | 0.84–9.84 | .101 |
| Female gender | 1.08 | 0.45–2.62 | .86 | |||
| NYHA > 2 | 1.83 | 0.72–4.69 | .2 | |||
| BAV | 0.86 | 0.36–2.10 | .73 | |||
| CAD | 0.91 | 0.31–2.65 | .86 | |||
| Diuretics | 2.14 | 0.89–5.15 | .09 | |||
| Diabetes (yes) | 1.22 | 0.43–3.51 | .7 | |||
| Hypertension (yes) | 1.01 | 0.43–2.36 | .98 | |||
| Dyslipidemia (yes) | 1.57 | 0.67–3.67 | .3 | |||
| LVEF < 50% | 4.17 | 1.03–16.84 | .04 | 1.99 | 0.48–8.92 | .343 |
| LVMI > 180 g/m 2 | 4.14 | 1.31–12.9 | .015 | 2.19 | 0.71–6.83 | .172 |
| BNP > 312 pg/mL | 7.79 | 2.69–26.44 | <.0001 | 5.58 | 1.82–20.16 | .002 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
