Ischemic heart disease is a major risk factor for morbidity and mortality in patients with end-stage renal disease. However, long-term benefits of percutaneous coronary intervention (PCI) relative to coronary artery bypass grafting (CABG) in those patients is still unclear in the drug-eluting stent era. We identified 388 patients with multivessel and/or left main disease with end-stage renal disease requiring dialysis among 15,939 patients undergoing first coronary revascularization enrolled in the Coronary REvascularization Demonstrating Outcome Study in Kyoto PCI/CABG Registry Cohort-2 (PCI: 258 patients and CABG: 130 patients). The CABG group included more patients with 3-vessel (38% vs 57%, p <0.001) and left main disease (10% vs 34%, p <0.001). Preprocedural Synergy between Percutaneous Coronary Intervention with Taxus and Cardiac Surgery score in the CABG group was significantly higher than that in the PCI group (23.5 ± 8.7 vs 29.4 ± 11.0, p <0.001). Unadjusted 30-day mortality was 2.7% for PCI and 5.4% for CABG. Cumulative 5-year all-cause mortality was 52.3% for PCI and 49.9% for CABG. Propensity score–adjusted all-cause mortality was not different between PCI and CABG (hazard ratio [HR] 1.33, 95% confidence interval [CI] 0.85 to 2.09, p = 0.219). However, the excess risk of PCI relative to CABG for cardiac death was significant (HR 2.10, 95% CI 1.11 to 3.96, p = 0.02). The risk of sudden death was also higher after PCI (HR 4.83, 95% CI 1.01 to 23.08, p = 0.049). The risk of myocardial infarction after PCI tended to be higher than after CABG (HR 3.30, 95% CI 0.72 to 15.09, p = 0.12). The risk of any coronary revascularization after PCI was markedly higher after CABG (HR 3.78, 95% CI 1.91 to 7.50, p <0.001). Among the 201 patients who died during the follow-up, 94 patients (47%) died from noncardiac morbidities such as stroke, respiratory failure, and renal failure. In patients with multivessel and/or left main disease undergoing dialysis, 5-year outcomes revealed that CABG relative to PCI reduced the risk of cardiac death, sudden death, myocardial infarction, and any revascularization. However, the risk of all-cause death was not different between PCI and CABG.
Several studies have shown that chronic kidney disease is associated with increased mortality after coronary revascularization, particularly in patients with end-stage renal disease (ESRD) requiring chronic dialysis. Several randomized trials have reported the outcomes after revascularization in complex coronary lesions. However, patients with ESRD undergoing chronic dialysis have been excluded from any randomized evaluation for the comparative efficacy of coronary revascularization strategies. Regarding observational studies comparing percutaneous coronary intervention (PCI) with coronary artery bypass grafting (CABG) in the drug-eluting stents (DES) era, only a few single-center studies with less complex coronary lesions with small sample size have been available so far. In the present study, we sought to investigate 5-year outcomes comparing PCI with CABG in ESRD patients on chronic dialysis with multivessel coronary and/or left main coronary disease using a large observational database in Japan.
Methods
The Coronary REvascularization Demonstrating Outcome Study in Kyoto (CREDO-Kyoto) PCI/CABG Registry Cohort-2 is a physician-initiated, noncompany-sponsored, multicenter registry that enrolled consecutive patients undergoing first coronary revascularization in 26 centers in Japan from January 2005 through December 2007. The relevant ethics committees in all 26 participating centers (see Supplementary Data A ) approved the research protocol. Because of retrospective enrollment, written informed consent from the patients was waived. However, patients who refused participation in the study when contacted for follow-up were excluded.
The study design and patient enrollment in the registry have been described in detail previously. Of 15,939 patients enrolled in the registry, the study population for the present subanalysis of the CREDO-Kyoto PCI/CABG Registry Cohort-2 consisted of 388 patients with ESRD with multivessel and/or left main coronary artery disease on long-term dialysis (258 patients with PCI and 130 patients with isolated CABG) excluding those patients who refused study participation (n = 99), who had concomitant noncoronary surgery (n = 609), who had acute myocardial infarction (MI; n = 4,892), who had single-vessel disease (n = 3,431), and those without chronic dialysis (n = 6,520).
Demographic, angiographic, and procedural data were collected from hospital charts according to the prespecified definitions by experienced research coordinators in an independent research organization (Research Institute for Production Development, Kyoto, Japan; see Supplementary Data B ). Definitions for clinical characteristics are described in Supplementary Data C .
The Synergy between Percutaneous Coronary Intervention with Taxus and Cardiac Surgery (SYNTAX) score was calculated in those with 3-vessel and/or left main coronary artery disease, in whom diagnostic coronary angiograms for both left and right coronary arteries were available for analysis. The SYNTAX scores were available in 119 patients (98%) for PCI and 106 patients (90%) for CABG, respectively. The SYNTAX score was calculated using the SYNTAX score calculator (available at http://www.syntaxscore.com ) by a dedicated SYNTAX score committee (see Supplementary Data D ) in a blinded fashion to the clinical data. Intra- and inter-observer variabilities of the SYNTAX score calculation in our group were previously reported.
Collection of follow-up information was conducted mainly through review of inpatient and outpatient hospital charts by clinical research coordinators in the independent research organization. Additional follow-up information was collected through contact with patients, relatives, and/or referring physicians by sending mail with questions on vital status and additional hospitalizations. Death, MI, stent thrombosis, and stroke were adjudicated by the clinical events committee (see Supplementary Data E ).
The primary outcome measure for the present analysis was death from any cause. Other prespecified end points included cardiac death, sudden death, stroke, MI, and any coronary revascularization. Death was regarded as cardiac in origin unless obvious noncardiac causes could be identified. Any death during the index hospitalization for coronary revascularization was regarded as cardiac death. Sudden death was defined as unexpected death in previously stable patients. MI was defined according to the definition in the Arterial Revascularization Therapy Study. Stroke during follow-up was defined as ischemic or hemorrhagic stroke requiring hospitalization with symptoms lasting >24 hours. Scheduled staged coronary revascularization procedures performed within 3 months of the initial procedure were not regarded as follow-up events but were included in the index procedure.
All continuous variables are expressed as the mean ± SD. Differences in baseline characteristics between the 2 groups were examined by unpaired t test and Fisher’s exact test. Cumulative incidence was estimated by the Kaplan-Meier method and differences were assessed using log-rank test. Propensity scores, which were the probabilities that a patient would undergo PCI, were estimated with multivariate logistic regression analyses including age, gender, body mass index, hypertension, dyslipidemia, diabetes mellitus, current smoker, heart failure mitral regurgitation grade 3 or 4, previous MI, previous stroke, peripheral arterial disease, atrial fibrillation, chronic kidney disease, hemodialysis, anemia, platelet count, chronic obstructive lung disease, liver cirrhosis, malignancy, emergency procedure, number of diseased vessels, left main disease, target of chronic total occlusion, target of proximal left anterior descending coronary artery, and institute. These variables were consistent with previous reports from the current registry. Continuous variables except age were dichotomized using clinically meaningful reference values or median values. We incorporated the 26 participating centers in the propensity score estimation as the stratification variable. The hazard ratios (HRs) of PCI compared with CABG were estimated by the stratified Cox proportional hazard models; the models included PCI or CABG as the covariate and were stratified by the quartiles of propensity score and institute to adjust for confounding. Effects of PCI compared with CABG for individual end points were expressed as HRs with 95% confidence intervals. All reported p values were 2-sided, and p values <0.05 were regarded as statistically significant.
All analyses were conducted by a statistician with the use of SAS software, version 9.3 (SAS Institute Inc Cary, North Carolina), and S-Plus, version 7.0 (Insightful Corp). The investigators had full access to the data and take responsibility for its integrity. All investigators have read and agreed to the manuscript as written.
Results
Baseline clinical characteristics were not different between the PCI and CABG groups except the rate of atrial fibrillation ( Table 1 ). Regarding the complexity of coronary artery anatomy, the CABG group included more patients with 3-vessel disease, left main disease, target of proximal left anterior descending artery, and target of chronic total occlusion. SYNTAX score was significantly higher in the CABG group. Median follow-up duration for the surviving patients was 1,821 days.
| Variable | PCI (n = 258) | CABG (n = 130) | p Value |
|---|---|---|---|
| Age | 66.2 ± 10.6 | 66.5 ± 8.7 | 0.75 |
| >75 years | 61 (24%) | 23 (18%) | 0.18 |
| Men | 187 (73%) | 104 (80%) | 0.11 |
| Body mass index (kg/m 2 ) | 22.3 ± 3.9 | 21.8 ± 2.9 | 0.20 |
| >25 | 207 (80%) | 111 (85%) | 0.21 |
| Hypertension | 229 (89%) | 107 (82%) | 0.08 |
| Diabetes mellitus | 167 (65%) | 77 (59%) | 0.29 |
| On insulin therapy | 80 (31%) | 37 (29%) | 0.61 |
| Current smoker | 46 (18%) | 20 (15%) | 0.55 |
| Ejection fraction (%) | 53.3 ± 14.0 | 52.1 ± 14.0 | 0.43 |
| Mitral regurgitation grade 3/4 | 28 (11%) | 9 (6.9%) | 0.21 |
| Previous myocardial infarction | 45 (17%) | 24 (19%) | 0.80 |
| Heart failure | 87 (34%) | 43 (33%) | 0.90 |
| Atrial fibrillation | 33 (13%) | 29 (22%) | 0.02 |
| Previous stroke | 40 (16%) | 23 (18%) | 0.58 |
| Peripheral artery disease | 68 (26%) | 39 (30%) | 0.45 |
| Anemia (hemoglobin <11.0 g/dl) | 155 (60%) | 73 (56%) | 0.46 |
| Platelet count <100 × 10 9 /L) | 16 (6%) | 8 (6%) | 0.99 |
| Chronic obstructive pulmonary disease | 2 (1%) | 2 (2%) | 0.48 |
| Liver cirrhosis | 13 (5%) | 6 (5%) | 0.86 |
| Malignancy | 18 (7%) | 11 (9%) | 0.60 |
| Procedural characteristics | |||
| Number of target coronary lesions or anastomoses | 1.7 ± 0.8 | 3.0 ± 1.1 | <0.001 |
| Number of coronary artery narrowed | 1.6 ± 0.7 | 2.4 ± 0.6 | <0.001 |
| Extent of coronary artery disease | |||
| 3 | 97 (38%) | 74 (57%) | <0.001 |
| 2 | 136 (52%) | 12 (9%) | <0.001 |
| Left main | 25 (10%) | 44 (34%) | <0.001 |
| Target of proximal left anterior descending artery | 133 (52%) | 109 (84%) | <0.001 |
| Target of chronic total occlusion | 49 (19%) | 44 (34%) | <0.001 |
| PCI profile | |||
| Drug-eluting stent use | 190 (73%) | (-) | n/a |
| Bare-metal stent use only | 46 (18%) | (-) | n/a |
| Balloon angioplasty only | 22 (9%) | (-) | n/a |
| Off-pump CABG | (-) | 78 (60%) | n/a |
| Emergency procedure | 7 (3%) | 8 (6%) | 0.10 |
| SYNTAX score ∗ | 23.5 ± 8.7 | 29.4 ± 11.0 | <0.001 |
| Medications at discharge | |||
| Ticlopidine/clopidogrel | 237 (93%) | 14 (11%) | <0.001 |
| Aspirin | 251 (97%) | 123 (95%) | 0.18 |
| Statins | 63 (24%) | 12 (9%) | <0.001 |
| Beta-blockers | 57 (22%) | 28 (22%) | 0.90 |
| Angiotensin converting enzyme inhibitor/angiotensin receptor blocker | 119 (46%) | 45 (35%) | 0.03 |
| Nitrates | 98 (38%) | 41 (32%) | 0.21 |
| Calcium channel blocker | 139 (54%) | 62 (48%) | 0.25 |
| Warfarin | 16 (6%) | 51 (39%) | <0.001 |
∗ SYNTAX scores were available in 119 patients for PCI (98%) and 106 patients for CABG (90%), respectively.
Unadjusted 30-day mortality was 2.7% for PCI and 5.4% for CABG ( Table 2 ). Cumulative 5-year incidence of all-cause death was not significantly different between the PCI and CABG groups ( Figure 1 ). Similarly, cumulative 5-year incidences of cardiac death, sudden death, and stroke were not different between the 2 groups ( Figure 1 ). Cumulative incidences of MI and any coronary revascularization after PCI were significantly higher than those after CABG ( Figure 1 ).
| Variable | PCI n = 258 | CABG n = 130 |
|---|---|---|
| Death (total) | 7 (3%) | 7 (5%) |
| Heart failure | 2 | 1 |
| Respiratory failure | 0 | 3 |
| Peripheral artery disease | 2 | 0 |
| Gastrointestinal | 0 | 2 |
| Stroke | 1 | 0 |
| Renal failure | 1 | 0 |
| Infection | 1 | 0 |
| Nonfatal stroke | 4 (2%) | 2 (2%) |
| Nonfatal myocardial infarction | 4 (2%) | 1 (1%) |
After adjusting for the propensity score and institute, the risk of PCI relative to CABG for all-cause death remained statistically insignificant ( Table 3 ). However, adjusted risks for cardiac death and sudden death after PCI were significantly higher than after CABG. The risk for any coronary revascularization after PCI was markedly higher than after CABG. The risk for stroke was not significantly different between the 2 groups.
