Studies evaluating long-term (≥5 years) outcomes of percutaneous coronary intervention (PCI) using drug-eluting stents compared with coronary artery bypass grafting (CABG) in patients with triple-vessel coronary artery disease (TVD) are still limited. We identified 2,978 patients with TVD (PCI: n = 1,824, CABG: n = 1,154) of 15,939 patients with first coronary revascularization enrolled in the Coronary Revascularization Demonstrating Outcome Study in Kyoto PCI/CABG Registry Cohort-2. The primary outcome measure in the present analysis was a composite of death, myocardial infarction (MI), and stroke. Median follow-up duration for the surviving patients was 1,973 days (interquartile range 1,700 to 2,244). The cumulative 5-year incidence of death/MI/stroke was significantly higher in the PCI group than in the CABG group (28.2% vs 24.0%, log-rank p = 0.006). After adjusting for confounders, the excess risk of PCI relative to CABG for death/MI/stroke remained significant (hazard ratio [HR] 1.38, 95% confidence interval [CI] 1.13 to 1.68, p = 0.002). The excess risks of PCI relative to CABG for all-cause death, MI, and any coronary revascularization were also significant (HR 1.38, 95% CI 1.10 to 1.74, p = 0.006; HR 2.81, 95% CI 1.69 to 4.66, p <0.001; and HR 4.10, 95% CI 3.32 to 5.06, p <0.001, respectively). The risk for stroke was not significantly different between the PCI and CABG groups (HR 0.88, 95% CI 0.61 to 1.26, p = 0.48). There were no interactions for the primary outcome measure between the mode of revascularization (PCI or CABG) and the subgroup factors such as age, diabetes, and Synergy Between PCI With Taxus and Cardiac Surgery score. In conclusion, CABG compared with PCI was associated with better long-term outcome in patients with TVD.
The Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgery (SYNTAX) randomized trial recently reported that 5-year cardiovascular outcome in patients with triple-vessel coronary artery disease (TVD) was better after coronary artery bypass grafting (CABG) than after percutaneous coronary intervention (PCI) using drug-eluting stents (DES). However, studies evaluating long-term (≥5 years) outcomes of PCI using DES compared with CABG in patients with TVD are still limited. We previously reported 3-year outcome in patients with TVD from the Coronary Revascularization Demonstrating Outcome Study in Kyoto (CREDO-Kyoto) PCI/CABG Registry Cohort-2, demonstrating that (1) consistent with the SYNTAX randomized trial results, PCI compared with CABG was associated with significantly higher risk for serious cardiovascular events; and (2) in contradiction to the SYNTAX randomized trial results, the benefit of using the SYNTAX score for risk stratification could not be clearly demonstrated. In the present study, we extended the follow-up to evaluate long-term (≥5 years) outcome of PCI compared with CABG in patients with TVD in real-world clinical practice.
Methods
As previously described in detail, the CREDO-Kyoto PCI/CABG Registry Cohort-2 is a physician-initiated non–company-sponsored multicenter registry enrolling consecutive patients who underwent first coronary revascularization among 26 centers in Japan from January 2005 to December 2007. The relevant ethics committees in all 26 participating centers ( Supplementary data A ) approved the research protocol. Because of retrospective enrollment, written informed consents from the patients were waived; however, we excluded those patients who refused participation in the study when contacted for follow-up. This strategy is concordant with the guidelines of the Japanese Ministry of Health, Labor, and Welfare.
Of 15,939 patients enrolled in the registry, we excluded 99 patients who refused to participate in the study, 609 patients with concomitant noncoronary surgery, 4,892 patients with acute myocardial infarction (MI) presentation, 6,357 patients with single- or double-vessel disease, and 1,004 patients with left main coronary artery disease. Therefore, the study population for the present analysis consisted of 2,978 patients with TVD (PCI: 1,824 patients and CABG: 1,154 patients; Figure 1 ).
The primary outcome measure in the present study was a composite of death, MI, and stroke. The secondary outcome measures included cardiac death, sudden cardiac death, noncardiac death, MI, stroke, hospitalization for heart failure, major bleeding, any coronary revascularization, and ischemia-driven coronary revascularization. Death was regarded as cardiac in origin unless obvious noncardiac causes could be identified. MI was defined according to the definition in the Arterial Revascularization Therapy Study. Stroke was defined as ischemic or hemorrhagic stroke either occurring during the index hospitalization or requiring hospitalization with symptoms lasting >24 hours. Hospitalization for heart failure was defined as hospitalization because of worsening heart failure requiring intravenous drug therapy. Bleeding was defined according to the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO) classification. GUSTO moderate or severe bleeding was adjudicated as a bleeding event. Any coronary revascularization was defined as either PCI or CABG for any reasons. 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. The detailed definitions of baseline clinical characteristics were described previously.
Experienced clinical research co-ordinators from the independent clinical research organization (Research Institute for Production Development, Kyoto, Japan; Supplementary data B ) collected baseline clinical, angiographic, and procedural characteristics from hospital charts or hospital databases according to prespecified definitions. Collection of follow-up information was mainly conducted through review of inpatient and outpatient hospital charts by the clinical research co-ordinators, and additional follow-up information was collected through contact with patients, relatives, and/or referring physicians by sending mail with questions regarding vital status, subsequent hospitalizations, and status of antiplatelet therapy. Death, MI, and stroke were adjudicated by the clinical event committee ( Supplementary data C ). Median follow-up duration for the surviving patients was 1,973 days (interquartile range 1,700 to 2,244). Complete 1-, 3-, and 5-year clinical follow-up information was obtained for 97.8%, 95.8%, and 71.4% of patients, respectively.
Categorical variables were presented as number and percentage and were compared with the chi-square test. Continuous variables were expressed as mean ± SD or median with IQR. Continuous variables were compared using the Student’s t test or Wilcoxon rank-sum test based on their distributions. Cumulative incidence was estimated by the Kaplan-Meier method, and differences were assessed with the log-rank test. The effects of PCI relative to CABG for the individual end points were expressed as hazard ratios (HRs) and their 95% confidence intervals (CIs). We estimated the HR by Cox proportional hazard models adjusting for 30 clinically relevant factors listed in Table 1 . Continuous variables were dichotomized by clinically meaningful reference values or median values. Proportional hazard assumptions for potential independent risk-adjusting variables were assessed on the plots of log (time) versus log [−log (survival)] stratified by the variable, and the assumptions were verified to be acceptable for all the variables. We incorporated the 26 participating centers in the Cox proportional hazard models as the stratification variable. As prespecified subgroup analyses, clinical outcomes were compared between the PCI and CABG groups stratified by age (≥75 and <75 years), diabetic status, and the SYNTAX score (<23, ≤23 to <33, ≥33). The same Cox proportional hazard models were conducted for the primary outcome measure in each subgroup analysis.
| PCI(n=1824) | CABG(n=1154) | P value | |
|---|---|---|---|
| (A) Clinical characteristics | |||
| Age (years) | 69.7±10.0 | 68.0±8.9 | <0.001 |
| Age >=75 years ∗ | 642(35%) | 305(26%) | <0.001 |
| Men ∗ | 1296(71%) | 844(73%) | 0.22 |
| Body mass index <25.0 (kg/m 2 ) ∗ | 1198(66%) | 809(70%) | 0.01 |
| Unstable angina pectoris | 181(10%) | 96(8.3%) | 0.14 |
| Hypertension ∗ | 1593(87%) | 970(84%) | 0.01 |
| Diabetes mellitus ∗ | 911(50%) | 642(56%) | 0.003 |
| on insulin therapy | 251(14%) | 216(19%) | <0.001 |
| Current smoker ∗ | 462(25%) | 279(24%) | 0.48 |
| Heart failure ∗ | 376(21%) | 256(22%) | 0.31 |
| Ejection fraction <=40% | 198(12%) | 161(15%) | 0.08 |
| Mitral regurgitation grade 3/4 ∗ | 108(5.9%) | 36(3.1%) | <0.001 |
| Prior myocardial infarction ∗ | 345(19%) | 291(25%) | <0.001 |
| Prior stroke (symptomatic) ∗ | 292(16%) | 173(15%) | 0.46 |
| Peripheral vascular disease ∗ | 211(12%) | 151(13%) | 0.22 |
| Estimated glomerular filtration rate <30 (ml/min/1.73m 2 ) , without hemodialysis ∗ | 103(5.7%) | 101(8.8%) | 0.001 |
| Hemodialysis ∗ | 97(5.3%) | 74(6.4%) | 0.21 |
| Anemia (Hemoglobin <11.0g/dl) ∗ | 283(16%) | 218(19%) | 0.02 |
| Thrombocytopenia (Platelet<10 ∗ 10 4 ) ∗ | 30(1.6%) | 22(1.9%) | 0.6 |
| Chronic obstructive pulmonary disease ∗ | 60(3.3%) | 25(2.2%) | 0.07 |
| Liver cirrhosis ∗ | 62(3.4%) | 34(3.0%) | 0.49 |
| Malignancy ∗ | 192(11%) | 118(10%) | 0.79 |
| (B) Procedural characteristics | |||
| Number of target lesions or anastomoses | 2.1±1.0 | 3.5±1.1 | <0.001 |
| Target of proximal left anterior descending coronary artery ∗ | 1172(64%) | 1118(97%) | <0.001 |
| Target of chronic total occlusion ∗ | 415(23%) | 593(51%) | <0.001 |
| Emergency procedure | 103(5.7%) | 37(3.2%) | 0.002 |
| SYNTAX score | 23(17-29) | 29(22.5-37) | <0.001 |
| Low <23 | 873(49%) | 257(25%) | <0.001 |
| Intermediate 23-32 | 637(36%) | 387(38%) | |
| High >=33 | 280(16%) | 374(37%) | |
| Total number of stents | 2.8±1.7 | ― | ― |
| Total stent length (mm) | 62.0±40.0 | ― | ― |
| Stent use | 1725(95%) | ― | ― |
| Drug-eluting stents use | 1326(77%) | ― | ― |
| Internal thoracic artery use | ― | 1131(98%) | ― |
| Off Pump | ― | 725(63%) | ― |
| (C) Baseline Medications | |||
| Antiplatelet therapy | |||
| Thienopyridine | 1800(99%) | 109(9.5%) | <0.001 |
| Ticlopidine | 1641(92%) | 107(98%) | |
| Clopidogrel | 150(8.4%) | 2(1.8%) | |
| Aspirin | 1794(98%) | 1135(98%) | 0.997 |
| Cilostazol ∗ | 185(10%) | 95(8.2%) | 0.08 |
| Other medications | |||
| Statins ∗ | 942(52%) | 349(30%) | <0.001 |
| Beta-blockers ∗ | 557(31%) | 300(26%) | 0.007 |
| Angiotensin converting enzyme inhibitor/ angiotensin receptor blocker ∗ | 1027(56%) | 346(30%) | <0.001 |
| Nitrates ∗ | 804(44%) | 391(34%) | <0.001 |
| Calcium channel blockers ∗ | 954(52%) | 579(50%) | 0.26 |
| Nicorandil ∗ | 479(26%) | 460(40%) | <0.001 |
| Warfarin ∗ | 150(8.2%) | 429(37%) | <0.001 |
| Proton pump inhibitors ∗ | 403(22%) | 470(41%) | <0.001 |
| Histamine type-2 receptor blockers ∗ | 425(23%) | 401(35%) | <0.001 |
∗ Risk adjusting variables selected for Cox proportional hazard models.
All statistical analyses were conducted using JMP 8.0 software (SAS Institute Inc., Cary, North Carolina) and SAS 9.2 (SAS Institute Inc.). All reported p values were 2 tailed, and p values <0.05 were considered statistically significant.
Results
Baseline characteristics were significantly different in several aspects between the PCI and CABG groups ( Table 1 ). The PCI group had significantly higher prevalence of patients with advanced age, hypertension, and severe mitral regurgitation, whereas the CABG group more often had patients with smaller body mass index, diabetes, previous MI, renal failure, and anemia. Regarding angiographic and procedural findings, the median SYNTAX score in the CABG group was significantly higher than that in the PCI group, and the CABG group had greater number of target lesions or anastomoses ( Table 1 ). Antiplatelet agents and evidence-based medicines, such as angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta blockers, and statins, were more often prescribed in the PCI group than in the CABG group ( Table 1 ).
The cumulative 5-year incidence of death/MI/stroke (primary outcome measure) in the PCI group was significantly higher than that in the CABG group (28.2% vs 24.0%, log-rank p = 0.006; Figure 2 , Table 2 ). After adjusting for confounders, the excess risk of PCI relative to CABG for the primary outcome measure remained significant (HR 1.38; 95% CI 1.13 to 1.68, p = 0.002). The cumulative 5-year incidence of and the adjusted risk for all-cause death were significantly higher in the PCI group than in the CABG group ( Figure 2 , Table 2 ). PCI compared with CABG tended to be associated with higher risk for cardiac death, although the cumulative 5-year incidence of cardiac death and sudden cardiac death were not significantly different between the 2 groups. The adjusted risks for MI and any coronary revascularization were also higher in the PCI group than in the CABG group. The risk for stroke was not significantly different between the PCI and CABG groups. PCI compared with CABG tended to be associated with higher risk for hospitalization for heart failure. The risk for major bleeding beyond 30 days after procedure was not significantly different between the 2 groups, although overall risk for major bleeding including bleeding events during the perioperative period was significantly higher in the CABG group than in the PCI group ( Table 2 ).
