Mortality Risk Factors in Chronic Renal Failure Patients after Coronary Artery Bypass Grafting

Abstract

Perioperative risk during coronary artery bypass grafting (CABG) is high in patients with chronic renal disease. We aimed to determine postoperative two-year mortality and identify the preoperative risk factors of mortality during CABG surgery in hemodialysis (HD)-dependent and HD-non-dependent CRF patients. We included 102 CRF patients who underwent CABG in Baskent University Hospital between 2000 and 2005. There were 47 patients with CRF undergoing HD (Group I) and 55 CRF patients without dialysis requirement (Group II). We retrospectively retrieved demographic variables; clinical, operative, and echocardiographic data; and biochemical parameters at the time of the operation and six months postoperation. Postoperative HD requirement in Group II patients and infectious complications were recorded. In the second postoperative year, mortality rate was 27.7% in group I and 16.4% in group II (p > .05). When preoperative risk factors evaluated by univariate Cox analysis, only age (RR = 1.06, p = .04) was a significant determinant of survival in Group I patients. Among the operative and postoperative risk factors of mortality such as duration of operation, numbers of coronary vessel bypass, HD requirement, and infection were investigated in Group I and II patients. Rate of infectious complication (including mediastinitis) was found to be a major determinant of mortality by multivariate Cox analyses in both group I (RR = 4.42, p ≤ .05) and group II (RR = 9.39, p ≤ .05). HD dependency did not increase mortality if the patients are younger and were electively prepared for CABG surgery. High infection rates have increased the postoperative mortality and hospitalization in CRF patients. Early diagnosis of infections in CRF patients is important for early recovery, shorter hospitalization, and lower mortality after CABG operation

Keywords:

• coronary artery disease
• bypass grafting
• hemodialysis
• mortality

INTRODUCTION

Prevalence of coronary artery disease (CAD) is high in chronic renal failure (CRF) and is a major cause of mortality and morbidity in these patients.[1],[2] In CRF, uremia-associated factors such as hypertension, hyperlipidemia, anemia, fluid overload, platelet dysfunction, heterotopic calcification due to secondary hyperparathyroidism, and hypercalcemia are known to disturb coronary artery circulation.[3] A majority of the CRF patients with CAD need myocardial revascularization. Coronary artery bypass grafting (CABG) has become the standard treatment for end-stage renal disease (ESRD) patients with CAD. It has been shown to yield better overall and angina-free survival than does percutaneous transluminal coronary angioplasty.[4],[5]

The impaired renal function results in several physiologic abnormalities that could contribute to adverse operative outcome.[6] Several studies have shown that perioperative risk during CABG is high in patients with ESRD,[3] but there has been a limited number of reports about the outcome of patients with mild to moderate renal failure.[7]

Our aim was to determine postoperative two-year mortality and identify and compare the preoperative, intraoperative, and postoperative risk factors of mortality during CABG surgery in hemodialysis (HD)-dependent and HD-non-dependent CRF patients.

PATIENTS AND METHODS

In this study, 102 CRF patients who underwent elective on-pump CABG operation between 2000 and 2005 in Baskent University Hospital were included. CRF patients were selected retrospectively from the database of Cardiovascular Surgery Department. The patients who had emergent CABG or concomitant surgery of the heart valves or other vessels were excluded.

In order to evaluate early postoperative (i.e., first six months) risk factors, the following parameters were evaluated in all patients at the time of the operation and after six months:

• demographic variables: age, sex, the etiology of ESRD, the duration of CRF, and smoking;

• clinical data: presence of diabetes mellitus (DM), hyperlipidemia (HL, total cholesterol >200 mg/dL), hypertension (HT), or previous myocardial infarction (MI); systolic and diastolic blood pressure, body mass index (BMI);

• operative data: numbers of coronary bypass grafts, cardiopulmonary bypass grafts, aortic cross-clamp time, total duration of hospitalization and intensive care unit stays;

• echocardiographic data: ejection fraction (EF), presence of pericardial effusion; and

• biochemical parameters: blood urea nitrogen (BUN), serum levels of creatinine, sodium, potassium, calcium, phosphorus, albumin, lipid profile, hemoglobin.

Also, postoperative hemodialysis requirement in Group II patients and infectious complications (pulmonary, urinary, mediastinal, operation site infections) were recorded.

The patients were divided into two groups. Group I included 47 ESRD patients (33 male and 14 female) who were on HD therapy. The age at the time of CABG surgery was 56.6 ± 9.9 years. Mean duration of CRF was 48.6 ± 9.0 months; mean duration of HD was 39.1 ± 8.1 months. The causes of renal failure in Group I were as follows: chronic glomerulonephritis (3 patients), diabetic nephropathy (15 patients), hypertension (3 patients), pyelonephritis (5 patients), amyloidosis (1 patient), solitary kidney (1 patient), and unknown (19 patients) (see Figure 1). All Group I patients underwent hemodialysis using hemophage dialyzers with an average blood flow rate of 300–350 mL/min and mean Kt/V maintained at >1.2 during each treatment, and received HD the day before the operation. Group II included 55 CRF patients (43 male and 12 female) without any dialysis requirement. The age at the time of surgery was 63.2 ± 8.5 years. Mean duration of CRF was 44.0 ± 12.0 months. Mean creatinine clearance was 27.3–56.8 mL/sec. The causes of CRF were chronic glomerulonephritis (1 patient), diabetic nephropathy (29 patients), hypertension (9 patients), pyelonephritis (4 patients), and unknown (12 patients; see Figure 1). None of the Group II patients needed HD before surgery.

Figure 1. Etiology of chronic renal failure.

All patients were followed postoperatively for two years to detect any mortality. Possible pre- and postoperative clinical and biochemical mortality risk factors were analyzed for each group.

Surgical Procedure

Preoperatively, the patients were submitted to dialysis the day before surgery. All operations were performed through a median sternotomy incision. CABG surgery was elective in all patients. The pericardium was opened, followed by general heparinization, aortic cannulation, and venous cannulation using a double-stage atrial cannula. Cardiopulmonary bypass was started. In situ LITA was always the graft of choice for revascularization of the LAD. LITA was harvested with a pedicle and preserved in a sponge with papaverine to avoid spasm. All operations were performed with moderate general hypothermia (30–32°C) and topical cooling by ice slush. Cardioplegic arrest was achieved by antegrade injection of cold-blood cardioplegia. Myocardial protection was performed by mild hypothermia at 28°C, and cardioplegic arrest was achieved by administering 10–15 mL/kg of cold crystalloid cardioplegia into the aortic root in antegrade fashion. All coronary anastomoses were done using a double-armed 7-0 polypropylene suture with a continuous suturing technique. The way of distal anastomosis of LITA-LAD was end-to-side. Proximal anastomoses were constructed on the ascending aorta with continuous double-armed 6-0 polypropylene sutures using a side clamp during rewarming.

Statistical Analysis

Statistical analysis was performed with SPSS 11.0 (SPSS Inc., Chicago, Illinois, USA). Data were presented as the mean ± the standard deviation (SD). Differences between groups were analyzed using student's t-test and chi-square test. The estimated probability of survival was calculated by the method of Kaplan- Meier survival estimates. The relative risk of mortality for the various preoperative, intraoperative, and postoperative risk factors were based on Cox proportional hazards analyses (univariate and multivariate step forward regression). Statistical significance was assumed if the p value was less than 0.05.

RESULTS

In all, 102 patients with CRF were included. Mean age was 60.17 (39–79) years, and 76 (74.5%) patients were male. Patients' demographic and clinical features are summarized in Table 1, and laboratory data at the time of operation and six months postoperatively are listed in Table 2. There were no significant difference between Groups I and II in terms of sex, CRF duration, prevalence of smoking, presence of DM/HL/HT/MI, numbers of coronary bypass grafts or cardiopulmonary bypass grafts, aortic cross-clamp time, total duration of hospitalization/intensive care unit, echocardiographic data (EF, presence of pericardial effusion), or biochemical parameters (preoperative and postoperative calcium, phosphorus, albumin, C-reactive protein, lipid profile, and postoperative hemoglobin; p > 0.05 for all). However, patients in Group II were older (63.2 ± 8.5 vs. 56.6 ± 9.9 years, p = 0.001); and have higher BMI (26.4 ± 4.0 vs. 24.7 ± 3.0, p = 0.026) and systolic (137.4 ± 18.4 vs. 122.1 ± 22.0 mmHg) and diastolic (79.7 ± 9.6 vs. 71.8 ± 9.7 mmHg) blood pressure than Group I patients. In Group I, preoperative hemoglobin levels was lower than in Group II (10.9 ± 1.4 vs. 12.0 ± 1.8, p = 0.001).

Table 1 Demographic and clinical features of patients in Group I and II

Patient characteristics	Group I (n = 47)	Group II (n = 55)	p
Age (yr.)	56.6 ± 9.9	63.2 ± 8.5	<0.001
Male gender (n/%)	33 (70.2%)	43 (70.8%)	>0.05
CRD duration (months)	48.6 ± 9.0	44.0 ± 12.0	>0.05
Systolic blood pressure (mmHg)	122.1 ± 22.0	137.4 ± 18.4	<0.001
Diastolic blood pressure (mmHg)	71.8 ± 9.7	79.7 ± 9.6	<0.001
DM (n, %)	17 (36.2%)	29 (53.7%)	>0.05
HT (n, %)	39 (83.0%)	46 (83.6%)	>0.05
HL (n, %)	19 (40.4%)	30 (61.2%)	>0.05
MI (n, %)	15 (42.9%)	20 (57.1%)	>0.05
Prevalence of smoking (n, %)	17 (43.6%)	22 (56.4%)	>0.05
EF	49.9 ± 10.8	48.0 ± 8.8	>0.05
BMI (kg/m^2)	24.7 ± 3.0	26.4 ± 4.0	0.02

Abbreviations: CRD = chronic renal disease, DM = diabetes mellitus, HT = hypertension, HL = hyperlipidemia, MI = myocardial infarct, EF = ejection fraction, BMI = body mass index.

Table 2 Comparison of zero and six months' laboratory data in patients in Group I and Group II

	0 months			6 months
	Group I	Group II	p	Group I	Group II	p
Hemoglobin, (gr/dL)	10.9 ± 1.4	12.0 ± 1.8	<0.001	10.7 ± 0.9	11.0 ± 1.0	>0.05
Calcium (mg/dL)	9.3 ± 0.8	9.1 ± 0.8	>0.05	9.0 ± 0.8	8.5 ± 0.8	>0.05
Phosphorus (mg/dL)	4.3 ± 1.6	3.9 ± 0.2	>0.05	4.4 ± 1.4	4.1 ± 0.6	>0.05
Albumin (gr/dL)	3.5 ± 0.5	3.8 ± 0.4	>0.05	3.4 ± 0.6	3.2 ± 0.5	>0.05
CRP (mg/L)	30.7 ± 8.0	22.9 ± 4.2	>0.05	34.1 ± 7.5	31.6 ± 9.8	>0.05
T.cholesterol (mg/dL)	185 ± 32.9	212.8 ± 66.9	>0.05	185.1 ± 68.4	196.7 ± 54.4	>0.05
Trygliceride (mg/dL)	171.8 ± 78.8	176.5 ± 100.9	>0.05	208.5 ± 139.4	136.7 ± 44.1	>0.05
HDL (mg/dL)	39.7 ± 7.2	39.7 ± 8.7	>0.05	44.2 ± 11.6	42.6 ± 7.80	>0.05

Intraoperative data are shown in Table 3 and postoperative complications in Table 4. These parameters did not differ except for amount of blood transfusions. Group I patients had more red cell transfusion (2.9 ± 0.4 vs. 1.6 ± 0.3, p < 0.05). In the first two years postoperatively, mortality rate was 27.7% in group I and 16.4% in group II (p > 0.05; see Figure 2).

Table 3 Intraoperative data

		Group I	Group II	p
Duration of operation (min)		222.6 ± 43.8	236.4 ± 62.9	>0.05
CPB time (min)		82.8 ± 26.0	84.9 ± 33.5	>0.05
Aortic cross-clamp time (min)		39.7 ± 18.4	44.8 ± 16.5	>0.05
Chest drainage time (min)		79.3 ± 34.7	94.9 ± 54.4	>0.05
Number of coronary bypass (n, %)	≤ 2	17 (36.1%)	12 (21.8%)	>0.05
	≥ 3	30 (63.9%)	43 (78.2%)	>0.05

Abbreviation: CPB = cardiopulmoner bypass.

Table 4 Postoperative data and complications

	Group I	Group II	p
Mechanical ventilation (hours)	53.5 ± 23.7	23.3 ± 7.1	>0.05
ICU stay (hours)	128.3 ± 25.3	110.4 ± 12.4	>0.05
Postoperative.stay (days)	15.5 ± 9.4	15.3 ± 8.4	>0.05
Tamponade (%)	1 (2.1%)	0 (0%)	>0.05
Mediastinitis (%)	3 (6.4%)	3 (5.5%)	>0.05
Infection (%)	5 (10.6%)	10 (18.2%)	>0.05
Arrhythmia (%)	1 (2.1%)	3 (5.4%)	>0.05
HD requirement (%)		9 (16.4%)
Blood transfusion (red cells/unit)	2.9 ± 0.4	1.6 ± 0.3	<0.05

Abbreviations: ICU = intensive care unit, HD = hemodialysis

Figure 2. Kaplan Meier proportion of surviving patients comparing HD-dependent and -nondependent CRF patients (log-rank test, p = 0.35).

Preoperative mortality risk factors, including age, HT, MI, DM, HL, history of smoking, EF, and BMI, were analyzed using univariate Cox proportional hazards. The univariate model revealed only age (RR = 1.06, p = 0.04) was significant determinant of survival in Group I patients. All preoperative risk factors from univariate analyses were included in the step forward multivariate Cox proportional hazards model. We did not detect any independent risk factor for mortality in multivariate Cox proportional hazards analysis (see Table 5).

Table 5 Multivariate Cox Proportional Hazards Analysis of Preoperative Risk Factors on Mortality for 24 months

	Group I			Group II
Factor	RR	95% CI	p	RR	95% CI	p
Age	1.07	0.99–1.16	>0.05	0.96	0.86–1.06	>0.05
BMI	1.10	0.83–1.44	>0.05	1.01	0.84–1.21	>0.05
EF	0.99	0.94–1.05	>0.05	0.97	0.87–1.07	>0.05
History of smoking	1.69	0.40–7.12	>0.05	0.74	0.10–5.30	>0.05
HT	0.80	0.70–8.65	>0.05	1.28	0.10–15.7	>0.05
HL	0.13	0.01–1.16	>0.05	7.81	0.81–74.7	>0.05
DM	0.65	0.09–4.46	>0.05	0.45	0.07–2.62	>0.05
MI	0.60	0.12–2.80	>0.05	1.77	0.17–18.5	>0.05

*All factors in the univariate analyses were included in the step forward multivariate model, and significant factors are presented.

Abbreiations: CI = coefficient interval, RR = relative risk, BMI = body mass index, EF = ejection fraction, HT = hypertension, HL = hyperlipidemia, DM; diabetes mellitus, MI = myocardial infarction.

Among the operative and postoperative risks of mortality, time of operation, numbers of coronary vessel bypass, HD requirement, and infection were investigated in both Group I and II patients. Rate of infectious complication (including mediastinitis) was found to be of significance both Group I (RR = 9.70, p = 0.0005) and Group II (RR = 6.25, p = 0.004) by univariate analyses. The results of the multivariate Cox proportional hazards analysis for independent operative and postoperative risk factors are presented in Table 6.

Table 6 Multivariate Cox Proportional Hazards Analysis of Intraoperative and Postoperative Risk Factors of Mortality

	Group I			Group II
Factor	RR	95% CI	p	RR	95% CI	p
Time of operation	1.00	0.98–1.01	>0.05	1.00	0.99–1.01	>0.05
Numbers of coronary vessel bypass	0.39	0.09–1.59	>0.05	1.38	0.15–12.4	>0.05
HD requirement				3.45	0.58–20.4	>0.05
Infection	4.42	0.93–20.8	<0.05	9.39	2.07–42.5	<0.05

*All factors in the univariate analyses were included in the step forward multivariate model, and significant factors are presented.

Abbreviations: CI = coefficient interval; RR = relative risk, HD = hemodialysis.

DISCUSSION

Coronary artery disease causes 47% of deaths in ESRD patients on maintenance dialysis.[1],[3],[4] The higher incidence of CAD in this patient population can be attributed to the presence of comorbid conditions that include hypertension, hyperlipidemia, renal anemia, fluid overload by arteriovenous shunt, heterotopic calcification due to secondary hyperparathyroidism, and abnormal carbohydrate metabolism that leads to accelerated atherosclerosis.[4] Due to the improved life expectancy in uremic patients, the need for surgical revascularization is also increased. In the last 10 years, the number of candidates requiring coronary intervention has been increased. Although percutaneous transluminal coronary angioplasty in CRF patients is feasible and provides relief of angina, aggressive restenosis limits its long term benefit. Thus CABG still remains a preferred therapeutic option in CRF patients.[3],[4] Since the 1970s, a series of studies have reported on early and late outcomes of CABG in CRF patients, yet there was no agreement on this issue. Some groups have shown that CRF patients with CABG have improved survival and quality of life, but others have demonstrated the opposite. Gelsomino et al. have shown that CRF patients with CABG have an increased but acceptable risk of perioperative morbidity and mortality.[8] Hospital mortality and morbidity rates are substantially high when these patients are compared with CABG patients with normal renal function.[4] In the current literature, studies evaluating mortality risk factors in CRF patients concerning dialysis dependency is limited.

In this study, we compared mortality risk factors in HD-dependent and HD-non-dependent CRF patients who underwent CABG. Intraoperative data and complications were not different in HD-dependent and non-dependent patients except for blood transfusion requirement, as HD patients had an increased need for blood transfusion. This is logical because the prevalence of anemia increases as GFR declines. Previous studies showed that while the prevalence of anemia is approximately 2% in earlier stages of chronic kidney disease, it's increasing to nearly 50% during stage 4 chronic kidney disease.[9],[10] Furthermore, other factors may exacerbate anemia in patients receiving HD treatment, including blood loss during HD from multiple blood draws and residual blood left in the extracorporeal unit.[11],[12]

However, blood transfusion requirement had no impact on mortality at two years follow-up by multivariate Cox proportional hazards analyses.

The two-year survival rates for Group I patients is 74.4% and 81.8% in Group II. Witczak et al.[7] found that the one-, three-, and five-year survival rates were 56%, 26%, and 23%, respectively. The higher survival rate in our study may be explained by the exclusion of emergent CABG or concomitant surgery of the heart valves or other vessels.

The independent preoperative factors of mortality vary widely in different studies. Many studies reported that low ejection fraction in echocardiographic evaluation is an independent risk factor.[7] Franga and colleagues[13] found that the only preoperative risk factor of late mortality is smoking. Witczak et al.[7] showed that age, DM, and dialysis treatment are independent risk factors in CRF patients. We didn't find any difference between the groups regarding the age, history of DM/HL/HT/MI, BMI, smoking, or ejection fraction. It is rather difficult to compare the results with similar reports: most studies have small patient groups, the demographical characteristics of the study group vary widely, and the types of CV operations may not be comparable.

When the postoperative risk factors, including time of operation, numbers of coronary vessel bypass, HD requirement, and infection, were analyzed by multivariate Cox hazards, infection incidence was the most important factor determining the mortality within two years' time. We think that HD dependency did not increase mortality if the patients are younger and were electively prepared for CABG surgery. High infection rate has increased the postoperative mortality and hospitalization in HD-dependent and HD–non-dependent CRF patients. Early diagnosis of infections, especially in dialysis-dependent and non-dependent CRF, is important for early recovery and shorter hospitalization in the postoperative period.