Preoperative blood urea nitrogen-to-left ventricular ejection fraction ratio is an independent predictor of long-term major adverse cardiac events in patients undergoing coronary artery bypass grafting surgery.

BACKGROUND: Long-term mortality rate following coronary artery bypass grafting (CABG) procedure is still considered to be high despite advances in surgical techniques and perioperative management. Identifying high-risk patients by using cost-effective and clinically useful parameters is needed.
METHODS: Patients who were admitted to our cardiology clinic with the diagnosis of coronary artery disease and underwent CABG between January 2008 and August 2010 were included. Study patients were followed-up for 112.6 ± 17.8 months for major adverse cardiac events (MACE) which were defined as all-cause mortality and new-onset decompensated heart failure (HF).
RESULTS: Patients in MACE (+) group were older (p < 0.001), had higher additive Euroscore (p < 0.001), and lower left ventricular ejection fraction (p < 0.001). Multivariate Cox regression analysis showed that additive Euroscore [odds ratio (OR) = 1.601; 95% confidence interval (CI) = 1.374-1.864; p < 0.001)] and blood urea nitrogen-to-left ventricular ejection fraction ratio (BUNEFr; OR = 1.028; 95% CI = 1.006-1.050; p = 0.011) independently predicted MACE. Receiver operating characteristic curve analysis demonstrated that BUNEFr had an area under curve of 0.794 and BUNEFr >33 had a sensitivity and specificity of 74% and 64%, respectively.
CONCLUSION: BUNEFr is a clinically useful and cost-effective parameter for the prediction of long-term mortality and new-onset decompensated HF in patients undergoing CABG.

1. Introduction

Coronary artery bypass grafting (CABG) surgery still remains a treatment of choice for revascularization of diffuse coronary artery disease (CAD) and three-vessel disease involving the left anterior descending artery [1]. Despite improvements in periprocedural care, surgical techniques, and operator’s experience, in-hospital and long-term mortality rates following CABG are still relatively high. Identifying those with a high risk for mortality is of utmost importance for tailoring guideline-directed therapy and for stricter follow-up. Several risk scores are developed for predicting both short- and long-term mortality, but the ideal risk score still remains debatable.

Renal functions are well-known prognostic factor for CAD [2]. In daily practice, elevated serum creatinine (sCr) level and decreased glomerular filtration rate (GFR) are considered as worsening in renal functions. Blood urea nitrogen (BUN) is another blood marker that is associated with renal functions, but its sensitivity and specificity is less compared with sCr and GFR. However, in addition to reflecting GFR, BUN may rise independent of changes in GFR or sCr owing to enhanced proximal tubular reabsorption of urea under the activation of the sympathetic nervous and renin–angiotensin–aldosterone systems (RAAS) [3]. Thus, BUN may reflect neurohormonal dysregulation in patients with depressed left ventricular ejection fraction (LVEF) [4-6] and is found to be a predictor for long-term mortality following CABG [7].

In the present study, we investigated the predictive value of the ratio of BUN to LVEF (BUNEFr) for the occurrence of all-cause mortality and new-onset decompensated heart failure (HF) following CABG.

2. Methods

In this prospective, observational, cohort study, patients who were admitted to our cardiology clinic with the diagnosis of CAD between January 2008 and August 2010 were included. The diagnosis of CAD constituted stable CAD, unstable CAD, ST-segment elevation myocardial infarction, and non-ST segment elevation myocardial infarction. Only patients who underwent CABG procedure following initial admittance to the hospital and subsequent coronary artery angiography (CAG) were evaluated in the final analysis. The decision for CABG or percutaneous coronary intervention was made by the careful evaluation of the patients by cardiology and cardiovascular surgery teams. Informed consents were obtained from all of the participants of the study.

Study patients were followed-up for 112.6 ± 17.8 months for major adverse cardiac events (MACE) which were defined as all-cause mortality and new-onset decompensated HF. After surgery, all patients were followed for MACE during the postoperative period. Data for MACE were obtained from the national and institutional databases, hospital records, routine clinical visits, or by phone interviews with patients and their families. The study was performed in accordance with the principles stated in the Declaration of Helsinki. The local ethics committee approved the study protocol.

Clinical characteristics, which consisted of multiple descriptors from each patient’s medical history and physical examination, were collected by physicians from the cardiology clinic for each patient prior to CAG and were stored in the database of coronary angiography laboratory in our institution. We recorded the baseline characteristics, including hypertension, diabetes mellitus, smoking history, family history for CAD, and blood’s parameters. Preoperative BUN value, which was obtained from a sample prior to CABG surgery, was used in our study. Both echocardiography and BUN measurements were performed on the same day before CABG surgery.

Additive Euroscore of each patient was calculated using an online web calculator in the website http://www.euroscore.org/calc.html. Preoperative myocardial infarction (MI) and chronic obstructive pulmonary disease were diagnosed in accordance to the latest guideline documents [8,9]. All patients were evaluated by carotid artery Doppler ultrasonography for the detection of carotid artery stenosis. Extracardiac arthropathy was defined as limb claudication, carotid artery stenosis more than 50%, chronic total occlusion of any artery in extremities, previous history of amputation due to peripheral artery disease (PAD), and previous history of intervention to abdominal aorta, extremities, and carotid arteries for the treatment of PAD.

Exclusion criteria were patients with GFR <30 mL/ min/1.73 m2, those on chronic dialysis, more than moderate valvular heart disease, re-do CABG cases, chronic liver disease, active infection, inflammatory diseases, malignancy, and emergent CABG procedures.

Standard transthoracic and Doppler echocardiographic examinations were performed using a 3.25-MHz transthoracic transducer connected to a Vivid 5 System (GE Vingmed Ultrasound AS, Horten, Norway). Two echocardiographers who were unaware of the study performed the examinations, and they were blinded to the echocardiograms and clinical status of each patient. Left ventricular end-systolic dimension, end-diastolic dimension, wall thickness, and left atrial volume were measured according to the guidelines of the American Society of Echocardiography [10]. Left ventricular end-systolic and end-diastolic volumes and ejection fraction were measured from the apical four- and two-chamber views using the modified Simpson method.

2.1. Statistical analysis

Continuous variables were presented as mean values (standard deviation) or medians with ranges, and the categorical variables were expressed as percentages. The variables were compared using a two-tailed Student t test for the continuous variables of normal distribution or the Mann–Whitney U test for the continuous variables of non-normal distribution. A Chi-square test was used for categorical variables. The effects of various variables on MACE were calculated by univariate regression analysis. In these analyses, the variables with unadjusted p < 0.1 were identified as confounding factors and included in the multivariate regression analyses to determine the independent predictors of MACE. The predictive values of BUN, LVEF, and BUNEFr were estimated by areas under the receiver operating characteristic (ROC) curve. We used the DeLong test to compare the area under the curve (AUC) with each of these parameters [11]. All statistical tests were two-tailed, and a p value <0.05 value was considered significant. All analyses were performed using SPSS version 16 (SPSS Inc., Chicago, IL, USA).

3. Results

Between January 2008 and August 2010, 948 patients were admitted to the hospital with the diagnosis of CAD. Among those, 304 patients had undergone CABG surgery, 32 had acute or chronic kidney disease, 30 had more than moderate valvular diseases, four had re-do CABG, five had chronic liver diseases, 12 had active infection, three had inflammatory diseases, four had malignancy, and in 14 patients CABG surgery was performed in emergency settings. These patients were excluded from the study. The remaining 202 patients constituted the study population. The mean age of patients was 61.2 ± 9.8 years and 38 (18.8%) were female. During the follow-up, 64 (31.6%) patients had MACE including 48 (23.7%) died and 16 (7.9%) developed new-onset HF.

Patients were divided into two groups according to MACE. Patients in the MACE (+) group were older (p < 0.001), had higher additive Euroscore (p < 0.001) and lower LVEF (p < 0.001). No significant difference was noted among patients regarding hypertension, diabetes, dyslipidemia, sex, smoke status, and body mass index. The number of patients with coronary artery disease with >50% stenosis was higher in the MACE (+) group (p = 0.02), and patients who had severe carotid artery stenosis were more likely to develop MACE (p = 0.01). No difference was noted for the presence of stenosis in left main stem. Revascularization using left internal mammary artery was associated with reduced MACE; however, this did not reach statistical significance (p = 0.082; Table 1).

Table 1

Baseline characteristics of the study population.

Variable	MACE (−)(n = 138)	MACE (+)(n = 64)	All patients(n = 202)	p
Age (y)	58.8 ± 8.4	65.3 ± 9.7	61.2 ± 9.8	<0.001
Male sex (%)	82.6	79.6	81.6	0.620
Diabetes mellitus (%)	49.2	37.5	48.7	0.119
Hyperlipidemia (%)	75.9	78.1	76.6	0.731
Hypertension (%)	58.7	68.7	46.7	0.173
Current smoking (%)	45.6	51.5	50.3	0.436
BMI (kg/m2)	28.6	281.1	28.5	0.341
Euroscore	2.8 ± 1.5	4.6 ± 1.8	2.95 ± 1.8	<0.001
Carotid stenosis (%)	2.27	10.9	5.1	0.010
COPD (%)	15.2	20.3	16.8	0.373
Diagnosis (%)
SAP	65.5	56.9	58
USAP/NSTEMI	26.8	34.1	33.9	0.392
STEMI	7.5	8.8	8
LVEF (%)	54.9	44.4	51.6	<0.001
LMCA >%50	10.4	12.6	11.2	0.632
Number of CAD	2.4 ± 0.7	2.6 ± 0.64	2.5 ± 0.68	0.024
Glucose (Fasting)	122 ± 35.6	133.4 ± 56	130.1 ± 51	0,038
BUN (mg/dL)	16.1 ± 5.3	18.8 ± 5.5	17.1 ± 5.4	0.001
BUNEFr	31.08 ± 12	45.5 ± 19	35.7 ± 16.1	<0.001
Serum creatinine (mg/dL)	0.96 ± 0.21	1.02 ± 0.21	0.98 ± 0.21	0.065
Total cholesterol (mg/dL)	177 ± 44.8	174.6 ± 45.4	176.8 ± 45	0.621
LDLc (mg/dL)	111.2 ± 37	114.8 ± 35.2	112.6 ± 36.7	0.592
HDLc (mg/dL)	35.5 ± 8.8	35.6 ± 9.4	35.5 ± 8.9	0.951
WBC (103/μm3)	7.9 ± 2	7.9 ± 2.1	7.9 ± 2.03	0.964
Neutrophil (103/μm3)	4.9 ± 1.8	5.03 ± 1.9	4.9 ± 1.8	0.721
Hb (g/dL)	13.5 ± 1.3	13.1 ± 1.4	13.4 ± 1.3	0.020
ASA (dis.) (%)	51.7	59.6	56.3	0.332
Clopidogrel (dis.) (%)	10.8	15.3	12.2	0.405
ß-Blocker (dis.) (%)	71.9	76.9	73.4	0.496
ACE/ARB (dis.) (%)	42.5	50	44.9	0.334
CCB (dis.) (%)	7.8	6.4	7.4	0.571
OAD/Insulin (dis.) (%)	24	18.9	22.6	0.531
Statin (dis.) (%)	70.8	61.29	67.7	0.188
Mortality, n (%)	0	48 (75)	48 (23.7)	<0.001
New onset HF, n (%)	0	16 (25)	16 (7.9)	<0.001
MACE, n (%)	0	64 (100)	64 (31.6)	<0.001

Data are presented as % or mean ± SD.

ACE = angiotensin converting enzyme; ARB = angiotensin-reception blocker; ASA = acetylsalicylic acid; BMI = body mass index; BUN = blood urea nitrogen; BUNEFr = blood urea nitrogen-to-left ventricular ejection fraction ratio; CAD = coronary artery disease; CCB = calcium channel blocker; COPD = chronic obstructive pulmonary disease; Dis = discharge; Hb = hemoglobin; HDLc = high-density lipoprotein cholesterol; HF = heart failure; LDLc = low-density lipoprotein cholesterol; LMCA = left main coronary artery; LVEF = left ventricular ejection fraction; MACE = major cardiac adverse events; NSTEMI = non-ST elevation myocardial infarction; OAD = oral antidiabetic; SAP=stable angina pectoris; STEMI=STelevationmyocardial infarction; USAP=unstable anginapectoris; WBC=whitebloodcell.

The sCr level was higher in patients with MACE (+), but this also did not reach statistical significance (p = 0.065). By contrast, both BUN (p < 0.001) and BUNEFr (p < 0.001) were significantly higher in MACE (+) group (Table 1).

We performed correlation analysis for BUNEFr which revealed that age (r = 0.269, p < 0.001), additive Euroscore (r = 0.486, p < 0.001), hypertension (r = 0.164, p = 0.021), diabetes (r = 0.145, p = 0.042), diffuse CAD (r = 0.232, p = 0.002), sCR (r = 0.324, p < 0.001), and hemoglobin (Hb; r =−0.212, p = 0.003) levels were significantly correlated with BUNEFr.

Univariate Cox analysis showed that additive Euroscore (p < 0.001), number of CAD with >50% stenosis (p = 0.019), BUN (p = 0.003), BUNEFr (p < 0.001), and Hb (p = 0.010) levels were associated with MACE (Table 2).

Table 2

Cox regression analysis.

Variables	Univariate analysis			Multivariate analysis
	OR	95% CI	p	OR	95% CI	p
Euroscore I	1.480	1.351–1.620	<0.001	1.601	1.374–1.864	<0.001
Number of coronary arteries with significant stenosis	1.770	1.100–2.848	0.019	1.171	0.688–1.994	0.560
BUN	1.059	1.020–1.100	0.003	1.003	0.933–1.079	0.931
Glucose	1.013	1.003–1.028	0.018
LVEF	0.948	0.929–0.968	<0.001
BUNEFr	1.043	1.031–1.054	<0.001	1.036	1.019–1.054	<0.001
Serum creatinine	2.734	0.967–7.730	0.058
Hemoglobin	0.777	0.641–0.943	0.010	1.086	0.869–1.357	0.467

BUN = blood urea nitrogen; BUNEFr = blood urea nitrogen-to-left ventricular ejection fraction ratio; CI = confidence interval; LVEF = left ventricular ejection fraction; OR = odds ratio.

We performed multivariate Cox regression analysis to identify independent predictors for MACE during the follow-up. Additive Euroscore [odds ratio (OR) = 1.601; 95% confidence interval (CI) = 1.374–1.864; p < 0.001)] and BUNEFr (OR = 1.028; 95% CI = 1.006–1.050; p = 0.011) were found to independently predict MACE (Table 2).

The ROC curve analysis demonstrated that BUNEFr, LVEF, and BUN had an AUC of 0.794, 0.629, and 0.711, respectively (Fig. 1). BUNEFr >33 had a sensitivity and specificity of 74% and 64%, respectively. Comparison of AUCs of BUNEFr, BUN, and LVEF revealed that BUNEFr predicted MACE with greater sensitivity and specificity (BUNEFr to LVEF: p = 0.0143; BUNEFr to BUN: p < 0.001; Table 3). We created a Kaplan–Meier graph for BUNEFr cutoff value of 33 which showed increased MACE beginning from even the first months following CABG (Fig. 2).

Fig. 1

Sensitivity and specificity of BUNEFr, BUN, and LVEF in predicting major adverse cardiac events. AUC = area under the curve; BUN = blood urea nitrogen; BUNEFr = blood urea nitrogen-to-left ventricular ejection fraction ratio; LVEF= left ventricular ejection fraction.

Table 3

Comparison of area under curves of BUNEFr, BUN, and LVEF.

	AUC Dif.	SE	95% CI	Z	p
BUNEFr to LVEF	0.082	0.0338	0.016–0.149	2.450	0.0143
BUNEFr to BUN	0.165	0.0310	0.104–0.225	5.310	<0.001

AUC = area under the curve; BUN = blood urea nitrogen; BUNEFr = blood urea nitrogen-to-left ventricular ejection fraction ratio; CI = confidence interval; LVEF = left ventricular ejection fraction.

Fig. 2

Kaplan–Meier analysis for survival when BUNEFr >33 mg/dL. BUNEFr = blood urea nitrogen-to-left ventricular ejection fraction ratio.

4. Discussion

In the present study, we found that during 10 years of follow-up, elevated BUNEFr and increased additive Euroscore predicted MACE in patients who underwent CABG procedure. To date, this is the first report to study role of BUNEFr in the prediction of MACE in patients with CAD.

Blood urea nitrogen is a blood parameter whose serum level is influenced by several factors including renal functions, neurohormonal, and sympathetic activity. In particular, serum BUN level increases in patients with HF due to passive reabsorption of urea from renal tubules in response to activation of sympathetic system and RAAS [4-6]. Activation of the neurohormonal system could also lead to renal vasoconstriction and decreased GFR and urea excretion [12,13]. Furthermore, insufficient blood volume secondary to low cardiac output stimulates the release of arginine vasopressin, which can facilitate the reabsorption of BUN in the collecting duct [14,15]. Thus, BUN may serve as an indicator of both cardiorenal dysfunction and neurohormonal activation [16]. BUN has been considered as a prognostic predictor of long-term mortality in acute and chronic HF [17,18]. In addition, data from several recent studies suggested that elevation in serum BUN level predicted worse outcomes in patients with acute MI, acute coronary syndrome, and following elective percutaneous coronary procedures [19-22]. Arnan et al. [23] reported that an increase in the BUN level was associated with an increased risk of stroke during the postoperative period following cardiovascular surgery.

Prior studies mainly included patients with LV systolic dysfunction and subsequent activation in sympathetic system and RAAS. By contrast, diastolic dysfunction can cause activation in both systems in a similar manner, HF symptoms due to elevation in LV end-diastolic filling pressures, and is associated with adverse cardiac outcomes including cardiovascular mortality [24-26]. Diastolic dysfunction can be frequently diagnosed in patients with systolic HF, and both entities share similar risk factors including diabetes and hypertension [27,28]. This can have an additional impact on activating the neurohormonal system and RAAS leading to elevation in BUN level [6,29]. Importantly, LVEF may have limitations in reflecting the renal functions and thus it is reasonable to consider BUNEFr as a more systemic marker than LVEF alone given the fact that comorbidities including diabetes and hypertension can also cause renal dysfunction and BUNEFr reflects the effect of cardiac function on the renal system not only through LV systolic function but also through a diastolic component [30].

It is important to note that BUN also has limitation in predicting prognosis in cardiovascular diseases because several other conditions can affect its level. Although LVEF has limitations in reflecting systemic mechanisms leading to adverse events, it is still the most reliable predictor of prognosis particularly in HF [31].Thus, combining BUN and LVEF as one parameter could have better accuracy for both a systemic and local marker.

In conclusion, although it is not possible to define precise pathophysiological mechanisms, we found that the BUNEFr value determined during the preoperative period could predict patients that were a high risk for new-onset HF and all-cause mortality in patients undergoing CABG procedure. It is important to note that both BUN and LVEF are easily obtainable parameters, and BUNEFr could be a cost-effective way to risk stratify patients prior to the CABG procedure.

4.1. Limitation

This is a single-center study with a limited number of patients. Due to inherent characteristics of observational studies, we cannot demonstrate the precise causative mechanism between BUNEFr and MACE.