B-type natriuretic peptide (BNP) levels have shown a correlation with outcomes in studies of aortic valve surgery. Results from multicenter trials of BNP in transcatheter aortic valve surgery (TAVR) are lacking. The aim of this study was to investigate the prognostic role of serial measurement of BNP in transfemoral TAVR. A total of 1,097 patients who underwent TAVR via transfemoral access were analyzed by tertile of baseline BNP. Of those, 933 with BNP levels at 30 days were divided into 2 groups on the basis of increases (334 patients) or decreases or no change (599 patients) in BNP compared with baseline. Patients in the low-tertile BNP group had a lower rate of death at 1 year than those in the higher tertile group (15.0% vs 23.0%, p <0.01) which was not significant in multivariate analysis. Over 1 year, BNP decreased from 1,258.13 ± 2,988.33 to 594.37 ± 1,087.30 (p <0.01) in the entire group. Patients in the BNP-rise group had higher rates of death at 1 year (20.3% vs 11.4%, p <0.01) and an overall increase in moderate or severe aortic regurgitation over 1 year (p <0.01). Multivariate predictors of 1-year mortality were moderate or severe aortic regurgitation (hazard ratio 2.04, 95% confidence interval 1.36 to 3.05, p <0.01), increase in BNP at 30 days (hazard ratio 1.82, 95% confidence interval 1.26 to 2.62, p <0.01) and Society of Thoracic Surgeons score (hazard ratio 1.05, 95% confidence interval 1.01 to 1.10, p = 0.03). In conclusion, increase in BNP at 30 days from baseline and moderate or severe aortic regurgitation at 30 days in patients who undergo transfemoral TAVR are independently associated with 1-year mortality. Increase in BNP at 30 days should prompt evaluation for causes of elevated wall stress, including aortic regurgitation.
Aortic regurgitation (AR) has been shown to have late effects on mortality in patients who undergo transcatheter aortic valve replacement (TAVR). Although relief of the outlet obstruction with TAVR in aortic stenosis (AS) would be expected to reduce the systolic wall stress of the left ventricle, significant AR and mitral regurgitation (MR) would continue to cause strain due to volume overload. Therefore, biochemical assays that reflect this elevation in wall tension may be useful. BNP is a hormone that is released mainly by the cardiac myocytes of the left ventricle in response to increasing wall tension. It has been shown to be elevated in patients with AS and to correlate with the early onset of symptoms and to decrease after successful surgical aortic valve replacement. Early data in balloon aortic valvuloplasty and TAVR have also suggested decreased levels of natriuretic peptide after relief of the outlet obstruction. The prognostic implications of serial changes in B-type natriuretic peptide (BNP) after TAVR have not yet been studied in large, multicenter randomized trials. For this reason, we sought to evaluate the importance of baseline and serial BNP measurements with respect to outcomes in the transfemoral (TF) cohort in the Placement of Aortic Transcatheter Valves (PARTNER) I trial.
Methods
The design and results of the PARTNER trial have been previously described. Briefly, the trial enrolled patients with symptomatic, severe AS, defined as an aortic valve area ≤0.8 cm 2 plus a peak velocity ≥4 m/s or a mean valve gradient ≥40 mm Hg, into 2 cohorts, A and B. Cohort B randomized patients who were deemed inoperable by 2 cardiac surgeons to TAVR using a TF approach or best medical therapy. Cohort A randomized patients who were at high risk for surgery to surgical aortic valve replacement or TAVR. In this group, the TAVR procedure was performed by either the TF or transapical approach depending on the suitability of the iliofemoral arterial access. In all cases, TAVR was performed with the Edwards SAPIEN transcatheter heart valve (Edwards Lifesciences, Irvine, California). Additional patients underwent TAVR by either the TF or transapical approach in a continued-access registry with identical inclusion and exclusion criteria. The study was approved by the institutional review board at each participating site, and all patients provided written informed consent.
Patients were recruited from cohort A, cohort B, and the continued access registry. We included only those patients who underwent TAVR using a TF approach. In addition to basic laboratory testing, BNP levels were measured before TAVR, at discharge from the hospital, and at 30 days, 6 months, and 1 year. For this analysis, those patients who had N-terminal pro-BNP (NT-proBNP) levels measured were excluded. The individual assay used to measure BNP was left to the discretion of the study site, although the individual assays were consistent throughout the time points measured. Standard 2-dimensional echocardiography was performed before TAVR and at follow-up as part of the study protocol. Measurements included left ventricular (LV) systolic function, volumes, and diameters; assessment of aortic valve hemodynamic status (gradients and area); and degrees of AR and MR. LV mass was measured in grams according to the guidelines of the American Society of Echocardiography. All images were analyzed by an independent echocardiographic core laboratory (Duke University Medical Center, Durham, North Carolina).
All patients who underwent TAVR using a TF approach with baseline BNP levels were analyzed. In addition, patients who had BNP levels at both baseline and at 30 days were divided into 2 groups. Group I consisted of those patients who had increases in BNP of ≥1 pg/ml from baseline (BNP-rise), and group II consisted of those patients who had no change or decreases in BNP of ≥1 pg/ml (BNP-fall). We compared the rate of adverse events between these 2 groups, including death, death and rehospitalization, and degrees of AR and MR after TAVR. Categorical variables were compared using Fisher’s exact test. Continuous variables, which are presented as mean ± SD, were compared using Student’s t test. Survival curves for time-to-event variables were constructed using Kaplan-Meier estimates and were compared using the log-rank test, and hazard ratio estimates correspond to Cox proportional-hazards modeling. The following variables were included in the multivariate analysis of all-cause death: age, female gender, body mass index, Society of Thoracic Surgeons risk score, diabetes, cerebrovascular disease, renal disease (creatinine ≥2 [177 μmol/L]), stroke or transient ischemic attack (in the past 6 to 12 months), baseline ejection fraction, moderate or severe AR at 30 days, moderate or severe MR at 30 days, baseline BNP, and increase in BNP at 30 days. A 2-sided α level of 0.05 was used for all testing. All statistical analyses were performed with SAS version 9.2 (SAS Institute Inc., Cary, North Carolina).
Results
A total of 2,601 patients enrolled in the PARTNER study from May 2007 to December 2011 had baseline BNP measured ( Figure 1 ). Of those, 1,097 patients who underwent TAVR through a TF approach were included for the baseline BNP analysis ( Table 1 ). 933 patients had both baseline and 30-day BNP measurements. Patients without 30-day BNP measurements had a higher median baseline BNP level, higher Society of Thoracic Surgeons scores, higher rates of previous pacemaker placement, and lower aortic valve mean gradients and peak velocities. In addition, those patients without 30-day BNP measurements had higher rates of major bleeding and all-cause death at 1 year.
| Variable | All Patients (n=1097) | Patients without 30-day BNP (n=164) | Patients with 30-day BNP (n=933) | P-Value |
|---|---|---|---|---|
| B-Type Natriuretic Peptide (pg/ml) baseline (mean) | 1332.76 ± 3068.68 (1097) | 1757.3 ± 3470.4 (164) | 1258.1 ± 2988.3 (933) | 0.08 |
| B-Type Natriuretic Peptide (pg/ml) baseline (median) | 527.00 [251.20, 1131.80] | 603.5 [321.5, 1811.5] | 527.0 [251.2, 1131.8] | 0.03 |
| Age (yrs) | 84.61 ± 7.57 (1097) | 85.3 ± 6.7 (164) | 84.5 ± 7.7 (933) | 0.19 |
| Men | 625 (57) | 96 (58.5) | 529 (56.7) | 0.66 |
| White | 1014 (92.4) | 152 (92.7) | 862 (92.4) | 0.90 |
| STS Score | 10.97 ± 3.96 (1095) | 12.1 ± 4.6 (164) | 10.8 ± 3.8 (931) | <0.01 |
| Logistic EuroSCORE | 25.39 ± 15.25 (1074) | 27.7 ± 17.1 (161) | 25.0 ± 14.9 (913) | 0.06 |
| New York Heart Association Class 3&4 | 1049 (95.6) | 161(98.2) | 888 (95.2) | 0.08 |
| Any Diabetes | 417 (38.1) | 64 (39) | 353 (37.9) | 0.79 |
| Renal disease (creatinine ≥ 2mg/dl) | 163 (14.9) | 30 (18.3) | 133 (14.3) | 0.19 |
| Hyperlipidemia | 919 (83.9) | 137 (83.5) | 782 (84) | 0.88 |
| Smoking | 463 (42.3) | 64 (39) | 399 (42.9) | 0.36 |
| Hypertension | 997 (91.1) | 151 (92.1) | 846 (90.9) | 0.62 |
| Chronic Obstructive Pulmonary Disease | 489 (44.6) | 80 (48.8) | 409 (43.8) | 0.24 |
| Coronary Artery Disease | 820 (74.9) | 121 (73.8) | 699 (75.1) | 0.72 |
| Prior Myocardial Infarction | 256 (23.6) | 4 (2.4) | 10 (1.1) | 0.14 |
| Prior Percutanous Coronary Intervention | 394 (36.1) | 51 (31.1) | 343 (37) | 0.15 |
| Prior Coronary Artery Bypass | 412 (37.7) | 51 (31.1) | 361 (38.8) | 0.06 |
| Stroke or Transient Ischemic Attack (last 6-12 months) | 225 (21.1) | 39 (24.2) | 186 (20.5) | 0.29 |
| Peripheral Vascular Disease | 307 (28.4) | 43 (26.4) | 264 (28.7) | 0.54 |
| Prior Balloon Aortic Valvuloplasty | 250 (23) | 41 (25.2) | 209 (22.6) | 0.48 |
| Permanent Pacemaker | 256 (23.4) | 48 (29.3) | 208 (22.3) | 0.05 |
| Echocardiographic Variable | ||||
| Left Ventricular Ejection Fraction (%) | 52.62 ± 13.18 (1080) | 51.5 ± 14.0 (160) | 52.8 ± 13.0 (920) | 0.27 |
| Left Ventricular End Diastolic Volume (ml) | 134.84 ± 52.61 (514) | 132.1 ± 62.5 (81) | 135.4 ± 50.6 (433) | 0.66 |
| Left Ventricular End Systolic Volume (ml) | 69.38 ± 44.81 (514) | 71.2 ± 53.0 (81) | 69.0 ± 43.2 (433) | 0.73 |
| Stroke Volume (ml) | 65.47 ± 20.74 (514) | 60.9 ± 22.5 (81) | 66.3 ± 20.3 (433) | 0.03 |
| Left Ventricular Mass (gm) | 255.85 ± 78.01 (989) | 263.5 ± 79.0 (143) | 254.6 ± 77.8 (846) | 0.20 |
| Aortic Valve Area (AVA)cm 2 | 0.66 ± 0.19 (1052) | 0.6 ± 0.2 (155) | 0.7 ± 0.2 (897) | 0.41 |
| Aortic Valve Peak Velocity (cm/s) | 415.16 ± 64.16 (1064) | 403.1 ± 70.9 (158) | 417.3 ± 62.7 (906) | 0.02 |
| Aortic Valve Mean Gradient (mmHg) | 43.56 ± 14.24 (1072) | 41.0 ± 15.3 (159) | 44.0 ± 14.0 (913) | 0.01 |
| Aortic Regurgitation (≥Moderate) | 122 (11.3) | 21 (13) | 101 (11) | 0.46 |
| Mitral Regurgitation (≥Moderate) | 241 (22.5) | 35 (22.2) | 206 (22.5) | 0.92 |
| Major Events at 1-year | ||||
| Death from any cause | 224 (20.6) | 90 (55.5) | 134 (14.6) | <0.01 |
| Myocardial Infarction | 14 (1.4) | 4 (2.4) | 10 (1.1) | 0.14 |
| Major Bleeding | 155 (14.8) | 48 (34.6) | 107 (11.9) | <0.01 |
Kaplan-Meier curves according to mean tertile of baseline BNP are shown in Figure 2 . The rate of death at 1 year was lower for the low- versus intermediate-tertile group (15.0% vs 23.9%, p <0.01) and the low- versus high-tertile group (15.0% vs 23.0%, p <0.01). Rates of death were similar between the intermediate- and high-tertile groups.
The baseline characteristics for the BNP-rise (n = 334) and BNP-fall (n = 599) groups are listed in Table 2 . Baseline characteristics were similar overall, including age, rates of coronary artery disease, and ratios of men. Patients in the BNP-fall group had a lower prevalence of creatinine ≥2 mg/dl (17.4% vs 12.6%, p = 0.04). Patients who had increases in BNP had a lower baseline rate of moderate or severe AR (7.4% vs 13.0%, p < 0.01). Rates of baseline moderate or severe MR and values for LV mass, LV end-diastolic volume, and LV end-systolic volume were similar.
| Variable | BNP-rise (n=334) | BNP Fall/No Change (N=599) | P-Value |
|---|---|---|---|
| B-Type Natriuretic Peptide (pg/ml) baseline (mean) | 663.59 ± 1124.24 (334) | 1589.64 ± 3592.57 (599) | <0.01 |
| B-Type Natriuretic Peptide (pg/ml) baseline (median) | 338.50 [170.30, 716.00] | 690.00 [334.00, 1352.00] | <0.01 |
| B-Type Natriuretic Peptide at 30-days pg/ml (median) | 1319.53 ± 3061.45 | 645.04 ± 1153.53 | <0.01 |
| B-Type Natriuretic Peptide at 30-days pg/ml (median) | 584.50 [286.00, 1121.00] | 328.00 [177.90, 665.00] | <0.01 |
| Age (years) | 84.29 ± 7.43 (334) | 84.61 ± 7.88 (599) | 0.55 |
| Men | 202 (60.5) | 327 (54.6) | 0.08 |
| White | 309 (92.5) | 553 (92.3) | 0.91 |
| STS Score | 10.52 ± 3.84 (334) | 10.91 ± 3.79 (597) | 0.13 |
| Logistic EuroScore | 24.06 ± 13.98 (325) | 25.50 ± 15.32 (588) | 0.15 |
| Any Diabetes | 133 (39.9) | 220 (36.8) | 0.34 |
| Renal disease (creatinine ≥ 2mg/dl) | 58 (17.4) | 75 (12.6) | 0.04 |
| Hyperlipidemia | 289 (86.8) | 493 (82.4) | 0.08 |
| Smoking | 144 (43.2) | 255 (42.6) | 0.86 |
| Hypertension | 306 (91.9) | 540 (90.3) | 0.42 |
| Coronary Artery Disease | 245 (73.6) | 454 (75.9) | 0.43 |
| Prior Myocardial Infarction | 75 (22.7) | 150 (25.3) | 0.38 |
| Prior Percutaneous Coronary Intervention | 122 (37) | 221 (37) | 1.00 |
| Prior Coronary Artery Bypass | 139 (41.7) | 222 (37.2) | 0.17 |
| Stroke or Transient Ischemic Attack (last 6-12 months) | 69 (21.2) | 117 (20.1) | 0.71 |
| Peripheral Vascular disease | 93 (28.1) | 171 (29.1) | 0.75 |
| Porcelain aorta | 15 (4.5) | 24 (4) | 0.72 |
| Prior Balloon Aortic Valvuloplasty | 60 (18.1) | 149 (25.1) | 0.01 |
| Echocardiographic Variable | |||
| Left Ventricular Ejection Fraction (%) | 54.69 ± 11.50 (329) | 51.75 ± 13.72 (591) | <0.01 |
| Left Ventricular End Diastolic Volume (ml) | 133.05 ± 52.88 (145) | 136.51 ± 49.48 (288) | 0.50 |
| Left Ventricular End Systolic Volume (ml) | 64.89 ± 42.92 (145) | 71.12 ± 43.22 (288) | 0.16 |
| Stroke Volume (ml) | 68.15 ± 20.36 (145) | 65.40 ± 20.25 (288) | 0.18 |
| Left Ventricular Mass (gm) | 252.02 ± 78.27 (299) | 255.94 ± 77.60 (547) | 0.48 |
| Aortic Valve Area (cm 2 ) | 0.70 ± 0.19 (313) | 0.64 ± 0.18 (567) | <0.01 |
| Aortic Valve Annulus Diameter (cm) | 1.96 ± 0.29 (254) | 1.91 ± 0.27 (483) | 0.01 |
| Aortic ValvePeak Velocity (m/s) | 4.04 ± .58 (318) | 4.25 ± .64 (588) | <0.01 |
| Aortic Valve Mean Gradient (mmHg) | 41.02 ± 12.53 (322) | 45.64 ± 14.51 (591) | <0.01 |
| Aortic Regurgitation (≥Moderate) | 24 (7.4) | 77 (13) | <0.01 |
| Mitral Regurgitation (≥Moderate) | 65 (20) | 141 (23.9) | 0.18 |
| Aortic Valve Size (mm) | 24.49 ± 1.5 (329) | 24.43 ± 1.5 (585) | 0.59 |
| Procedural Events | |||
| Migration or Embolization | 5 (3.6) | 1 (0.4) | 0.03 |
| Intraaortic Balloon Pump during procedure | 8 (2.4) | 0 (0) | <0.01 |
| Cardiopulmonary Bypass | 2 (0.6) | 5 (0.8) | 1.00 |
Valve sizes in the 2 groups were similar. There was an overall low, but slightly higher rate of intra-aortic balloon pump use (2.4% vs 0.0%, p <0.01) and migration or embolization of the valve (3.6% vs 0.4%, p = 0.03) in the BNP-rise group. The rates of cardiopulmonary bypass use between the groups were similar.
Figure 3 shows the impact of TAVR on levels of BNP in the 2 groups. For the 2 groups combined, there was a significant decrease in BNP from baseline to 1 year (1,258.13 ± 2,988.33 vs 594.37 ± 1,087.30, p <0.01). After an abrupt decrease in BNP at 30 days, levels stabilized out to 1 year for the BNP-fall group, whereas the BNP-rise group had an overall trend of elevated BNP to 1 year. The mean ejection fraction of the BNP-rise group at 30 days was slightly higher (55.96 ± 10.03% vs 54.51 ± 11.35%, p = 0.05), and mean aortic gradients were similar in the 2 groups (17.83 ± 7.75 vs 18.28 ± 7.63 mm Hg, p = 0.40). Rates of moderate or severe AR were higher in the BNP-rise group compared with the BNP-fall group at 30 days (20.9% vs 12.3%, p <0.01) and 6 months (18.9% vs 12.0%, p <0.01), and there was an overall trend of increased rates of AR from baseline in this group ( Figure 3 ). Although rates of moderate or severe AR were higher at baseline in the BNP-fall group, there was no overall change in these rates of AR in this group out to 1 year after TAVR. Rates of moderate or severe MR were similar at baseline but higher in the BNP-rise group at 30 days (24.0% vs 18.3%, p = 0.04) and at 6 months (23.9% vs 15.4%, p <0.01). There was an overall decrease in the severity of moderate or severe MR in the BNP-fall group ( Figure 3 ).
