Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Online Materials 1 2 Full title: Coronary Artery Bypass Graft Surgery and Percutaneous Coronary 3 Interventions in Patients with Unprotected Left Main Coronary Artery Disease 4 Contents 5 Ⅰ Online Methods 6 Data source and management 7 Revascularization procedures 8 Definitions 9 Ⅱ Online Tables 10 Table 1. Indicators for either CABG or PCI procedure 11 Table 2. Candidate patient-level covariates for adjustment 12 Table 3. Discrimination and calibration results for multivariable logistic regression 13 models 14 Table 4. Proportional hazards assumption of the Cox regressions 15 Table 5. Baseline characteristics of patients after adjustment with inverse 16 probability-of-treatment weighting 17 Table 6. 30-day outcomes for patients treated with CABG and PCI 18 Table 7. Baseline characteristics of patients with baseline SYNTAX scores 19 Table 8. Baseline characteristics of patients without contraindications to CABG 20 Table 9. Procedural characteristics of patients with baseline SYNTAX scores 21 Table 10. Procedural characteristics of patients without contraindications to CABG 22 Table 11. 30-day and 3-year outcomes for patients with baseline SYNTAX scores 23 Table 12. 30-day and 3-year for patients without contraindications to CABG 24 Ⅲ Online Figures 1 1 Figure 1. Patient Flowchart. 2 ULMD = unprotected left main disease; CABG = coronary artery bypass graft; PCI 3 = percutaneous coronary intervention; STEMI = ST-segment elevation myocardial 4 infarction. 5 6 7 8 9 10 11 12 13 14 15 16 Figure 2. Revascularization Strategy Selection for Patients with ULMD at Our Institution During 2004 and 2010. ULMD = unprotected left main disease; CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention. Figure 3. Unadjusted Kaplan-Meier Curves of 3-Year Outcomes According to Revascularization Strategy and SYNTAX Score. (A) patients with low to intermediate-risk SYNTAX score (≤32); (B) patients with high-risk SYNTAX score (>32). CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; MI = myocardial infarction. Figure 4. Adjusted Survival Curves of 3-Year Outcomes According to Revascularization Strategy and SYNTAX Score. 17 (A) patients with low to intermediate-risk SYNTAX score (≤32); (B) patients with 18 high-risk SYNTAX score (>32). The HRs were reported for patients undergoing PCI 19 with those undergoing CABG as reference. 20 21 22 23 CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; MI = myocardial infarction. Figure 5. Unadjusted Kaplan-Meier Curves of 3-Year Outcomes According to Revascularization Strategy and Diabetes Mellitus. 24 (A) patients without diabetes mellitus; (B) patients with diabetes mellitus. 25 CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; 2 1 2 3 MI = myocardial infarction. Figure 6. Adjusted Survival Curves of 3-Year Outcomes According to Revascularization Strategy and Diabetes Mellitus. 4 (A) patients without diabetes mellitus; (B) patients with diabetes mellitus. The HRs 5 were reported for patients undergoing PCI with those undergoing CABG as reference. 6 CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; 7 8 9 10 11 12 13 14 15 16 MI = myocardial infarction. Figure 7. Unadjusted Hazard Ratios of Clinical Outcomes According to Selected Major Subgroups. a Data are Kaplan-Meier estimates of cumulative event rate at 3 years, expressed as percent of patients. b ULM bifurcation and SYNTAX Score subsets analyses were conducted in 2,752 patients whose baseline angiogram was available. CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; MI = myocardial infarction; LVEF = left ventricular ejection fraction. Figure 8. Patients Distribution and 3-Year All-cause Mortality According to 17 Revascularization Strategy and SYNTAX Score. 18 Patients were divided into low or intermediate risk (≤32) and high risk (>32) groups 19 according to the SYNTAX score in the PCI and the CABG group. Data are expressed 20 as percent of patients. 21 CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention. 22 3 1 Online Methods 2 Data source and management 3 The two primary databases used in the study were our institutional CABG registry and 4 PCI registry. These registries were developed in 1999 and 1997, respectively, after the 5 approval of the institutional review board, for the purpose of collecting information on 6 all consecutive patients who had undergone CABG or PCI in our institution. The two 7 registries contain information on demographics, comorbidities, hemodynamic state, 8 left ventricular function, diseased vessels and vessels for which bypass or angioplasty 9 was attempted, operator identifiers, and in-hospital adverse outcomes. The PCI 10 registry also contains information on the types of device used for each patient, 11 including bare-metal stents and drug-eluting stents. Uniform definitions for these 12 elements are used in our study. These clinical, angiographic, procedural or operative, 13 and outcome data were collected with the use of a dedicated computer-based reporting 14 system. We performed routine data auditing to ensure timeliness, completeness and 15 accuracy of the data. 16 For in-hospital data collection, we established 11 modules with 392 items, among 17 which 184 items composed the core dataset and 208 composed the extended dataset. 18 The core dataset is a mandatory set of variables believed to be necessary for the 19 accurate representation of a clinical practice, risk model development, and analysis. 20 The extended dataset includes less critical but still important fields that are strongly 21 encouraged to be collected but are not mandatory for submission. The creation of 22 custom fields is permissible such that the sites can collect specific data that are 23 desirable. The complete in-hospital data is requested to be submitted within 30 days 24 after CABG or PCI. 25 Data variables have been added for internal quality and integrity checks. If any 4 1 component of the required mandatory data elements is missing or any of the 2 validation roles of logic are not meet, then the data will be automatically rejected. The 3 Medical Research & Biometrics Center (MRBC) of National Center for 4 Cardiovascular Diseases (NCCD) back-end checks data for completeness and integrity. 5 To further audit the accuracy of the data, we randomly select 5-10% of in-hospital 6 data annually. Two qualified physicians from the MRBC abstract all of the medical 7 record data independently. After discussion between the two physicians, the final 8 standard data are established. Next, we compare the data between MRBC data 9 abstraction and the database. We defined the accuracy of the data elements as present 10 when the standard data and the submitted data are exactly the same or both are 11 missing for categorical variables. For continuous variables, we defined the accuracy 12 of the data as it was within ±10% of the value recorded in the standard data. The 13 accuracy report is fed back to each data collection staff, and we also request that staffs 14 review and correct the data in error. 15 5 1 Revascularization procedures 2 1. CABG 3 Anesthesia was managed by inhalation of isoflurane with the addition of fentanyl or 4 sufentanil, and propofol was administered continuously until the end of the procedure 5 if necessary. Surgical revascularization was performed using standard bypass 6 techniques. CABG procedures were performed under cardiopulmonary bypass or 7 off-pump, at the surgeon’s discretion. For on-pump CABG, a standard 8 cardiopulmonary bypass was established, and moderate systemic hypothermia (28°C 9 to 32°C) and perfusion with antegrade intermittent cold crystalloid cardioplegia were 10 used. Heparin was given to achieve activated clotting times of 480 seconds or above 11 before institution of cardiopulmonary bypass. For off-pump CABG, stabilization 12 devices were used to provide a motionless anastomosis site, and heparin was 13 administered before the start of the first distal anastomosis to achieve an activated 14 clotting time of 300 to 350 seconds. On-pump CABG involved aortic cross-clamping 15 and cardioplegic arrest, while off-pump CABG was performed with a partial 16 occlusion clamp. Whenever feasible, Y or T grafts were used to avoid partial aortic 17 clamping when performing proximal anastomoses. Whenever possible, complete 18 revascularization was attempted, and the internal thoracic artery was used 19 preferentially for revascularization of the left anterior descending artery. The 20 remaining vessels were to be bypassed either using another arterial conduit or the 21 saphenous vein in the configuration decided by the surgeon. During reperfusion, the 22 bypass grafting was completed with proximal anastomoses to the ascending aorta. The 23 decision to switch to cardiopulmonary bypass during the procedure was based on 24 significant hemodynamic instability or ventricular arrhythmia. After separation from 25 cardiopulmonary bypass or on completion of all anastomoses, protamine was given to 6 1 reverse the effects of heparin. Postoperatively, starting within the first 24 hours, 2 aspirin therapy (100 mg/d) was recommended and should be continued indefinitely. 3 2. PCI 4 All patients undergoing PCI received aspirin plus clopidogrel (loading dose, 300 or 5 600 mg) before the coronary intervention. Vascular access method and the choice of 6 the specific type of drug-eluting stents (ie, sirolimus- or paclitaxel- or everolimus- or 7 zotarolimus-eluting stents) was left to the interventionalist’s discretion. Lesions at the 8 ostium or shaft without involvement of ULM bifurcation were usually treated with 9 single stents. ULM bifurcation lesions were treated using different stenting strategies 10 including provisional stenting or 2-stent technique in the vast majority of cases. Stent 11 strategies to treat distal bifurcation lesions included: crossover stenting with side 12 branch balloon angioplasty, provisional or dedicated T stenting, simultaneous kissing 13 or V stenting (TF approach only), Culotte or crush technique (including “step crush” 14 technique involving sequential balloon crushing of side branch stent followed by main 15 vessel stenting). Final kissing balloon post-dilation was performed in cases with 16 suboptimal results after crossover stenting at the side branch ostium and, in most 17 cases, with 2-stent implantation. During the procedure, unfractionated heparin (100 18 U/kg) was administered to all patients. Proximal optimization technique (POT) with 19 additional bigger balloons was performed to optimize stent apposition. Intravascular 20 ultrasound was recommended to assess the baseline characteristics and the final 21 results. Adjunctive devices including cutting balloon angioplasty or rotablator were 22 used in selected patients with severe calcified or fibrous plaque. Glycoprotein IIb/IIIa 23 antagonists and an intra-aortic balloon pump were used if clinically indicated. After 24 the procedure, aspirin plus clopidogrel therapy was continued indefinitely. After the 25 procedure, aspirin was prescribed at a dose of 300 mg daily for 3 months, followed by 7 1 100 mg daily indefinitely; clopidogrel 75 mg daily was prescribed for at least 1 year. 2 For patients with stents requiring anticoagulation (chronic atrial fibrillation for 3 example), it was recommended to follow the American College of 4 Cardiology/American Heart Association ST elevation myocardial infarction 5 guidelines for triple therapy after stenting. 6 8 1 Definitions 2 Death was defined as death from any cause. 3 Myocardial infarction occurred when there were clinical signs and symptoms of 4 ischemia that were distinct from the presenting ischemic event and meeting at least 1 5 of the following criteria: 6 1. Spontaneous (>48 h after PCI, and/or after CABG) 7 A. New, significant Q waves in at least 2 contiguous leads of an ECG that were not 8 present with the presenting ischemic event; 9 B. Patients whose most recent cardiac markers measured before reinfarction, which 10 were normal, require an increase in CK-MB or troponin above the 99th percentile 11 limit of normal and at least ≥20% above the most recent value. 12 2. Within 48 h after PCI: 13 A. Patients with normal biomarker values (preprocedure) who then develop an 14 increase in biomarker values >5 times the 99th percentile ULN or if the baseline 15 values are elevated and are stable or falling, a rise of cTn values ≥20%. In addition, 16 symptoms suggestive of myocardial ischemia or new ischemic electrocardiographic 17 changes or angiographic findings consistent with a procedural complication or 18 imaging demonstration of new loss of viable myocardium are required. 19 B. Stent thrombosis associated with MI when detected by coronary angiography or 20 autopsy in the setting of myocardial ischemia and with a rise and/or fall of cardiac 21 biomarker values with at least 1 value above the 99th percentile ULN. 22 C. For patients with elevated baseline (preprocedure) cardiac biomarkers, there are 2 23 possible scenarios. In these scenarios, electrocardiographic changes or symptoms are 24 not required to qualify. 25 i. Patients with cardiac markers above the ULN (preprocedure) assumed to be in 9 1 the midst of an acute MI. 2 ii. Patients with elevated biomarkers with a characteristic rise and fall in 3 biomarker levels preprocedure most likely have completed their presenting infarct. 4 Further rises in cardiac markers must be ≥20% above the most recent value to be 5 coded as reinfarction. 6 D. Patients with new, significant Q waves in at least 2 contiguous leads of an ECG 7 that were not present with the presenting ischemic event. 8 9 3. Within 48 h after CABG: A CABG-related MI was defined by elevation of cardiac biomarker values >5 times 10 the 99th percentile upper reference limit in patients with normal baseline cardiac 11 troponin values (≤99th percentile upper reference limit) plus either new pathological 12 Q waves; new left bundle-branch block, angiographically documented new graft, or 13 native coronary artery occlusion; or imaging evidence of new loss of viable 14 myocardium or new regional wall motion abnormality. 15 Stroke was confirmed by a neurologist on the basis of imaging studies and was 16 defined as follows: 17 1. A focal neurologic deficit of central origin lasting >72 hours, or 18 2. A focal neurologic deficit of central origin lasting >24 hours, with imaging 19 evidence of cerebral infarction or intracerebral hemorrhage, or 20 3. A nonfocal encephalopathy lasting >24 hours with imaging evidence of cerebral 21 infarction or hemorrhage adequate to account for the clinical state, or 22 Retinal arterial ischemia or hemorrhage is included in the definition of stroke. 23 Repeat revascularization was defined as any repeat PCI or CABG. All stages of a 24 staged index PCI procedure will be considered part of the index revascularization 25 procedure and not a repeated revascularization. 26 10 1 Online Tables 2 Table 1. Indicators for Either CABG or PCI Procedure Indicators for Procedure CABG Complex coronary anatomy Unable to take anti-platelet medications Patient refused PCI Serious allergic reaction to stainless steel or contrast agent History of coagulopathy PCI Suitable coronary anatomy for stenting End-stage COPD No graft material Patient refused CABG Emergent revascularization is necessary Age≥80 years Poor distal vessel Prior CABG 3 4 CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; COPD = chronic obstructive pulmonary disease. 5 6 11 1 Table 2. Candidate Patient-Level Covariates for Adjustment Risk Factors 2 3 1. Age 8. Dyslipidemia 15. eGFR* 2. Sex 9. COPD 16. Ejection fraction 3. Weight 10. PVD 17. No. of diseased vessel† 4. Height 11. Family history of CAD 18. SYNTAX score‡ 5. Smoker 12. Previous MI 19. Revascularization procedure 6. Diabetes mellitus 13. Previous PCI 20. Year of procedure§ 7. Hypertension 14. Unstable Angina *The eGFRs were estimated from serum creatinine values with the Chronic Kidney Disease Epidemiology Collaboration equation. 4 †No. of diseased vessel was included in the propensity score calculation as a covariate in 5 the overall cohort analysis, and age (<65 and ≥65 years), diabetes mellitus, ejection fraction 6 (≥50% and <50%) and EuroSCORE subsets. 7 8 9 10 11 ‡SYNTAX score was included in the propensity score calculation as a covariate in the SYNTAX-score subset and cohorts with coronary ULM bifurcation. §Year of surgery was included in the multivariable regression model as each class indicating a calendar of year. COPD = chronic obstructive pulmonary disease; PVD = peripheral vascular disease; CAD = 12 coronary artery disease; MI = myocardial infarction; PCI = percutaneous coronary 13 intervention; eGFR = estimated glomerular filtration rate; SYNTAX = Synergy between 14 Percutaneous Coronary Intervention with Taxus and Cardiac Surgery. 15 12 1 TTable 3. Discrimination and Calibration Results for Multivariable Logistic 2 RRegression Models p value for Models C statistic Hosmer-Lemeshow test Overall cohort 0.831 0.359 Patients with baseline SYNTAX Scores 0.871 0.148 Patients without contraindications to CABG 0.832 0.297 <65 years 0.830 0.148 ≥65 years 0.836 0.859 No diabetes 0.840 0.388 Diabetes 0.812 0.732 Ejection fraction ≥50% 0.822 0.004 Ejection fraction <50% 0.853 0.730 Two-vessel disease 0.766 0.875 Three-vessel disease 0.738 0.434 No bifurcation 0.939 0.581 Bifurcation 0.863 0.005 SYNTAX score ≤32 0.819 0.952 SYNTAX score >32 0.717 0.471 EuroSCORE ≤1 0.879 0.420 2≤ EuroSCORE ≤3 0.836 0.327 EuroSCORE >3 0.870 0.182 Subgroups 3 SYNTAX = Synergy between Percutaneous Coronary Intervention with Taxus and Cardiac 4 Surgery; CABG = coronary artery bypass graft; EuroSCORE, the European System for Cardiac 5 Operative Risk Evaluation. 6 13 1 Table 4. Proportional Hazards Assumption of the Cox Regressions Patient Group Outcome Overall cohort Patients with baseline SYNTAX Scores Patients without contraindications to CABG SYNTAX score ≤32 SYNTAX score >32 No diabetes Diabetes <65 years 14 p value All-cause death 0.172 Death, MI, and stroke 0.049 Cardiac death 0.910 Myocardial infarction 0.998 Stroke 0.592 Repeat revascularization 0.001 All-cause death 0.197 Death, MI, and stroke 0.064 Cardiac death 0.357 Myocardial infarction 0.774 Stroke 0.800 Repeat revascularization 0.004 All-cause death 0.089 Death, MI, and stroke 0.017 Cardiac death 0.768 Myocardial infarction 0.907 Stroke 0.801 Repeat revascularization 0.002 All-cause death 0.731 Death, MI, and stroke 0.705 Repeat revascularization 0.003 All-cause death 0.041 Death, MI, and stroke 0.013 Repeat revascularization 0.084 All-cause death 0.850 Death, MI, and stroke 0.309 Repeat revascularization 0.019 All-cause death 0.015 Death, MI, and stroke 0.061 Repeat revascularization 0.012 All-cause death 0.214 ≥65 years Male Female Ejection fraction ≥50% Ejection fraction <50% Two-vessel disease Three-vessel disease No bifurcation Bifurcation EuroSCORE ≤1 2≤ EuroSCORE ≤3 EuroSCORE >3 1 Death, MI, and stroke 0.003 All-cause death 0.528 Death, MI, and stroke 0.820 All-cause death 0.360 Death, MI, and stroke 0.035 All-cause death 0.202 Death, MI, and stroke 0.806 All-cause death 0.451 Death, MI, and stroke 0.414 All-cause death 0.235 Death, MI, and stroke 0.065 All-cause death 0.627 Death, MI, and stroke 0.301 All-cause death 0.058 Death, MI, and stroke 0.003 All-cause death 0.228 Death, MI, and stroke 0.091 All-cause death 0.542 Death, MI, and stroke 0.399 All-cause death 0.156 Death, MI, and stroke 0.041 All-cause death 0.505 Death, MI, and stroke 0.093 All-cause death 0.724 Death, MI, and stroke 0.974 SYNTAX = Synergy between Percutaneous Coronary Intervention with Taxus and Cardiac 2 Surgery; CABG = coronary artery bypass graft; EuroSCORE, the European System for 3 Cardiac Operative Risk Evaluation. 4 15 1 2 Table 5. Baseline Characteristics of Patients After Adjustment With Inverse Probability-of-Treatment Weighting CABG PCI (n = 2604) (n = 1442) p value Age, yrs 61.2 ± 11.7 61.3 ± 18.2 0.74 Women 20.6 20.1 0.57 Hypertension 61.9 61.3 0.54 Hyperlipidemia 55.5 56.8 0.23 Diabetes mellitus 29.5 29.9 0.72 Smoker 50.2 52.1 0.08 COPD 0.9 1.1 0.39 Family history of CAD 19.2 18.6 0.51 Body Mass Index* , g/m2 25.5 ± 3.8 25.5 ± 5.2 0.60 Creatinine, μmol/l† 85.1 ± 28.7 83.4 ± 30.9 0.89 Previous MI 33.0 34.5 0.15 Previous PCI 17.1 15.4 0.03 Peripheral vascular disease 5.2 4.8 0.53 Unstable angina 51.1 50.5 0.50 61.4 ± 10.0 61.5 ± 13.0 0.72 Left main only 2.9 3.0 Left main + 1-vessel disease 10.5 10.6 Left main + 2-vessel disease 24.3 22.9 Left main + 3-vessel disease 62.3 63.5 Variable Demographics Risk factors Cardiovascular conditions Ejection fraction, % Diseased coronary vessels 0.52 3 Values are mean ± SD or n (%). 4 *Calculated as weight in kilograms divided by height in meters squared. 5 †To convert μmol/l to mg/dl, multiply serum creatinine values by 0.0113. 6 CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; COPD = 7 chronic obstructive pulmonary disease; CAD = coronary artery disease; MI = myocardial 8 infarction. 16 Table 6. 30-Day Outcomes for Patients Treated With CABG and PCI p value Adjusted OR† (95% CI) p value 2.72 (1.11–6.68) 0.03 5.30 (2.83–9.94) < 0.001 52 (3.6) 1.23 (0.86–1.76) 0.26 2.14 (1.70–2.70) < 0.001 8 (0.3) 10 (0.7) 2.27 (0.89–5.75) 0.09 4.44 (2.35–8.41) < 0.001 Myocardial infarction 55 (2.1) 45 (3.1) 1.49 (1.00–2.23) 0.05 2.44 (1.88–3.18) < 0.001 Stroke 16 (0.6) 2 (0.1) 0.22 (0.05–0.98) 0.05 0.14 (0.04–0.50) 0.002 0 17 (1.2) Not effective 0.94 Not effective 0.98 CABG (n = 2604) PCI (n = 1442) Unadjusted OR (95% CI) All-cause death 8 (0.3) 12 (0.8) Death, MI, and stroke 77 (3.0) Cardiac death Outcome, No. (%)* Repeat revascularization *Data are Kaplan-Meier estimates of cumulative event rate at 3 years, expressed as percent of patients. †Inverse probability-of-treatment weighted Cox proportional-hazards regression was used with adjustment for all patient-level variables in Online Table 1. The HRs were reported for patients undergoing PCI with those undergoing CABG as reference. CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; OR = odds ratio; MI = myocardial infarction. 17 Table 7. Baseline Characteristics of Patients With Baseline SYNTAX Scores CABG PCI (n = 1310) (n = 1442) p value Age, yrs 62.0 ± 9.2 59.9 ± 10.7 < 0.001 Women 248 (18.0) 308 (21.4) 0.11 Hypertension 874 (66.7) 782 (54.2) < 0.001 Hyperlipidemia 866 (66.1) 722 (50.1) < 0.001 Diabetes mellitus 439 (33.5) 348 (24.1) < 0.001 Smoker 745 (56.9) 671 (46.5) < 0.001 COPD 15 (1.1) 12 (0.8) 0.38 Family history of CAD 268 (20.5) 168 (11.7) < 0.001 Body Mass Index* , g/m2 25.4 ± 2.9 25.7 ± 3.1 0.01 Creatinine, μmol/l† 86.8 ± 21.6 81.3 ± 18.7 < 0.001 Previous MI 512 (39.8) 340 (23.6) < 0.001 Previous PCI 156 (11.9) 324 (22.5) < 0.001 70 (5.3) 74 (5.1) 0.32 Unstable angina 609 ( 46.5) 962 (66.7) < 0.001 Ejection fraction, % 60.5 ± 8.5 63.1 ± 7.2 < 0.001 63 (4.8) 106 (7.4) Left main + 1-vessel disease 193 (14.7) 304 (21.1) Left main + 2-vessel disease 403 (30.8) 535 (37.1) Left main + 3-vessel disease 651 (49.7) 497 (34.5) Variable Demographics Risk factors Cardiovascular conditions Peripheral vascular disease Diseased coronary vessels Left main only < 0.001 Values are mean ± SD or n (%). *Calculated as weight in kilograms divided by height in meters squared. †To convert μmol/l to mg/dl, multiply serum creatinine values by 0.0113. CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; COPD = chronic obstructive pulmonary disease; MI = myocardial infarction; CAD = coronary artery disease. 18 Table 8. Baseline Characteristics of Patients Without Contraindications to CABG CABG PCI (n = 2551) (n = 1404) p value Age, yrs 61.9 ± 9.0 59.4 ± 10.3 < 0.001 Women 465 (18.2) 297 (21.2) 0.03 Hypertension 1637 (64.2) 763 (54.3) < 0.001 Hyperlipidemia 1505 (59.0) 706 (50.3) < 0.001 Diabetes mellitus 789 (30.9) 337 (24.0) < 0.001 Smoker 1364 (53.5) 659 (46.9) < 0.001 COPD - - - Family history of CAD 572 (22.4) 165 (11.8) < 0.001 Body Mass Index* , g/m2 25.5 ± 3.0 25.7 ± 3.1 0.006 Creatinine, μmol/l† 86.2 ± 22.8 81.1 ± 18.6 < 0.001 Previous MI 972 (38.1) 332 (23.6) < 0.001 Previous PCI 250 (9.8) 314 (22.4) < 0.001 146/2,551 (5.7) 70 (5.0) 0.33 Unstable angina 1135 (44.5) 933 (66.5) < 0.001 Ejection fraction, % 60.2 ± 8.2 63.2 ± 7.2 < 0.001 Left main only 18 (0.7) 106 (7.5) Left main + 1-vessel disease 121 (4.7) 298 (21.2) Left main + 2-vessel disease 407 (16.0) 522 (37.2) Left main + 3-vessel disease 2005 (78.6) 478 (34.0) Variable Demographics Risk factors Cardiovascular conditions Peripheral vascular disease Diseased coronary vessels < 0.001 Values are mean ± SD or n (%). *Calculated as weight in kilograms divided by height in meters squared. †To convert μmol/l to mg/dl, multiply serum creatinine values by 0.0113. CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; COPD = chronic obstructive pulmonary disease; CAD = coronary artery disease; MI = myocardial infarction. 19 Table 9. Procedural Characteristics of Patients With Baseline SYNTAX Scores Total No. /No. (%) of Patients CABG patients 1310 Off-pump procedure 689/1310 (52.6) 3.4 ± 0.8 Grafts per patient Mean venous graft 2.3 ± 0.9 Mean arterial graft 1.1 ± 0.5 IMA-to-LAD graft 1237 (94.4) Intraoperative IABP 3 (0.1) PCI patients 1442 Ad hoc PCI 92 (6.4) PCI procedure access Transradial approach 837 (58.0) Transfemoral approach 605 (42.0) Total number of stents per patient 2.2 ± 1.7 Mean diameter of stent, mm 3.4 ± 0.5 27.4 ± 16.0 Total stent length, mm Two-stent strategy 400 (33.9) Crush 271 (67.8) T-stent 55 (13.8) V- or kissing stent 52 (13.0) Culotte 22 (5.5) Final kissing balloon 591 (50.0) Dissection in target lesion 41 (2.8) Guidance with IVUS 560 (38.8) Use of glycoprotein IIb/IIIa antagonist 118 (8.2) Use of IABP 121 (8.4) Procedural success 1424 (98.8) Values are mean ± SD or n (%). CABG = coronary artery bypass graft; IMA = internal mammary artery; LAD = left anterior descending branch; IABP = intra-aortic balloon pump; PCI = percutaneous coronary intervention; IVUS = intravascular ultrasound. 20 Table 10. Procedural Characteristics of Patients Without Contraindications to CABG Total No. /No. (%) of Patients CABG patients 2551 Off-pump procedure 1,357 (53.2) Grafts per patient 3.4 ± 0.8 Mean venous graft 2.2 ± 0.9 Mean arterial graft 1.2 ± 0.7 IMA-to-LAD graft 2404 (94.2) Intraoperative IABP 4 (0.1) PCI patients 1404 Ad hoc PCI 92 (6.6) PCI procedure access Transradial approach 820/ (58.4) Transfemoral approach 584 (41.6) Total number of stents per patient 2.2 ± 1.2 Mean diameter of stent, mm 3.4 ± 0.5 27.6 ± 16.1 Total stent length, mm Two-stent strategy 395 (34.5) Crush 269 (68.1) T-stent 54 (13.7) V- or kissing stent 51 (12.9) Culotte 21 (5.3) Final kissing balloon 572 (49.9) Dissection in target lesion 41 (2.9) Guidance with IVUS 550 (39.2) Use of glycoprotein IIb/IIIa antagonist 114 (8.1) Use of IABP 118 (8.4) Procedural success 1387 (98.8) Values are mean ± SD or n (%). CABG = coronary artery bypass graft; IMA = internal mammary artery; LAD = left anterior descending branch; IABP = intra-aortic balloon pump; PCI = percutaneous coronary intervention; IVUS = intravascular ultrasound. 21 Table 11. 30-Day and 3-Year Outcomes for Patients With Baseline SYNTAX Scores 3-Year after the Procedure* 30-Day after the Procedure CABG (n = 1,310) PCI (n = 1,442) Adjusted OR† (95% CI) p value CABG (n = 1,310) PCI (n = 1,442) Adjusted HR† (95% CI) p value All-cause death 3 (0.2) 12 (0.8) 5.30 (2.83–9.94) < 0.001 32 (2.5) 55 (3.8) 1.27 (1.06–1.53) 0.01 Death, MI, and stroke 37 (2.8) 52 (3.6) 2.28 (1.75–2.98) < 0.001 120 (9.3) 112 (7.5) 1.04 (0.87–1.24) 0.67 Cardiac death 3 (0.2) 10 (0.7) 4.44 (2.35–8.41) < 0.001 15 (1.2) 37 (2.6) 3.35 (2.13–5.26) < 0.001 Myocardial infarction 29 (2.2) 45 (3.1) 1.60 (1.18–2.15) 0.002 40 (3.2) 61 (4.2) 1.78 (1.38–2.29) < 0.001 Stroke 7 (0.5) 2 (0.1) 0.14 (0.04–0.50) 0.002 57 (4.4) 18 (1.1) 0.22 (0.14–0.34) < 0.001 0 17 (1.2) Not effective 0.98 29 (2.4) 146 (9.9) 4.99 (3.76–6.61) < 0.001 Outcome, No. (%) Repeat revascularization *Data are Kaplan-Meier estimates of cumulative event rate at 3 years, expressed as percent of patients. †Inverse probability-of-treatment weighted Cox proportional-hazards regression was used with adjustment for all patient-level variables in Online Table 1. The HRs were reported for patients undergoing PCI with those undergoing CABG as reference. CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; OR = odds ratio; HR = hazard ratio; MI = myocardial infarction. 22 Table 12. 30-Day and 3-Year Outcomes for Patients Without Contraindications to CABG 3-Year after the Procedure* 30-Day after the Procedure CABG (n = 2551) PCI (n = 1404) Adjusted OR† (95% CI) p value CABG (n = 2551) PCI (n = 1404) Adjusted HR† (95% CI) p value All-cause death 8 (0.3) 12 (0.9) 5.49 (2.93–10.29) < 0.001 61 (2.4) 52 (3.7) 1.78 (1.36–2.31) < 0.001 Death, MI, and stroke 75 (2.9) 52 (3.7) 2.29 (1.81–2.88) < 0.001 228 (8.9) 107 (7.6) 0.98 (0.85–1.14) 0.82 Cardiac death 8 (0.3) 10 (0.7) 4.56 (2.41–8.64) < 0.001 33 (1.3) 35 (2.5) 2.55 (1.82–3.58) < 0.001 Myocardial infarction 54 (2.1) 45 (3.2) 2.60 (2.00–3.38) < 0.001 73 (2.9) 61 (4.3) 2.15 (1.72–2.69) < 0.001 Stroke 15 (0.6) 2 (0.1) 0.16 (0.04–0.55) 0.004 106 (4.2) 16 (1.1) 0.18 (0.12–0.26) < 0.001 0 16 (1.1) Not effective 0.98 48 (1.9) 139 (9.9) 4.98 (3.94–6.29) < 0.001 Outcome, No. (%) Repeat revascularization *Data are Kaplan-Meier estimates of cumulative event rate at 3 years, expressed as percent of patients. †Inverse probability-of-treatment weighted Cox proportional-hazards regression was used with adjustment for all patient-level variables in Online Table 1. The HRs were reported for patients undergoing PCI with those undergoing CABG as reference. CABG = coronary artery bypass graft; PCI = percutaneous coronary intervention; OR = odds ratio; HR = hazard ratio; MI = myocardial infarction. 23 Online Figures Figure 1. Patient Flowchart. 24 Figure 2. Revascularization Strategy Selection for Patients with ULMD at Our Institution During 2004 and 2010. 25 Figure 3. Unadjusted Kaplan-Meier Curves of 3-Year Outcomes According to Revascularization Strategy and SYNTAX Score. 26 Figure 4. Adjusted Survival Curves of 3-Year Outcomes According to Revascularization Strategy and SYNTAX Score. 27 Figure 5. Unadjusted Kaplan-Meier Curves of 3-Year Outcomes According to Revascularization Strategy and Diabetes Mellitus. 28 Figure 6. Adjusted Survival Curves of 3-Year Outcomes According to Revascularization Strategy and Diabetes Mellitus. 29 Figure 7. Unadjusted Hazard Ratios of Clinical Outcomes According to Selected Major Subgroups. 30 Figure 8. Patients Distribution and 3-Year All-cause Mortality According to Revascularization Strategy and SYNTAX Score. 31