Impact of Chronic Kidney Disease on Long-Term Outcome of Coronary Artery Bypass Grafting in Patients With Diabetes Mellitus

Abstract

Background: The aim of this study was to compare the short- and long-term outcomes of CABG in diabetes mellitus (DM) patients according to eGFR.

Methods and Results: A total of 573 DM patients receiving CABG between 1997 and 2012 were stratified according to preoperative eGFR: normal or mild chronic kidney disease (CKD), eGFR ≥60 ml/min/1.73 m²; moderate CKD, eGFR 30–60 ml/min/1.73 m²; severe CKD, eGFR <30 ml/min/1.73 m²; and severe CKD requiring hemodialysis (HD). Off-pump and bilateral internal thoracic artery (BITA) grafting rates were 83.4 and 62.3%, respectively. Mediastinitis and in-hospital mortality rates were both 1.4%. On logistic regression analysis, preoperative congestive heart failure and CKD severity were independent predictors of postoperative renal failure and major complications. The mean follow-up period was 5.7 years (range, 0–15.5 years). Estimated 5-year survival (92.9±1.6%, 82.8±3.3%, and 47.3±7.0%, respectively, P<0.001) significantly decreased with declining kidney function. On Cox hazard modeling, CKD severity was an independent predictor of major cerebrocardiovascular events (normal/mild: hazard ratio [HR], 1; moderate: HR, 1.35; severe: HR, 1.83; HD: HR, 2.0, P=0.016) and of overall survival (normal/mild: HR, 1; moderate: HR, 1.65; severe: HR, 5.96; HD: HR, 10.93, P<0.001). BITA grafting was a strong protective factor for overall survival (HR, 0.63; P=0.022).

Conclusions: In DM patients, early- and long-term outcomes after CABG are strongly influenced by CKD progression. (Circ J 2016; 80: 110–117)

Diabetes mellitus (DM) is an important risk factor in advanced coronary artery disease (CAD), and it influences survival after coronary revascularization, including coronary artery bypass grafting (CABG).^1–3 Nephropathy is one of the main complications in patients with DM, and, at present, DM is the leading cause of progression to chronic kidney disease (CKD) requiring chronic hemodialysis (HD).^4–7 CKD itself not only worsens systemic atherosclerosis, resulting in CAD,^8–10 but also has an impact on the outcome of coronary revascularization.^11–18 In addition, DM and CKD frequently coexist.^8,9 Despite these facts, little is known about the impact of the severity of CKD on outcome CABG in DM patients. The aim of this study was to investigate the relationship between the preoperative severity of CKD and postoperative results in DM patients undergoing CABG.

Methods

This study was approved by the Institutional Review Board of Kyoto Prefectural University of Medicine. Between July 1997 and December 2012, a total of 1,177 consecutive patients underwent isolated CABG at Kyoto Prefectural University of Medicine performed by 2 surgeons (H.Y. and D.K.). Informed consent was obtained preoperatively from each patient. Of these, the 573 patients diagnosed with DM were enrolled, and retrospectively analyzed. The cause of CKD was not considered in this study. Since January 2000, we have performed off-pump CABG (OPCAB) as the first choice for surgical revascularization regardless of the level of urgency.

Definition of CKD

The preoperative severity of CKD was classified according to estimated glomerular filtration rate (eGFR), calculated using the Modification of Diet in Renal Disease formula for Japanese patients: eGFR (ml/min/1.73 m²)=194×(serum creatinine [mg/dl]^–1.094×(age[years])^–0.287×0.739 (if female).¹⁹ In this study, we did not consider the severity of proteinuria as a diagnosis of CKD. The patients were divided into 4 groups based on eGFR: normal or mild CKD, eGFR >60 ml/min/1.73 m²; moderate CKD, eGFR=30–60 ml/min/1.73 m²; severe CKD not requiring chronic HD, eGFR <30 ml/min/1.73 m²; and severe CKD dependent on HD.

Definition of DM

DM was defined as a previous diagnosis of DM and preoperative hemoglobin (Hb) A1c ≥6.5%, with oral hypoglycemic agents or insulin injection, or dietary and exercise therapy. Patients without these treatments with preoperative HbA1c ≥6.5% were also defined as having DM.

Revascularization Strategy

The strategy for multi-vessel revascularization has been reported previously.²⁰ Briefly, the left anterior descending artery was revascularized via in situ internal thoracic artery (ITA) grafting, mostly achieved with right ITA grafting. The left circumflex artery was revascularized with another ITA or a great saphenous vein (SVG) graft. The right coronary artery (RCA) was revascularized using an SVG or a gastroepiploic artery (GEA) graft. A GEA graft was used in patients with a severely calcified ascending aorta and/or marked stenosis (>90%) of the proximal RCA.

Surgical Technique

All surgical procedures were performed through a median sternotomy. The ITA and GEA were harvested in a skeletonized fashion using a harmonic scalpel (Harmonic Scalpel; Ethicon Endo-Surgery; Cincinnati, OH, USA). After systemic heparinization (1 mg/kg in OPCAB, 3 mg/kg in conventional CABG), the ITA were transected. All arterial conduits were dilated using papaverine hydrochloride (0.3 mg/ml). The SVG was harvested via direct exposure from the lower leg. Proximal anastomosis was performed with an anastomotic device (Novare Surgical Systems; Cupertino, CA, USA), and an intracoronary shunt tube (Anastaflow; Edwards Lifescience; Irvine, CA, USA) was routinely used in OPCAB. The ITA grafts were anastomosed with an 8-0 polypropylene suture. Distal anastomosis of other grafts was performed with a 7-0 or 8-0 polypropylene suture. To evaluate the graft flow, a transit time flow trace (MediStim VQ-1101, MediStim ASA, Oslo, Norway) was performed. In conventional CABG, cardiopulmonary bypass was established between the ascending aorta and bicavae. Cardiac arrest was maintained by both ante- and retrograde cardioplegic solution. Distal anastomosis was the same as performed in the off-pump technique.

Endpoints and Definition

The operative data and incidence rates of in-hospital complications were extracted from hospital records. The duration of follow-up was calculated from the date of operation to that of death or the last direct contact by telephone or interview. Follow-up was performed via direct patient contact or telephone and a postcard survey each year. The endpoints were overall death and major adverse cerebrocardiac events (MACCE). MACCE were defined as cardiac death, repeat revascularization, readmission for congestive heart failure, and stroke. Cardiac death included death caused by arrhythmia, myocardial infarction (MI), heart failure, and sudden death. Major in-hospital complications included in-hospital death, reoperation, prolonged ventilation >72 h or reintubation, newly required dialysis, deep sternal wound infection (DSWI), and stroke. Postoperative renal failure was defined as newly required HD, or creatinine increased to ≥2-fold the preoperative baseline and ≥2.0 mg/dl. Perioperative MI was defined as elevated creatinine kinase-MB >100 IU/L. Low-output syndrome was defined as the need for intra-aortic balloon pump, extracorporeal membrane oxygenator, or >5 µg∙kg^–1∙min^–1 catecholamine >24 h. Mitral regurgitation (MR) was investigated using transthoracic echocardiography and divided into 5 grades: none, 0; mild, 1+; moderate, 2+; moderate-severe, 3+; and severe, 4+.

Statistical Analysis

Continuous variables are expressed as mean±SD. Comparisons of clinical characteristics and outcomes between each group were performed using Kruskal-Wallis test for skewed variables, and chi-squared test for categorical variables. Kaplan-Meier survival curves were used to evaluate the long-term results. Log-rank test was used to compare survival curves. Odds ratios (OR) and 95% confidence intervals (CI) for the association between patient background and in-hospital complications were estimated using logistic regression analysis. Independent predictors of long-term outcome were determined using the Cox proportional hazard models and expressed as hazard ratios (HR) and 95% CI. Multivariate analyses were performed using a stepwise downward selection model. Results were considered significant at P<0.05. All statistical calculations were performed using SPSS 22.0 (SPSS, Chicago, IL, USA).

Results

Patient background is summarized in Table 1. The mean age in the moderate and severe CKD groups was significantly older than in the other groups. The prevalence of DM on oral medication decreased with increasing CKD level. The rate of insulin users (64.6%) and mean New York Heart Association (NYHA) classification were significantly higher, and mean ejection fraction (EF) and body mass index (BMI) were lower in the HD group. The prevalence of hypertension was higher in patients with more than moderate CKD. There were no differences in the prevalence of male sex, number of diseased vessels, left main disease, history of smoking, percutaneous coronary intervention or MI, MR >grade 2, hyperlipidemia, chronic obstructive pulmonary disease, peripheral artery disease, stroke, atrial fibrillation (AF), urgency, or redo surgery.

Table 1. Patient Preoperative Characteristics

	All (n=573)	Normal/mild CKD (n=310)	Moderate CKD (n=181)	Severe CKD (n=34)	HD (n=48)	P-value
Age (years)	66.8±8.8	65.6±9.0	68.8±8.2	68.3±10.3	65.7±7.8	<0.001
Male	422 (73.6)	235 (75.8)	127 (70.2)	23 (67.6)	37 (77.1)	0.424
BMI (kg/m2)	23.5±3.5	23.7±3.5	23.4±3.2	24.3±4.0	21.8±3.8	0.003
EF (%)	59.3±14.7	61.0±13.6	58.4±15.5	57.3±16.1	53.3±15.3	0.011
EF <35%	38 (6.8)	13 (4.3)	15 (8.6)	4 (11.8)	6 (12.8)	0.052
Diseased vessels	2.4±0.8	2.4±0.8	2.5±0.8	2.4±0.9	2.4±0.8	0.877
LMT ≥50%	215 (37.5)	117 (37.7)	70 (38.7)	12 (35.3)	16 (33.3)	0.966
NYHA class (mean)	1.9±0.9	1.8±0.9	1.9±0.9	2.0±0.9	2.4±0.9	<0.001
Previous PCI	164 (28.7)	80 (25.9)	60 (33.3)	8 (23.5)	16 (33.3)	0.259
Previous MI	290 (51.9)	162 (53.5)	95 (54.0)	15 (45.5)	18 (38.3)	0.201
Acute MI	17 (3.0)	8 (2.6)	6 (3.4)	2 (6.1)	1 (2.1)	0.707
MR ≥2	42 (7.3)	23 (7.4)	14 (7.7)	2 (5.9)	3 (6.3)	0.972
DM	573 (100)	310 (100)	181 (100)	34 (100)	48 (100)	1.000
Diet/Exercise	99 (17.3)	62 (21.7)	24 (13.3)	8 (23.5)	5 (10.4)	<0.001
Drug	289 (50.4)	171 (53.9)	92 (50.8)	14 (41.2)	12 (25.0)	<0.001
Insulin user	185 (32.3)	77 (24.4)	65 (35.9)	12 (35.3)	31 (64.6)	<0.001
Hemoglobin A1c (%)	7.1±1.3	7.2±1.2	7.1±1.3	7.0±1.1	6.2±1.2	<0.001
Hypertension	534 (72.0)	172 (67.2)	145 (80.1)	28 (82.4)	38 (79.2)	<0.001
Hyperlipidemia	315 (55.1)	172 (55.7)	102 (56.4)	17 (50.0)	24 (50.0)	0.797
Smoking	282 (49.4)	152 (49.4)	94 (51.9)	14 (41.2)	22 (45.8)	0.653
COPD	18 (3.1)	10 (3.2)	7 (3.9)	0 (0)	1 (2.1)	0.660
eGFR (ml/min/1.73 m2)	60.8±28.3	80.2±20.1	48.1±7.9	23.2±5.4	10.1±6.6	<0.001
Cre (mg/dl)	1.4±1.6	0.7±0.1	1.1±0.2	2.3±0.8	5.9±2.6	<0.001
PAD	77 (13.5)	0 (0)	30 (16.6)	3 (8.8)	8 (16.7)	0.330
Stroke	71 (12.4)	36 (11.7)	24 (13.3)	5 (14.7)	4 (8.3)	0.795
AF	24 (4.2)	38 (12.3)	10 (5.6)	0 (0)	1 (2.1)	0.417
Emergency	21 (3.7)	13 (4.2)	10 (5.5)	1 (2.9)	1 (2.9)	0.445
Redo	19 (3.3)	9 (2.9)	8 (4.4)	0 (0)	3 (6.3)	0.300

Data given as n (%) or mean±SD. AF, atrial fibrillation; BMI, body mass index; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; Cre, creatinine; DM, diabetes mellitus; EF, ejection fraction; eGFR, estimated glomerular filtration rate; HD, hemodialysis; LMT, left main trunk; MI, myocardial infarction; MR, mitral regurgitation; NYHA, New York Heart Association; PAD, peripheral artery disease; PCI, percutaneous coronary intervention.

Operative Results and In-Hospital Outcome

Intraoperative and postoperative results are summarized in Table 2. The rate of OPCAB was 83.4% in total, and increased with rising CKD (P=0.086). Bilateral ITA (BITA) grafting was performed in 62.3% of all patients in this study. The rate of BITA use was relatively lower in the severe CKD group, while in the HD group it was 62.5%. Other arterial grafts were rarely used in the severe CKD or HD group. The rate of SVG use was associated with elevated CKD (P=0.073). There were no differences regarding the rate of complete revascularization, mean number of distal anastomoses, or intra-aortic balloon pumping. There were 10 in-hospital deaths (1.4%) in this study. Rates of in-hospital mortality and major complications were significantly higher in the severe CKD+HD group. Postoperative renal failure was rare in the normal or mild CKD group. The incidence of DSWI was significantly higher in the HD group (6.3%). The duration of intensive care unit stay was significantly longer in the normal or mild CKD group. There were no differences observed in the rates of re-exploration for bleeding, prolonged ventilation, perioperative stroke, AF, perioperative MI, or low-output syndrome.

Table 2. Operative Details and Patient Postoperative Characteristics

	All (n=573)	Normal/mild CKD (n=310)	Moderate CKD (n=181)	Severe CKD (n=34)	HD (n=48)	P-value
Off-pump	478 (83.4)	249 (80.3)	155 (85.6)	29 (85.3)	45 (93.8)	0.086
No. distal anastomoses	3.3±1.2	3.4±1.2	3.3±1.1	3.1±1.1	3.3±1.0	0.591
Complete revascularization	496 (86.6)	274 (88.4)	150 (82.9)	29 (85.3)	43 (89.6)	0.330
Bilateral ITA	357 (62.3)	205 (66.1)	105 (58.0)	17 (50.0)	30 (62.5)	0.136
SVG	348 (60.7)	174 (56.1)	117 (64.6)	25 (73.5)	32 (66.7)	0.073
Radial	51 (8.9)	37 (11.9)	14 (7.7)	0 (0)	0 (0)	0.008
GEA	57 (9.9)	35 (11.3)	20 (11.0)	0 (0)	2 (4.2)	0.093
IABP	85 (14.9)	43 (14.0)	26 (14.4)	7 (21.2)	9 (18.8)	0.607
Re-exploration for bleeding	10 (1.8)	4 (1.3)	5 (2.8)	0 (0)	1 (2.1)	0.559
ICU stay (days)	3.0±3.9	9.2±21.5	3.7±5.8	3.9±3.9	3.8±3.6	<0.001
Prolonged ventilation	21 (3.7)	2.3±2.0	8 (4.4)	3 (8.8)	2 (4.2)	0.273
Renal failure	28 (4.9)	8 (2.6)	16 (8.8)	9 (26.5)	0 (0)	<0.001
Stroke	9 (1.6)	3 (1.0)	5 (2.8)	1 (2.9)	0 (0)	0.312
AF	111 (19.4)	3 (1.0)	35 (19.3)	7 (20.6)	8 (16.7)	0.990
PMI	22 (3.8)	61 (19.8)	7 (3.9)	1 (2.9)	2 (4.2)	0.993
LOS	33 (5.8)	12 (3.9)	11 (6.1)	1 (2.9)	2 (4.2)	0.841
Any infection	46 (8.1)	19 (6.1)	14 (7.8)	5 (14.7)	6 (12.5)	0.266
DSWI	8 (1.4)	4 (1.3)	0 (0)	1 (2.9)	3 (6.3)	0.010
Hospital stay (days)	21.3±27.5	4 (1.3)	19.4±9.6	32.6±36.2	26.8±49.9	0.006
Major complications	74 (12.9)	20.2±28.4	29 (16.0)	11 (32.4)	13 (27.1)	<0.001
Hospital mortality	8 (1.4)	2 (0.6)	2 (1.1)	2 (5.9)	2 (4.2)	0.011

Data given as n(%) or mean±SD. DSWI, deep sternal wound infection; GEA, gastroepiploic artery; IABP, intra-aortic balloon pumping; ICU, intensive care unit; ITA, internal thoracic artery; LOS, low output syndrome; PMI, perioperative MI; SVG, saphenous vein graft. Other abbreviations as in Table 1.

Results of multivariate logistic regression analysis are presented in Table 3. Preoperative kidney function was an independent risk factor for major postoperative complications (normal/mild: OR, 1; moderate: OR, 1.7; severe: OR, 3.2; and HD: OR, 2.03, respectively; P<0.001). Preoperative NYHA classification was also a strong predictor of major complications (OR, 1.8; P=0.001). Concerning renal failure, moderate CKD and severe CKD were very strong predictors (moderate: OR, 8.4; severe: OR, 21.5, P<0.001, respectively). Preoperative HD was the only predictor of DSWI (OR, 4.8; P=0.039); insulin users and BITA grafting were not predictors for DSWI.

Table 3. Independent Risk Factors for Renal and Major In-Hospital Complications

	Adjusted OR	95% CI	P-value
Major complications
NYHA (1 increment)	1.82	1.48–2.28	<0.001
Group	–	–	<0.001
Normal/Mild CKD	1	–	–
Moderate CKD	1.71	1.002–2.92	0.049
Severe CKD	3.20	1.58–6.49	0.001
HD	2.03	1.01–4.08	0.046
Renal failure
NYHA (1 increment)	2.06	1.42–2.98	<0.001
Group	–	–	<0.001
Normal/Mild CKD	1	–	–
Moderate CKD	8.39	2.44–28.79	0.001
Severe CKD	21.48	5.89–78.38	<0.001
HD	–	–	–

The following variables were entered: age, male sex, BMI, LMT, insulin user, PAD, hyperlipidemia, NYHA, EF, renal function classification, redo, MR, AF, preoperative stroke, off-pump, emergency, and BITA. CI, confidence interval; OR, odds ratio. Other abbreviations as in Tables 1,2.

Long-Term Outcomes

The follow-up rate was 95.1%, and mean follow-up period was 5.7±3.9 years (range, 0–15.5). Freedom from MACCE at 5 years was 75.9±2.7% for normal/mild CKD, 63.1±4.1% for moderate CKD, 56.7±10.2% for severe CKD, and 47.2±10.7% for HD (log-rank, P=0.004; Figure 1A). There were 131 deaths in the follow-up period. The causes of death were cardiac death (n=31, 26.7%), cancer (n=25, 19.1%), pneumonia (n=19, 14.5%), renal failure (n=11, 8.4%), and stroke (n=10, 7.6%). Freedom from cardiac death (Figure 1B) and overall survival (Figure 2) was significantly decreased with elevated CKD level. Overall survival at 5 years was 92.9±1.6% for normal/mild CKD, 82.8±3.3% for moderate CKD, 56.2±10.3% for severe CKD, and 40.1±9.5% for HD (P<0.001). Detailed causes of death are given in Table S1, and freedom from repeat revascularization and stroke in Figure S1.

Figure 1.

Freedom from (A) major adverse cerebrocardiovascular events and (B) cardiac death according to level of chronic kidney disease. HD, hemodialysis.

Figure 2.

Rate of overall survival according to level of chronic kidney disease. HD, hemodialysis.

The results of the Cox hazard model are summarized in Table 4. On multivariate analysis, preoperative CKD level was a strong predictor of MACCE. While moderate CKD had borderline significance (HR, 1.35; P=0.067), severe CKD and dependence on HD were independent risk factors for MACCE (severe CKD: HR, 1.83; HD: HR, 2.0; P=0.016). Preoperative NYHA classification was an independent risk factor for MACCE (HR, 1.33; P<0.001).

Table 4. Multivariate Predictors of MACCE

	Adjusted HR	95% CI	P-value
MACCE
Group	–	–	0.016
eGFR ≥60	1	–	–
eGFR 30–60	1.35	0.98–1.86	0.067
eGFR <30	1.83	1.04–3.20	0.036
HD	2.00	1.19–3.38	0.010
NYHA (1 increment)	1.33	1.13–1.55	<0.001
Overall survival
Group	–	–	<0.001
eGFR ≥60	1	–	–
eGFR 30–60	1.65	1.08–2.54	0.022
eGFR <30	5.96	3.30–10.78	<0.001
HD	10.93	5.93–20.16	<0.001
Age (1-year increment)	1.01	1.04–1.09	<0.001
Male sex	1.82	1.17–2.81	0.008
PAD	1.93	1.25–2.98	0.003
AF	3.20	1.63–6.29	0.001
MR ≥2	2.16	1.21–3.85	0.009
EF (1% increment)	0.98	0.97–0.99	<0.001
BITA grafting	0.63	0.42–0.93	0.022
Insulin user	1.33	0.88–1.99	0.173

Following variables were entered: age, male sex, BMI, LMT, insulin user, PAD, hyperlipidemia, NYHA, EF, renal function classification, redo, MR, AF, preoperative stroke, off-pump, emergency, and BITA. HR, hazard ratio; MACCE, major adverse cerebrocardiac event. Other abbreviations as in Tables 1–3.

The severity of CKD was found to be a very strong predictor of overall death (moderate CKD: HR, 1.65; severe CKD: HR, 5.96; HD: HR, 10.92; P<0.001). Age (HR, 1.07; P<0.001), male sex (HR, 1.82; P=0.008), peripheral artery disease (HR, 1.93; P=0.003), MR≥2 (HR, 2.16; P=0.009), and AF (HR, 3.2; P<0.001) were also independent risk factors. BITA grafting was a strong protective factor against overall death (HR, 0.63; P=0.022). Insulin usage was not an independent risk factor for MACCE or overall death.

Subgroup analysis was performed in the severe CKD and HD groups to assess the efficacy of BITA grafting. Neither freedom from cardiac death nor overall survival were significantly different between the single ITA (SITA) and BITA groups (P=0.245 and 0.263, respectively; Figure 3). On Cox hazard modeling, however, BITA grafting was an independent protective factor against death (HR, 0.37; P=0.033; Table 5).

Figure 3.

(A) Freedom from cardiac death and (B) overall survival in the severe chronic kidney disease+hemodialysis group, according to graft type. BITA, bilateral internal thoracic artery; SITA, single internal thoracic artery.

Table 5. Independent Factors for Overall Survival in the Severe CKD+HD Group

	Adjusted HR	95% CI	P-value
Age (1-year increment)	1.08	1.01–1.15	0.016
Body mass index (1 kg/m2 increment)	0.86	0.75–0.97	0.016
EF (1% increment)	0.95	0.92–0.98	<0.001
BITA grafting	0.37	0.14–0.92	0.033
Hyperlipidemia	1.3	0.88–1.99	0.055
MR ≥2	4.8	0.89–26.2	0.068

The following variables were entered: age, male sex, BMI, LMT, insulin user, PAD, hyperlipidemia, NYHA, EF, HD, MR, preoperative stroke, off-pump, emergency, and BITA. Abbreviations as in Tables 1–4.

Discussion

In this study, we have shown that preoperative CKD severity is a strong predictor of both short- and long-term outcomes in DM patients undergoing CABG.

DM is an important risk factor that significantly influences both early and late outcomes after CABG. This is because DM patients have more comorbidities, including CKD and complex coronary lesions, such as a narrow diameter or multivessel disease accompanied by endothelial dysfunction.²¹ eGFR measured on blood test is more accurate for assessing the kidney function as compared with serum creatinine level.^4–6,19 Previous studies showed that the prognosis after CABG worsens as CKD progresses, which is associated with the inflammatory response, activation of the renin-angiotensin-aldosterone system, electrolyte disturbance, hypertension, dyslipidemia, nephrogenic anemia, left ventricular hypertrophy, and endothelial dysfunction.^8–18 In this study, all patients had DM and the prevalences of moderate and severe CKD were 31.6% and 14.3%, respectively. These rates were higher than in the STS database report (moderate CKD, 24%; severe CKD, 2.5%).¹⁴ This difference was due to the fact that all patients in the present study had DM.

The characteristics of this study were a high rate of OPCAB (83.4%) and the use of BITA (62.3%). OPCAB was frequently adopted in the severe CKD group. This may indicate the prevalence of comorbidity in the severe CKD group, such as calcified aorta, clinical or subclinical cerebrovascular disease, impaired left ventricular function, or chronic HD. There have been many favorable reports on OPCAB, especially for high-risk patients.^22–24 In the present study, in-hospital mortality was excellent (0.6%), excluding the severe CKD and HD groups. This rate, however, was similar to those of previous reports (2.5–14.8% in patients with severe CKD),^13,16,18,24 because all patients had DM in the present series, while those studies included 30–80% of patients without DM. OPCAB was not an independent protective factor against in-hospital death. This was mainly due to the small number of events (8 patients). We believe, however, that high-risk patients with moderate or severe CKD are good candidates for OPCAB.

As expected, the level of CKD was an independent predictor of major complications and renal failure. Clearly, the postoperative rate of newly required HD rose inversely with declining renal function in this study. Boulton et al reported that in-hospital mortality increased significantly as CKD progressed.²⁵ Zakeri et al also found that renal function was significantly correlated with mortality and newly required HD.¹³ Despite the marked long-term benefits of BITA grafting, one of the concerns regarding its use is postoperative DSWI.^26–29 In the present study, the overall rate of DSWI was acceptable (1.4%), but it was significantly higher in the HD group. This rate, however, was similar to those of previous reports including HD cohorts.^16,25 On multivariate analysis, chronic HD was the only independent predictor of DSWI. Minakata et al reported a similar result whereby severe CKD including HD was an independent risk factor for developing any infection.¹⁶ Deo et al showed that harvesting BITA with a skeletonized technique is not associated with DSWI because it can minimize tissue damage, and preserve the blood supply to the sternum.²⁸ Therefore, the present lower incidence of DSWI even in DM patients may be mainly attributable to the skeletonized technique. These results, however, should not be directly applied to other races because obesity is also a known risk factor for DSWI, and the mean BMI of Japanese study subjects, including that in the present study, was around 23 kg/m²,^16,23–25 while that in the US was 27–30 kg/m².^13,14,26

With regard to long-term results, CKD severity was strongly correlated with overall survival, cardiac death, and MACCE among DM patients. Severe CKD with or without HD was associated with a markedly high HR for overall death, but the exact mechanism of the association between CKD and death after CABG remains unclear.^4–10 As mentioned here, CKD itself is a risk factor for developing cardiovascular disease, and cardiac dysfunction also worsens CKD. In the presence of DM, this pathogenesis can be more complicated, and may worsen the cardiovascular disease, resulting in death or MACCE. On Kaplan-Meier survival curve analysis, the rate of MACEE in the moderate CKD group was similar to that in the severe CKD or HD groups, but the types of MACCE were different: the events that occurred in the moderate CKD group were relatively mild. Indeed, the leading cause of MACCE in the moderate CKD group was re-admission for heart failure (32.8%) due to hypertension, rapid AF, or volume overload, and repeat revascularization ranked second (28.1%). Cardiac death (36.7%), stroke (26.7%), and heart failure (26.7%) were the main causes of MACCE in the severe CKD+HD group (data not shown).

Historically, insulin usage is a strong risk factor affecting both short- and long-term outcomes after CABG.^1,2 Although insulin usage was not an independent predictor of overall death in this study, there is a possibility that advanced DM contributes to the poorer prognosis after CABG because insulin users comprised 64.6% of HD patients in this study. Suzuki et al noted similar long-term outcomes in insulin users and the other DM patients with multiple arterial grafting.³⁰ They used BITA grafting for approximately 69% of patients, and this was similar to the present rate (62.3%).

In the present study we also found that BITA grafting was an independent protective factor for survival in the overall cohort and on subgroup analysis. The long-term benefits of BITA grafting in DM patients with multivessel disease have already been shown.^26–29 As for BITA grafting in CKD patients, the efficacy of BITA grafting is one of the concerns when treating patients with markedly impaired renal function.^31,32 Kinoshita et al noted superior midterm results for skeletonized BITA grafting as compared with SITA grafting in patients with CKD.³³ Their recent study using propensity-matched analysis showed favorable results for BITA grafting in patients with a markedly impaired renal function (eGFR <30 ml/min/m²).³⁴ In the present subgroup analysis, BITA grafting had good efficacy even in DM patients with eGFR <30 ml/min/m². Therefore, the present study suggests that BITA grafting has potential benefits even for DM patients with impaired kidney function.

Preoperative AF, MR, and reduced EF were independent risk factors for predicting overall death. In this study, we mainly used OPCAB, and therefore there may have been a selection bias because mitral valve surgery was not conducted for relatively high-risk patients with borderline severity of MR.

Study Limitations

This study has some limitations. First, it was a single-center, nonrandomized, retrospective study. Second, the overall subject group was relatively small, resulting in insufficient statistical power to create risk models. Third, we did not consider the precise cause of CKD despite the fact that all patients had DM in this study. Fourth, we could not examine renal outcomes in the long-term. Fifth, although this study demonstrated the benefits of BITA grafting, a potential selection bias for graft selection existed, based on the preoperative status not represented in the data. Finally, we did not examine the relationship between graft patency and long-term outcome.

Conclusions

Among patients with DM, long-term morbidity and mortality after CABG markedly increased with CKD progression.

Acknowledgments

The authors thank Professor Satoshi Teramukai (Department of Biostatics, Kyoto Prefectural University of Medicine, Kyoto, Japan), for statistical discussion and valuable comments on this study.

Conflict of Interest

None declared.

Supplementary Files

Supplementary File 1

Table S1. Cause of late death

Figure S1. Freedom from repeat revascularization and stroke.

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-15-0776

References

• 1.    Mohanmmadi  S,  Degenais  F,  Mathie  P,  Kingma  JG,  Doyle  D,  Baillot  R, et al. Long-term impact of diabetes and its comorbidities in patients undergoing isolated primary coronary artery bypass graft surgery. Circulation 2007; 116(Suppl I): 220–225.
• 2.    O’Keef  JH,  Blackstone  EH,  Sergeant  P,  McCallister  BD. The optimal mode of coronary revascularization for diabetics: A risk-adjusted long-term study comparing coronary angioplasty and coronary bypass surgery. Eur Heart J 1998; 19: 1696–1703.
• 3.    Puskas  JD,  Sadiq  A,  Vassiliades  TA,  Kilgo  PD,  Lattouf  OM. Bilateral internal thoracic artery grafting is associated with significantly improved long-term survival, even among diabetic patients. Ann Thorac Surg 2012; 94: 710–715; discussion 715–716.
• 4.    Go  AS,  Chertow  GM,  Fan  D,  McCulloch  CE,  Hsu  CY. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004; 351: 1296–1305.
• 5.    Muntner  P,  Hamm  LL,  Kusek  JW,  Chen  J,  Whelton  PK,  He  J. The prevalence of nontraditional risk factors for coronary heart disease in patients with chronic kidney disease. Ann Intern Med 2004; 140: 9–17.
• 6.    Gupta  R,  Birnbaum  Y,  Uretsky  BF. The renal patient with coronary artery disease: Current concepts and dilemmas. J Am Coll Cardiol 2004; 44: 1343–1353.
• 7.    Collins  AJ,  Foley  R,  Herzog  C,  Chavers  B,  Gilbertson  D,  Ishani  A. Excerpts from the United States Renal Data System 2007 annual data report. Am J Kidney Dis 2008; 51: S1–S320.
• 8.    Schrier  RW. Role of diminished renal function in cardiovascular mortality: Marker or pathogenetic factor? J Am Coll Cardiol 2006; 47: 1–8.
• 9.    Keane  WF. Metabolic pathogenesis of cardiorenal disease. Am J Kidney Dis 2001; 38: 1372–1375.
• 10.    Shlipak  MG,  Fried  LF,  Crump  C,  Bleyer  AJ,  Manolio  TA,  Tracy  RP, et al. Elevations of inflammatory and procoagulant biomarkers in elderly persons with renal insufficiency. Circulation 2003; 107: 87–92.
• 11.    Ashrith  G,  Lee  VV,  Elayda  MA,  Reul  RM,  Wilson  JM. Short- and long-term outcomes of coronary artery bypass grafting or drug-eluting stent implantation for multivessel coronary artery disease in patients with chronic kidney disease. Am J Cardiol 2010; 106: 348–353.
• 12.    Hillis  GS,  Croal  BL,  Buchan  KG,  El-Shafei  H,  Gibson  G,  Jefferey  RR, et al. Renal function and outcome from coronary artery bypass grafting: Impact on mortality after a 2.3-year follow up. Circulation 2006; 113: 1056–1062.
• 13.    Zakeri  R,  Freemantle  N,  Barnett  V,  Lipkin  GW,  Bonser  RS,  Graham  TR, et al. Relation between mild renal dysfunction and outcomes after coronary artery bypass grafting. Circulation 2005; 112: 270–275.
• 14.    Cooper  WA,  O’Brien  SM,  Thourani  VH,  Guyton  RA,  Bridges  CR,  Szczech  LA, et al. Impact of renal dysfunction on outcomes of coronary artery bypass surgery: Results from the Society of Thoracic Surgeons National Adult Cardiac Database. Circulation 2006; 113: 1063–1070.
• 15.    Han  SS,  Shin  N,  Baek  SH,  Ahn  SY,  Kim  DK,  Kim  S, et al. Effects of acute kidney injury and chronic kidney disease on long-term mortality after coronary artery bypass grafting. Am Heart J 2015; 169: 419–425.
• 16.    Minakata  K,  Bando  K,  Tanaka  S,  Takanashi  S,  Konishi  H,  Miyamoto  Y, et al. Preoperative chronic kidney disease as a strong predictor of postoperative infection and mortality after coronary artery bypass grafting. Circ J 2014; 78: 2225–2231.
• 17.    Domoto  S,  Tagusari  O,  Nakamura  Y,  Takai  H,  Seike  Y,  Ito  Y, et al. Preoperative estimated glomerular filtration rate as a significant predictor of long-term outcomes after coronary artery bypass grafting in Japanese patients. Gen Thorac Cardiovasc Surg 2012; 62: 95–102.
• 18.    Holzmann  MJ,  Ahnve  S,  Hammar  N,  Jorgensen  L,  Klerdal  K,  Pehrsson  K, et al. Creatinine clearance and risk of early mortality in patients undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg 2005; 130: 746–752.
• 19.    Matsuo  S,  Imai  E,  Horio  M,  Yasuda  Y,  Tomita  K,  Nitta  K, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 2009; 53: 982–992.
• 20.    Ohira  S,  Doi  K,  Numata  S,  Yamazaki  S,  Yamamoto  K,  Fujita  S, et al. Does age at operation influence short and long-term outcomes in off-pump coronary artery bypass grafting? Circ J 2015; 79: 2177–2185.
• 21.    Goraya  TY,  Leibson  CL,  Palumbo  PJ,  Weston  SA,  Killan  JM,  Pfeifer  EA, et al. Coronary atherosclerosis in diabetes mellitus: A population based autopsy study. J Am Coll Cardiol 2002; 40: 946–953.
• 22.    Stamou  SC,  Jablonski  KA,  Hill  PC,  Bafi  AS,  Boyce  SW,  Corso  PJ. Coronary revascularization without cardiopulmonary bypass versus the conventional approach in high-risk patients. Ann Thorac Surg 2005; 79: 552–557.
• 23.    Marui  A,  Okabayashi  H,  Komiya  T,  Tanaka  S,  Furukawa  Y,  Kita  T, et al. Benefits of off-pump coronary artery bypass grafting in high-risk patients. Circulation 2012; 126(11 Suppl 1): S151–S157.
• 24.    Kinoshita  T,  Asai  T,  Murakami  Y,  Suzuki  T,  Kambara  A,  Matsubayashi  K. Preoperative renal dysfunction and mortality after off-pump coronary artery bypass grafting in Japanese. Circ J 2010; 74: 1866–1872.
• 25.    Boulton  BJ,  Kilgo  P,  Guyton  RA,  Puskas  JD,  Lattouf  OM,  Chen  EP, et al. Impact of preoperative renal dysfunction in patients undergoing off-pump versus on-pump coronary artery bypass. Ann Thorac Surg 2011; 92: 595–601.
• 26.    Raza  S,  Sabik  JF 3rd,  Masabni  K,  Ainkaran  P,  Lytle  BW,  Blackstone  EH. Surgical revascularization techniques that minimize surgical risk and maximize late survival after coronary artery bypass grafting in patients with diabetes mellitus. J Thorac Cardiovasc Surg 2014; 148: 1257–1264; discussion 1264–1266.
• 27.    Kieser  TM,  Rose  MS,  Aluthman  U,  Montgomery  M,  Louie  T,  Belenkie  I. Toward zero: Deep sternal wound infection after 1001 consecutive coronary artery bypass procedures using arterial grafts: Implications for diabetic patients. J Thorac Cardiovasc Surg 2014; 148: 1887–1895.
• 28.    Deo  SV,  Shah  IK,  Dunlay  SM,  Erwin  PJ,  Locker  C,  Altarabsheh  SE, et al. Bilateral internal thoracic artery harvest and deep sternal wound infection in diabetic patients. Ann Thorac Surg 2013; 95: 862–869.
• 29.    Hemo  E,  Mohr  R,  Uretzky  G,  Katz  G,  Popovits  N,  Pevni  D, et al. Long-term outcomes of patients with diabetes receiving bilateral internal thoracic artery grafts. J Thorac Cardiovasc Surg 2013; 146: 586–592.
• 30.    Suzuki  T,  Asai  T,  Nota  H,  Kinoshita  T,  Fujino  S,  Takashima  N. Similar outcome in insulin-dependent and noninsulin-dependent diabetic patients after off-pump coronary artery bypass grafting with multiple skeletonized arterial conduits. Ann Thorac Surg 2015; 99: 1562–1567.
• 31.    Kumada  Y,  Ishii  H,  Aoyama  T,  Kamoi  D,  Kawamura  Y,  Sakakibara  T, et al. Long-term clinical outcome after surgical or percutaneous coronary revascularization in hemodialysis patients. Circ J 2014; 78: 986–992.
• 32.   JCS Joint Working Group. Guidelines for the clinical application of bypass grafts and the surgical techniques (JCS 2011) published in 2012: Digest version. Circ J 2013; 77: 1608–1641.
• 33.    Kinoshita  T,  Asai  T,  Murakami  Y,  Hiramatsu  N,  Suzuki  T,  Kambara  A, et al. Efficacy of bilateral internal thoracic artery grafting in patients with chronic kidney disease. Ann Thorac Surg 2010; 89: 1106–1111.
• 34.    Kinoshita  T,  Asai  T,  Suzuki  T. Off-pump bilateral skeletonized internal thoracic artery grafting in patients with chronic kidney disease. J Thorac Cardiovasc Surg 2015; 150: 315–321.e3, doi:10.1016/j.jtcvs.2015.04.058.