We sought to study the prognostic utility of serum brain natriuretic peptide (BNP) in patients with significant primary mitral regurgitation (MR) and preserved left ventricular (LV) ejection fraction (EF). Consecutive 548 asymptomatic patients (age 62 ± 13 years and 66% men) with ≥3 + primary MR and preserved LVEF on echo at rest, evaluated at our center from 2005 to 2008 were studied. Baseline clinical and echo data were recorded and the Society of Thoracic Surgeons (STS) score was calculated. Mean STS score was 4 ± 1%. Mean LVEF, mitral effective regurgitant orifice, indexed LV end-systolic diameter, and right ventricular systolic pressure (RVSP) were 62 ± 4%, 0.55 ± 0.3 cm 2 , 1.6 ± 0.3 cm/m 2 , and 38 ± 15 mm Hg; 43% had flail. Median log-transformed brain natriuretic peptide (lnBNP) was 4.1 (interquartile range 3.30 to 5.0), corresponding to an absolute BNP value of 60 pg/ml (only 13% had an absolute BNP value >250 pg/ml). At 7.4 ± 2 years, 493 patients (90%) had mitral surgery (92% repair) and nonmalignancy death occurred in 53 patients (10%). On multivariate Cox analysis, higher STS score (hazard ratio [HR] 1.50, 95% CI 1.20 to 1.88), higher baseline RVSP (HR 1.17, 95% CI 1.02 to 1.35), and higher ln BNP (HR 2.51, 95% CI 1.86 to 3.39) predicted death, whereas mitral surgery (HR 0.17, 95% CI 0.09 to 0.30) was associated with improved survival (all p <0.01). Eighty-nine percent of deaths occurred in patients with lnBNP >4.1. Addition of lnBNP to a model of STS score, baseline RVSP, and mitral surgery provided incremental prognostic utility (chi-square for mortality increased from 137 to 162, p <0.001). In conclusion, in asymptomatic patients with ≥3 + primary MR and preserved LVEF, the addition of BNP improved risk stratification and higher BNP independently predicted reduced survival.
Brain natriuretic peptide (BNP) is a hormone secreted from myocardial cells in response to either diastolic stretch indicating volume overload, or wall stress indicating pressure overload, and is a marker of LV dysfunction It is stable, generally unaffected by minute-to-minute daily activity–related perturbations. Multiple previous reports have demonstrated incremental utility of BNP in patients with primary mitral regurgitation (MR). Most of these studies had a small sample size and relatively short follow-up. In addition, these studies reported end points such as development of symptoms/LV dysfunction, and/or need for mitral valve (MV) surgery (rather than survival). In this study, we sought to determine whether serum BNP prospectively measured at the time of initial evaluation is a predictor of long-term mortality in a large group of asymptomatic patients with significant asymptomatic primary MR and preserved LVEF, many of whom went on to surgical intervention.
Methods
This was a retrospective observational cohort study of 548 consecutive asymptomatic (or those with minimal atypical symptoms) patients with ≥III + primary MR and relatively preserved left ventricular ejection fraction (LVEF) (all >55%, 93% patients with LVEF ≥60%) who were seen and evaluated in our institution from 2005 to 2008. To be included in the study, all patients had a comprehensive echocardiogram and BNP level measured within 30 days of each other (>90% on the same day). We excluded the following patients: any degree of concomitant aortic or mitral stenosis, above moderate aortic regurgitation, history of hypertrophic obstructive cardiomyopathy, infective endocarditis, and previous MV surgery. The present study was approved by the institutional review board. Baseline demographics, clinical status, and echocardiographic parameters were electronically recorded prospectively at the time of the initial encounter. For the present study, all data were manually extracted from the electronic medical records. The presence of atrial fibrillation was recorded on the basis of history and electrocardiographic data. We recorded the type of MV surgery (repair vs replacement) along with concomitant procedure (coronary artery bypass grafting, MAZE, pulmonary vein isolation, or left atrial appendage ligation/excision). The Society of Thoracic Surgeons (STS) score was calculated in all patients.
For BNP assay, all blood samples were collected into ethylenediaminetetraacetic acid Vacutainer tubes and analyzed according to standard clinical laboratory routine. Plasma BNP (pg/ml) was determined by chemiluminescence immunoassay on site (Biosite Diagnostics, San Diego, CA).
All patients underwent comprehensive echocardiograms using commercial instruments (Philips Medical Systems, Bothell, Washington; Siemens Medical Solution Inc, Malvern, Pennsylvania; and General Electric, Milwaukee, WA). LVEF, indexed LV dimensions, and left atrial area were measured according to the guidelines. The severity of MR was ascertained using multiple previously described techniques. This included qualitative visual assessment (as a percentage of left atrial size) and measurement of effective regurgitant orifice area and vena contracta width. Because of the severity of MR, diastolic function was not reported. The presence of flail mitral leaflet was recorded. Right ventricular systolic function was measured qualitatively (normal, mild, moderate, or severe). Right ventricular systolic pressure (RVSP) was measured at rest.
The date of the patient’s baseline echocardiography at our institution was defined as the beginning of the observational period. Follow-up was ascertained by chart review, and we recorded the date at which events occurred. Mortality data were obtained from review of medical records or state and nationally available databases (last inquiry in September 2014). Primary outcome was nonmalignancy death. For the current survival analysis, we excluded 3 patients who died because of a documented malignancy that was diagnosed during follow-up. However, these 3 patients were censored at the time of their death. There were no patients with additional noncardiac causes of death (e.g., renal/liver/neurologic issues).
Continuous variables are expressed as mean ± SD, or median and interquartiles for skewed distributions, and compared using the Student t test or analysis of variance (for normally distributed variables) or the Mann–Whitney test (for nonnormally distributed variables). Categorical data are expressed as percentage and compared using the chi-square test or Fisher’s exact test, as appropriate. Because the distribution of BNP was skewed, values were logarithmically transformed for analysis. Correlation between continuous variables was assessed using the Spearman’s correlation coefficient. To assess outcomes, Cox proportional hazards analysis was performed. We created a parsimonious model in which prespecified relevant variables, associated with adverse outcomes in patients with primary MR, were included. MV surgery was included as a time-dependent covariate in Cox survival analysis. For each patient who underwent MV surgery, the analysis time was modeled so that only the person-time after MV surgery was included in the surgical group. The person-time before the occurrence of MV surgery was included in the nonsurgical category. Although STS score has only been validated to predict 30-day postoperative mortality, we used it in the Cox survival analysis, as it is a composite of many factors that are known to be associated with adverse postoperative events in the longer term. Hazard ratios (HRs) with 95% CIs were calculated and are reported. In addition, cumulative proportion of events as a function over time was obtained by the Kaplan–Meier method, and event curves were compared using a log-rank test in which proportional hazards were not violated and a generalized Wilcoxon (Breslow’s) test in which the survival curves clearly cross and the proportional hazards were violated. Additionally, we assessed the classification of risk using net reclassification improvement. Statistical analysis was performed using SPSS, version 11.5, (IBM Corp, Armonk NY) and Stata, version 10.0, (StataCorp, College Station TX). A p value <0.05 was considered significant.
Results
Baseline and echocardiographic characteristics of the study sample are listed in Tables 1 and 2 . In the study sample, 222 patients (40%) had dilated LV and 442 patients (81%) had a dilated left atrium. Relevant characteristics of the study population divided into BNP quartiles are provided in Supplementary Table 1 . There was no correlation between BNP and baseline LVEF ( r = 0.001, p = 0.9) or time to surgery ( r = 0.03, p = 0.5).
| Variable | |
|---|---|
| Age (years) | 62±13 |
| Male gender | 366 (67%) males |
| Body mass index (kg/ m2) | 25± 4 |
| Hypertension | 140 (26%) |
| Diabetes mellitus | 13 (2%) |
| Obstructive coronary artery disease | 56 (11% |
| Atrial fibrillation | 127 (30%) |
| Hyperlipidemia | 122 (22%) |
| Smoker | 173 (32%) |
| Society of Thoracic Surgeons Score (%) | 4.0±1 |
| Angiotensin converting enzyme inhibitor treatment | 203 (37%) |
| Beta Blocker treatment | 220 (40%) |
| Oral anticoagulant treatment | 90 (16%) |
| Serum Creatinine (mg/ml) | 1±0.3 |
| Log-transformed brain natriuretic peptide | 4.2±1.2 |
| Serum hemoglobin (mg/dl) | 14.2±2 |
| Low density lipoprotein cholesterol (mg/dl) | 108±34 mg/dl |
| High density lipoprotein cholesterol (mg/dl) | 59±18 |
| Triglycerides (mg/dl) | 109±67 mg/dl |
| Variable | |
|---|---|
| LV ejection fraction (%) | 62±4 |
| Indexed LV end systolic diameter (cm/m2) | 1.6±0.3 |
| Indexed left atrial diameter (cm/m2) | 4.7± 0.8 |
| Mitral effective regurgitant orifice (cm2) | 0.55±0.3 cm2 |
| Mitral leaflet flail | 238 (43%) |
| Mitral valve prolapse | |
| Anterior | 70 (13%) |
| Posterior | 242 (44%) |
| Bileaflet | 236 (43%) |
| Right ventricular systolic pressure mm Hg | 38±15 mmHg |
| Tricuspid regurgitation | |
| None-trivial | 244 (44%) |
| Mild | 195 (36%) |
| Moderate | 70 (13%) |
| Moderate-severe | 36 (6%) |
| Severe | 3 (0.5%) |
| Right ventricular function | |
| Normal | 523 (95%) |
| Mildly reduced | 20 (4%) |
| Moderately reduced | 4 (0.5%) |
| Severely reduced | 2 (0.2%) |
In the present study, 493 patients (90% of the cohort) underwent MV surgery at a median time of 4 days (interquartile range 2 to 7 days) from the baseline echocardiogram. Each of these had at least a class IIa indication using American College of Cardiology/American Heart Association guidelines prevalent at the time of surgery. Of these, 454 patients (92%) were MV repairs and 39 patients (8%) were MV replacements (only 2 were mechanical valve prostheses). Additional procedures performed at the time of MV surgery were left atrial appendage ligation/excision in 137 patients (25%), maze procedure/pulmonary vein isolation in 114 patients (21%), tricuspid valve repair in 51 patients (9%), and coronary artery bypass grafting in 56 patients (10%). Patients who did not undergo MV surgery (n = 55) had a significantly lower STS score (3.4 ± 1 vs 4.1 ± 0.9, p <0.01), similar LVEF (62% vs 62%, p = 0.2) and similar effective regurgitant orifice area (0.57 ± 0.3 vs 0.55 ± 0.2 cm 2 , p = 0.7). The primary reason to not operate in these patients was the personal preference of the patient after a thorough clinical evaluation and detailed discussion with the attending cardiologist. No patient in this subgroup had a noncardiac condition precluding MV surgery.
During a mean follow-up of 7.4 ± 2 years, death was observed in 53 patients (10%). There were 3 additional deaths because of malignancy (censored at the time of event but excluded from survival analysis, as discussed previously). In the surgical group, there were no 30-day postoperative deaths. There were 7 patients (1.3%) who were admitted for management of congestive heart failure during long-term follow-up.
We subsequently performed survival analysis for the primary outcome using Cox proportional hazards method. Neither quadratic nor cubic transformations of log-transformed brain natriuretic peptide (lnBNP), RVSP, or STS score were significant predictors of outcomes when forced into the Cox model, which already included these variables in a nontransformed form. The results of univariate analysis, testing the relevant predictors associated with mortality in these patients, are provided in the Supplementary Table 2 . The results of multivariate analysis are listed in Tables 3 and 4 . As shown in Figure 1 , addition of lnBNP to STS score, baseline RVSP, and MV surgery provided incremental prognostic utility in these patients with significant increases in chi-square for mortality at every stage (chi-square increased from 37 to 80 to 137 to 162, p value for differences <0.001). The addition of lnBNP to STS score, mitral surgery, and RVSP resulted in significant net reclassification improvement (0.3 [0.17 to 0.44]). The presence of a flail MV was not associated with adverse outcomes on univariate Cox analysis (HR 1.07, 95% CI 0.61 to 1.88, p = 0.8).
| Hazard Ratio (95% confidence interval) | p-value | |
|---|---|---|
| Log-transformed brain natriuretic peptide (for every unit increase) | 2.51 (1.86-3.39) | <0.001 |
| Society of Thoracic Surgeons score (for every unit increase) | 1.50 (1.20-1.88) | <0.001 |
| Baseline right ventricular systolic pressure (for every mm Hg increase) | 1.17 (1.02- 1.35) | <0.001 |
| Mitral surgery (time-dependent covariate) | 0.17 (0.09- 0.30) | <0.001 |
| Hazard Ratio (95% confidence interval) | p-value | |
|---|---|---|
| Log transformed brain natriuretic peptide (for every unit increase) | 2.16 (1.57-2.00) | <0.001 |
| Baseline right ventricular systolic pressure (for every mm Hg increase) | 1.12 (1.02- 1.39) | <0.001 |
| Age (for every year increase) | 1.11 (1.07-1.15) | <0.001 |
| Mitral surgery (time-dependent covariate) | 0.26 (0.14- 0.49) | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
